Home

WINDOWS IMAGE MANAGER USER`S MANUAL

Contents

1. WIM User s Manual 63 6 5 Geo Operations 5 Get Vector Ojects This function loads vector objects such as points transects rectangles drifter tracks or vectors from a text file and displays them on top of an image allows to pick values from an image or calculate statistics of a predefined list of points transects rectangular areas or drifter tracks Note starting from version 6 46 WIM can read the ESRI shapefiles http www esri com library whitepapers pdfs shapefile pdf as other vector objects Currently the shape objects are written into the image or into the bitmap but they don t get into the Vector Objects table A sample shapefile USstates shp with US state boundaries is installed in the Maps folder For example you can have the coordinates longitudes and latitudes of a series of points specified in a text file and after loading the points as vector objects you will see the points plotted on the image After selecting a particular vector object it starts blinking and you can get statistics of the image corresponding to the location of a vector object e g 3 x 3 pixel window corresponding to a point object A related function Geo Read Vector Objects from HDF reads similar vector objects from a binary HDF file Each image has its set of Vector Objects that you can see in the Vector Objects table select View Vector Objects or click on the icon on the toolbar After completion the imported vector
2. The projections n3a n3b s3a s3b are specific implementations of the Polar Stereographic projection used with the SSM I products delivered by NSIDC see SSMI Products The Swath projection is specific to the SeaWiFS Level 1A and Level 2 products and shows the image in the satellite view projection SeaWiFS global mapped images are in the Global equal angle projection CoastWatch images can only be in Linear or Mercator projections The simplest conversion between image coordinates and an area of Earth is accomplished by the Linear projection which is acceptable for small areas In case of a Linear projection the coordinate conversion between the video screen and geographic longitude and latitude is performed with two pairs of coefficients A B and C D The top left corner of any image is assumed to have video screen coordinates of X 0 Y 0 with X increasing to the right and Y increasing to the bottom Longitude is assumed negative on the Western Hemisphere and latitude is assumed to be negative on the Southern Hemisphere Lon A B Xy Lat C D Y where X and Y are the screen coordinates pixel and line respectively The default coefficient values are zero If at least one of them is different from zero you will see the calculated longitude and latitude of the current pixel when you drag with the right button of the mouse See section 5 Useful hints how to find the linear geo conversion coefficients for your
3. small endian format cf the next section For displaying a float image the values are scaled linearly into 256 bins between Min value and Max value see View Settings General Color Scaling An easy way to find the minimum and maximum values and stretch the color coding between them is to use View LUT Stretch The minimum and maximum values can then be viewed in the View Settings General Scaling for Integer Float Some image operations are either not available or not convenient to use for float image buffers In that case it is recommended to convert the float image to Byte with Transf Convert and a suitable scaling type Float Bigendian Like the previous option Float Bigendian reads a Float32 raster image but in the big endian format The term big endian or small endian refers to whether the most significant bytes of a 4 byte float number are in the end or in the beginning The bigendian format files are usually generated by Sun SPARC SGI and others systems and are a popular format in the Unix world GOES SST A special version of the raster byte image with the upper left corner at 180W 60N and the lower right corner at 30W 45S These images are always 3000 columns by 2100 rows and in a simple Linear projection gc coefficients are respectively 180 0 05 60 0 05 The scaling of the SST is linear with slope 0 15 Intercept 3 15 When read as GOES SST WIM assigns the correct scaling and
4. Optionally allows to load a LUT file from disk or to save the current LUT into a file The Palette type is automatically changed to Custom Each of the Red Green and Blue LUT components can be modified either by the scrollbars or by mouse movements The resulting bitmap can be viewed by selecting Preview To test the operation check out spectrum ut after loading it with File Lookup Table Load LUT Scale to Clipboard Copies the current color table to the Clipboard as a bitmap You can then either save it as a bitmap or paste it directly into another program for combining your image with the corresponding color scale If Value Scaling options have been selected View Settings General Value Scaling real geophysical values are shown e g the CZCS Pigment concentration in mg m or temperature in C otherwise they are the pixel values It is essentially the same operation as Examine Color Table plus the transfer of the result to the Clipboard WIM User s Manual 36 6 3 View Options 6 3 View View Options Toolbar Show or hide the toolbar Status Bar Show or hide the status bar Image List Show or hide the List of Images dialog box Zoom Allows to zoom the image bitmap at up to 64X magnification This zoom operation acts at the bitmap level whereas the Transf Zoom actually produces a zoomed image buffer Annotate Writes a horizontal color scale at selected location into the image buffer an
5. The following is a list of files that are normally installed with WIM If you are short of disk space you may delete all the image files Only the exe files and the dll files are essential to run WIM wim exe wimLE exe hd414m dil hm414m dil wl1b dll gctp dll tilemap dil PointTypes xml Wim hip geotifcp exe Wim paf WAM pdf Look up tables LUT anomaly lut anomaly5 lut anomaly7 lut chl lut chl1 lut chl1 white end lut chi2 lut chi3 lut chl raw spectrum raw Igbp Ignd lut mld lut petes24 lut pigment lut spectrum lut sst lut sst_kuring lut Map datasets WIM main program the limited evaluation version of WIM dynamic link library for HDF files dynamic link library for HDF files dynamic link library for AVHRR Level 1B files dynamic link library for GCTP projections dynamic link library for MODIS Land projections extensible list of symbols for point objects in XML format old help file for WIM NOT updated with newer features utility called by WIM when saving GeoTIFF tags to TIFF files WIM User s Manual in Adobe Acrobat format User s Manual for WIM Automation Module WAM sample LUT files for anomaly images sample LUT files for suitable for ocean color and other types of images a sample raw LUT files Various sample LUT files e g one for the mixed layer depth mld or one used by the IGBP Global Land Cover Characterization program http
6. WINDOWS IMAGE MANAGER Image Display and Analysis Program for Microsoft Windows with Special Features for Satellite Images USER S MANUAL Version 8 May 2015 Mati Kahru WIM is a software tool for satellite image analysis under Microsoft Windows XP and Vista WIM Automation Module WAM is the library applications and tools for automating WIM tasks WimLE is the evaluation version of WIM that includes a limited set of functions For the latest version and other news please check the WIM World Wide Web home page http www wimsoft com For information or orders send E mail inquiries to wim wimsoft com WIM is subject to copyright and may NOT be copied without an explicit permission from the author Software Warranty The WIM software is licensed solely on an as is basis i e the entire risk as to its quality and performance is with the purchaser In no event will Mati Kahru be liable for direct indirect incidental or consequential damages resulting from any defect in the software DON Contents Like Na 7 Hardware and Software Requirements ccccccccssssseeseeeeeeeeeeeeeeeeeeeeenees 9 Installing WIM asa GANANG 10 Getting StOMO LanG KANAN ANA ANAN GAGA 11 5 WIM SS AA TA 14 6 MENU SY SCC UU AA NAA ERA AA NAAAGAAAAS 16 6 1 File File Operati nS ama ANAN AA AA ees 16 PW ANAN ANA 16 Filter to 9 BC gene eo BUAN 16 OG ce ccc GASA ANA ce sme eda ees AA AA 16 Copy LED AA 27 C t LE AA AA A
7. 0 725 1 10 um NDVI is calculated according to the formula NDVI Ch2 Ch1 Ch1 Ch2 and ranges from 1 to 1 A new image buffer is allocated with the corresponding pixel values equaling 700 NDVI 100 NDVI is a rough index of the amount of green plant biomass In general NDVI is negative for water near zero for clouds and bare soil and changes between 0 05 and 0 6 for vegetative surfaces Holben 1986 NDV has been used to detect thick surface phytoplankton blooms green soup of the blue green algae Prangsma and Roozekrans 1989 As NDVI for water surface is normally negative since water is nearly a black body at near infrared channel 2 positive values indicate dense accumulations of surface floating algae In fact positive values result only from very dense accumulations and even small negative numbers indicate surface floating algae In order to show negative pixel values 100 is added to NDVI values multiplied by 100 NDVI values less than 0 1 are converted to zero For testing the operation you can calculate the NDVI image for a sample pair of images channels 1 and 2 from sbaltic lan use File Open Erdas Lan to read it Note that due to the transformation the pixel values below 100 are actually negative As another example try it with sample images s califf img and s calif2 img of Southern California You can see reduced vegetation density or greenness in the Los Angeles and San Diego Tijuana urban areas The
8. 3 15 When read as GOES SST WIM assigns the correct scaling and projection These files can also be read as generic raster images img but then the user has to manually set the size scaling and projection Pixel values 0 2 4 are used as flags for Space land cloud The 24 hour files seem to be using only the flag with the pixel value 0 File names are of the form sst1 YYYY DDD HH for the hourly results and sst3 YYYY DDD HH for the three hourly results e g a 1 hour result is sst1_ 2002 062 14 anda 3 hr result is ssii 2002 062 12 Please note that these files have no extension In Windows the extension is used to associate files with the application e g WIM Therefore it is not possible to set up an association so that by just clicking on a GOES SST file the data will be automatically read into WIM Instead the File Open GOES SST has to be used WIM User s Manual 108 21 MERIS Products 15 Altimetry Products Maps of Sea Level Anomaly MSLA in the netCDF format obtained from data of various sensors TOPEX POSEIDON Jason ERS 1 2 Envisat can be read with WIM The maps are provided on a MERCATOR 1 3 grid Resolutions in kilometers in latitude and longitude are thus identical and vary with the cosine of latitude e g from 37 km at the equator to18 5 km at 60 N S Units are centimeters The files are in a NetCDF format and filenames are as follows hh msla oer tp or merged h date nc e g T P map for January 23 2
9. Be careful when editing the string as WIM may become unstable with certain settings oo BO Set Colors Allows you to specify the coloring sequence look up table LUT by the Start pixel value and the End pixel value Both parameters range from 0 to 255 If Start is smaller than End all pixels less and equal to Start will get the first color in the palette usually black and all pixels greater and equal to End get the last color of the palette usually white The pixels with values between Start and End will be colored by a linear sequence of the available colors Pixel values different from O but smaller than Start will get the second color in the LUT Respectively if End is less than 255 pixel values less than 255 but larger than End will get the second color from the end If Start is greater than End the opposite color sequence will be assigned If you are CT sa running more than one color intensive application Color Definition x simultaneously Windows may run out of available colors and has to replace colors on inactive windows by temporally matching them to the colors of active windows In 15 16 or 24 bit color modes the selection of colors is practically unlimited ja zi zl Pixel Values WIM User s Manual Geophysical Values 51 0 02 10 00 OK Cancel 6 3 View Options In case of a Byte image the values below the Start and End pixel values show the respective geophysical values For example
10. and Monthly compositing since 2002 Each dataset has corresponding files in JPG XML NX GZ and HDF GZ formats Please note that the datasets that are useable by WIM are those in netCDF format nc and not those in HDF After downloading the compressed files nc gz need to de uncompressed e g with the command gzip d gz After that they are directly readable with WIM e g by double clicking on the filename in Windows Explorer The XML files provide additional information about scaling etc but are not needed in WIM Each Level 3 file has the following datasets count mean stdev min max Most useful of those is mean The stdev is actually not the standard deviation but the sum of the squares of the input MERIS pixel values divided by the number of values The Level 3 data are provided on global sinusoidal equal area grid Global ISIN of approximately 9 km resolution 4320 x 2160 for the monthly images and of approximately 4 5 km resolution 8640 x 4320 for the daily images WIM User s Manual 118 Wind products 22 Cross Calibrated Multi Platform CCMP Ocean Surface Wind products This dataset is a multi satellite climate data record for ocean surface wind that spans nearly 21 years Atlas et al 2008 2009 The product is derived through cross calibration and assimilation of ocean surface wind data from SSM TMI AMSR E SeaWinds on QuikSCAT and SeaWinds on ADEOS II Cross calibration is performed by Remote Sensing Systems
11. on a new line If you select an existing file it s contents will be lost before new data is recorded The selected points are shown on the screen by a black line connecting them Up to 256 points can be shown on screen saving to the disk file can go on until there is space on the disk Do not forget to deselect the Point Save option when you don t need it any more by selecting the menu option again If the image projection is set see View Settings Projection the x and y coordinates recorded in the file are the Longitude and Latitude otherwise they are the screen video coordinates If Value Scaling has been selected View Settings General Value Scaling the recorded values are in their respective units otherwise they are the pixel values WIM User s Manual 57 6 4 Examine Operations Alih Profiles Plots the currently selected line or rectangular area In case of a line the beginning of the line is positioned on the left and the end of the line on the right In case of a rectangle you may get either plots of rows horizontal lines or columns vertical lines If the height of the area is greater than the width columns are plotted otherwise the rows are plotted In case of the column plot the left side on the plot corresponds to the top of the rectangular area If Value Scaling has been selected View Settings General Value Scaling the plotted values are in their respective units otherwise they are pixel
12. operation is used for the horizontally and vertically polarized microwave signals see Chapter 7 on SSM I Products To prevent division by zero the result is set to 1 0 if TBH TBV 0 Primary Production Calculates an image of phytoplankton net primary production NPP mg C m day based on the Behrenfeld and Falkowski 1997 Vertically Generalized Production Model VGPM and using images of surface chlorophyll a concentration Chl a mg m3 incident photosynthetically active radiation PAR Einstein m Day and sea surface temperature SST C Please see the following options dialog box for Primary Production calculations Currently there are options for photo inhibition temperature model of PP oot and euphotic zone depth calculation More options will be added in the future The original VGPM model used the Morel and Berthon 1989 euphotic zone depth parameterization as a function of surface chlorophyll That model was later revised by Morel and Maritorena 2001 In the Southern Ocean mixed layer is deeper at similar chlorophyll concentrations and a different parameterization Mitchell Kahru unpublished is available as an option Global and regional images of Chl and PAR are available among others from the Goddard DAAC hitp oceancolor gsfc nasa gov ftp html and SST images are available among others from the PO DAAC http oodaac jpl nasa gov sst It is assumed that the sequence of images is Chl a PAR SST and the us
13. the longitudes The longitudes here are converted by WIM from the 0 360 degree range to the usual 180 180 range WIM reads only the latitude and longitude values and ignores the rest Vector Vectors can be read from two kinds of files The type was usually set in View Settings Misc but currently the reading routine is expected to automatically recognize the vector type The default vector format is x1 y1 x2 y2 that specifies the start and end coordinates of the vectors either in image or geographical coordinates The image coordinates are in pixels Y increases from top to down X increases from left to right with pixel 0 0 in the top left corner and are assumed if the projection of the image is Unmapped In the projection is specified then the geographic coordinates are assumed x1 is the start longitude y1 is the start latitude x2 is the end longitude and y1 is the end latitude You can easily create a vector file even with an editor A sample vector file est256 vec in image coordinates for a sample image is provided The file starts with a comment line followed by vector data each on a separate line e g Comment 180 100 170 111 180 110 168 98 Another format is a specialized one that is used to plot the velocity vectors derived from the motion analysis of a sequence of 2 images Multi Motion Detect Here the format specifies the start coordinates x y in video coordinates the velocity components u and
14. values Optionally allows to save the values on the selected line or rectangle as ASCII integers in a file i RGB Image Uses three image bands for the Red Green and Blue components respectively to construct a true color image This is a special type of Int32 image in memory When right clicking on the RGB image the three components Red Green Blue are shown You need to stretch the Red Green and Blue bands separately in order to create the desired result The Low and High scrollbar levels for each band correspond respectively to the Min and Max values of the component pixel values between which the color scale is linearly stretched They are similar to the Start and End values in View Set Colors The selected three sets of Min and Max values are saved in the computers registry and are used as default values next time this operation is used In order to be close to a true color image the bands should roughly correspond to the red green and blue spectral bands For example with SeaWiFS images the true color images are usually constructed from bands 670 nm Red 555 nm Green and 412 nm Blue A sample image file sw Ila mapped byte hdf in the Imagesihdflseawifs folder of the WIM CD can be used for testing this operation Different features such as land turbid water clear water aircraft contrails clouds can be separated by stretching the Min and Max values for the R G B components When applying this operation
15. 0 For the MCSST and GLI Level 3 images the Shift is automatically set at 180 and the left side of the image corresponds to the zero meridian For the OCTS Level 3 global images produced by NASDA you have to manually set Shift 160 whereas the OCST Level 3 images produced by NASA no shift is necessary i e it is 0 See the following figure on the influence of the Shift parameter Shift 0 gt WIM User s Manual 46 6 3 View Options Shift 160 gt When saving the file as HDF file the Shift value will be recorded and will be set when the file is read again with WIM It is obvious that the Shift value of 160 is convenient when looking at imagery covering both sides of the international dateline e g the Pacific Ocean With Shift O the area of interest will be in two separate parts see the figures at the top Note that Level 3 in netCDF formats are provided in Grid projection which is in practice equivalent to the Global equal angle projection if using a constant longitude and latitude steps WIM User s Manual 47 6 3 View Options General Extensions Projection HDF Options Misc E Show HDF global LUT Stretch image attributes before loading after loading ho Minimal width 4 Prompt for geolocation 10 height of image Iv Ha file if proj info not found Select SDS buffers to tead when loading Mirror automatically acending orbit data Iv jv Save as compressed
16. 250749 42 656075 70 241949 42 650208 2 70 233148 42 517025 70 268937 42 524066 WIM User s Manual 62 6 5 Geo Operations Dy Get Map Overlay Reads a database file with coordinates and creates an overlay file with the selected features WIM includes low resolution global databases for creating overlays of coastlines gshhs_crude b country boundaries boundlow dat and rivers riverlow dat Corresponding high resolution 1 km files are provided in the Maps folder of the WIM CD WIM includes functions for reading GSHHS Global Self consistent Hierarchical High resolution Shorelines shorelines format Wessel and Smith 1996 GSHHS files have extension b and include datasets in 5 different resolutions in order of increasing accuracy crude low inter high full The resolution of the full GSHHS dataset is better than 100 m The sizes of GSHHS files change from 167 KB for the crude to over 87 MB for the full dataset The installation of the GSHHS full dataset can be skipped in the Custom installation option When creating the overlay the user can choose the level of shorelines to include e g whether to include only the land sea shorelines and or lakes inland water bodies or higher level shorelines like islands in lakes etc The time for scanning these GSHHS datasets for coastlines increases proportionally with the size of the dataset The additional fine scale details are not necessary and sometimes distrac
17. 2C otherwise they are the pixel values Scan Scans animates through the whole color palette It shifts the colors in the 256 color palette one step every predefined number of milliseconds 10 1000 between the currently selected color range Start and End To end the color animation select any menu This operation is not available for 16 color standard VGA and high color more than 256 colors modes Random As Color Scan but generates random color palettes Not available for other than 256 color modes To end the color animation select any menu Contour Lines Creates a new image buffer with contour lines in a specified pixel value for a selected range of pixel values To get a smoother looking picture the image is usually smoothed with a median filter before applying the contour finding algorithm Bigger kernel size of the filter produces stronger smoothing The filtering option for 2 and 4 byte per pixel image buffers is available starting with WIM version 6 12 If Copy source is selected the image itself is copied to the new image buffer otherwise a blank image buffer is used to draw the contour lines Having the contours on a blank image gives you more options e g you may want to manually edit the contour lines with Edit Draw before you overlay the contours on the image with Multi Overlay Image You have to select the Start End and Step values for which the contour lines are drawn into the image buffer The Start En
18. AA ae eee Ra ia 79 Ste oe aa ama AA AA AA 79 Salo NE AL AA PAA AA ARA AA PAANE RP 79 ANU aa A AA AA 80 PAGO APA AA AAO 80 6 7 Multi Multiple Image Operations 11 00700000usanamanasasasasasanansraaaaa nna 81 Pao ko 2 gt aril ANA AANI 81 Band gt i NANANA NAA ANA 81 NM ME a A AE AA AA AA T 83 ll AA AA AA AA 83 WIM User s Manual 4 1 Introduction BULA AA T PAANO 84 Divide WIM ne ie a Pe aaa een eee err 84 Insert Imaje iaaa ene eee cee ae ee 84 Linear CoD PAPA AA AA AA AA AA 84 Mask Ww Image NANANA va savass causes seca aa vevscduneeet ceased es anaeseecurteess 85 Motion BI J L o iicascccs A seca ccs KAANAK BARA 85 LC 3 ee re ee geo ee RPC OP ne Pe eRe ee ee Peet eer res 86 Overlay Mage seston Reinet siete 86 Polarization FRAUUO BANAAG AA 87 Primary Produc NANG ai GE 87 SIMA OG RANI NG sissies aceasta AA S 88 SITU IM AGG ANAN NAAN KANAN 89 SST GNA ONS eich NATAN eth NPA aa cake ANA bA a gn art 90 Subtract Mage i Pe Tn se aaa 90 T rbidity chl Ra RA 90 Vegetation Index PARAAN APA 91 Ba PAGA AA AA AA AA AA T 92 1 Eze E2 0 APAPAP AA APAT PA PAA PP MAMBO 92 91110 0 NAAPI PANAMAN PANAMA P PRAAN PAMANA HPNPA MAN PARNANAPHPNPAM AN PAMANA ikana KAA AKANA KAA Nania 92 Clean 2 AASA AA 92 De r arar a vd oxi gal aA ra aa bx vale AAPAAA 92 FMR TATATATA AT TT 92 SUED PAA AA AA AA AA ASAP A AA 92 6 9 Segm Segmenlallon an ea aaiae raaa aaa ieee 93 Find BOGGS osc rh ets eee ee AA cheered 93 TANA
19. Atmospheric correction is usually performed to approximate the water column reflectance defined as Ry 7 Ly A Ea 2 where Ly is the water leaving radiance and Ey is the downwelling irradiance entering the water is the spectral band e g Stumpf 1992 While the full atmospheric correction is beyond the scope of a general program like WIM a simple approximation is performed here by assuming Ey 1 r cos To where E is the solar constant r is the normalized Earth Sun distance 9 is the solar zenith angle for calculation see Stumpf 1992 and T is the transmission through the atmosphere T A exp t A 2 Tg 4 COS9 where qt is the Rayleigh optical depth tg is the gaseous absorption optical depth i e for ozone and water vapor Then a simple correction is performed by dividing the pixel value by Eg The optical depths are specific for each sensor and can be modified with the View Attributes dialog box The default values for NOAA AVHRR bands 1 and 2 are taken from the DECCON program by Stumpf and Townsley Both for a 1 or 2 byte pixel source image buffer the output will be a 1 byte pixel buffer For 1 byte pixel buffer the correction is not performed for values O and 255 i e for out of range values When calculating the output pixels are scaled according to the albedo minimum and maximum values View Attributes If no albedo range has been specified the range is assumed to be from 0 0 to 25 5 and the corr
20. CHLO in ImagesihdaflSeaWiFS Level3 If you load images with the Open icon on the toolbar menu File Open you need to pay attention to the type of images that you are trying to read WIM can read and write many types of images but you have to select the correct Files of type in the dialog The next time you are trying to read or write an image the type used last time is already selected as default for you In each folder you will be shown only the files that have the appropriate extension for the selected file type You can see all the files by typing in the File name edit box If your files are not shown then you can add your extensions to the default extensions in View Settings Extensions To look up pixel values hold down the right button of your mouse and move the mouse on the image In the top part of the image window frame you will see values in the format X Y Lon Lat PV SV where X and Y are image coordinates X distance in pixels from the leftmost column Y distance in pixels from the top row Lon and Lat are the corresponding longitude and latitude note the order the pixel value PV and the scaled value SV in geophysical units If the image has no projection then the corresponding Lat and Lon are not shown Select an area of interest e g a rectangle by moving the mouse with the left button down Assuming that you successfully opened composite hdf select an area of interest and apply the Histogram function Hist
21. Manual 89 6 7 Multiple Image Operations SST ch4 ch5 Calculates the sea surface temperature image from AVHRR channel 4 and channel 5 images It assumes that the current image is the channel 4 image and prompts for the channel 5 image number The calculated SST image is put into a newly allocated image buffer The source image values are interpreted depending on the Value Scaling flag in the View Settings General dialog box if either Pixel Value or x 10 is selected the image values are assumed to be temperature values multiplied by 10 if Value Scaling SST SMHI is selected it is assumed that the SST SMHI transformation is used see Transf Convert The brightness temperatures of a single channel are influenced by changes in the atmospheric water vapor The split window method McClain et al 1985 uses the values at two different channels to correct for the atmospheric interference Here the coefficients estimated by Coll et al 1991 are used T T4 1 54 0 22 T4 T5 T4 T5 where T is the true sea surface temperature 74 and 75 are respectively the brightness temperatures in AVHRR channels 4 and 5 Subtract Image The operation is similar to Multi Difference It subtracts a specified image from the current image pixel wise The images must be of the same size Value Scaling see View Settings General is used for calculating the real value before the subtraction The resulting image is created a
22. NAVOCEANO format The US Naval Oceanographic Office NAVOCEANO provides near real time SST datasets of the to calculate SSTs at a near real time rate The Level 2 datasets are converted to either individual or composited images by WIM The NAVOCEANO Level 3 datasets are read as five sequential global images WIM makes a distinction between raster images and bitmap type objects e g bitmaps GIF PNG JPEG and other similar types Whereas in raster images the emphasis is on the digital value of a pixel a number does not have a color the purpose of a bitmap object is its visual the appearance with colors There are many programs that deal with bitmaps MS Paint has been a perennial Windows companion WIM is not a program for reading and visualizing bitmaps WIM performs operations on images represented by numbers and not on bitmaps WIM transforms digital raster images to bitmaps and can save as BMP GIF JPEG PNG TIFF and other bitmap type files but not vice versa WIM User s Manual Hardware and Software Requirements 2 Hardware and Software Requirements WIM is designed for the current Windows operating systems Support for the older Windows versions currently Win95 Win98 and WimMe is being gradually dropped and the latest versions are preferred currently Windows XP WIM depends on many system components provided by Windows Many of those components have undergone significant changes since Windows 95 As WIM tries
23. al 1999 The Power Law band ratio is a simple Linear regression in the log log transformed space A simple Chl algorithm uses typically the blue and green bands of the are normalized water leaving radiances Lwn 443 and Lwn 555 A simple CDOM colored dissolved organic matter algorithm was proposed Kahru and Mitchell 2001 for SeaWiFS using Lwn 443 and Lwn 510 CDOM 10 0 393 0 872 log Lwn443 Lwn510 where CDOM is expressed as CDOM absorption at 300 nm in m These coefficient values are the default values when the Band Ratio function is used for first time When changed the new coefficient values will retain their value in your computer s registry for easy access next time For OCTS bands Lwn 443 and Lwn 520 the coefficients are Coeff1 0 411 and Coeff2 0 703 The resulting image of the Band Ratio calculation is always a 4 byte per pixel FLOAT image As it is hard to guess a convenient color range for the newly created FLOAT image as a first step you can do View LUT Stretch to stretch the colors to the range between actual minimum and maximum pixel values A FLOAT image can be transformed to the more convenient BYTE image with Transf Convert BYTE and choosing a convenient scaling For Chl use the Log Chl scaling For CDOM a simple Linear scaling with Slope 0 01 and Intercept 0 0 is usually OK but limits the upper values to 2 55 Another important aspect is filtering out pixels that are out of the Va
24. another custom LUT is loaded is the grayscale The raw LUT file is a binary file of 768 bytes organized in the following order 256 red values 8 bit integers 256 green values and 256 blue values The raw LUT format is the same as the raw palette format of the Hierarchical Data Format HDF for 8 bit raster data A sample raw LUT file included with WIM is spectrum raw provided with the HDF code and utilities which is another continuous spectrum lookup table A popular LUT file for chlorophyll and other images in the raw format Chl raw is included You can create your own raw LUT files by modifying the current LUT and then saving with File Save LUT raw Save LUT Saves the current look up table LUT to a LUT file The LUT file is an ASCII file with the structure given under File Load LUT You can interactively edit the current LUT with Edit LUT Edit Save LUT raw Saves the current look up table LUT to a raw LUT file The raw LUT file is a binary file of 3 times 256 bytes consecutive 256 bytes for the Red Green and Blue components respectively The raw LUT format is the same as the raw palette format of HDF for 8 bit raster data You can interactively edit the current LUT with Edit LUT Edit Save Info Saves the current image size and optionally geographic projection parameters to an info file corresponding to the current image The info file is not used for image file formats that have all the necessar
25. are easiest to work with They are provided in spatial bins of 4 km 36 km and 1 degree and temporal bins of 1 day 8 days month year see http acdisx gsfc nasa gov data dataset MODIS 03 Ocean 03 Level3 Mapp ed index html Monthly and other composites can be downloaded or generated with WIM Multi Composite MODIS Level 2 products are in the highest possible resolution 1 km for ocean channels 250 or 500 m for some land channels but have the disadvantage of being in the satellite view projection In contrast with SeaWiFS Level 2 products that include algorithms for calculating the latitude and longitude of each pixel MODIS provides a separate product MOD03 that provides the many ancillary parameters such as the zenith and azimuth angles for the sensor and the sun satellite location etc Most significantly the latitude and longitude arrays are provided in MODOS files for the corresponding MODIS Level 2 products The latitude and longitude arrays LLA can also be supplied in the MODOCQC product group and GSUB subset files After selecting the Scientific Datasets to be read from MODIS Level 2 products if the latitude and longitude arrays LLA are not found in the same file the user has to select a file where the LLA e g the corresponding MOD03 or QC product This can be quite a challenge as the filenames are very long and hard to memorize Starting with WIM version 6 15 a smart algorithm is being used to help the user to locate the ri
26. by the user Projection parameters and a few other characteristics of the image NOAA satellite number year month day hour minute are also read from the CCAR image header Baldwin and Emery 1993 Emery 1995 Correct time of the image is needed for making the solar correction Transf Solar Correction and calculating the turbidity index with Multi Turbidity ch7 ch2 When saving a CoastWatch image as HDF the CoastWatch header data is automatically converted to HDF attributes these attributes e g the geo referencing attributes specifying the projection are then used by WIM when the HDF file is read The Copy TO function makes it possible to selectively copy attributes from one image buffer to another This useful for example when remapping an image to the projection of another image and creating a new remapped image The problem of which attributes to get from the source image and which attributes from the target projection image is not easily solved automatically For Level 1B files the Attributes function is replaced with the diagnostic log of processing of the Level 1B file The following example shows the global and local attributes of a SeaWiFS Level 2 image WIM User s Manual 39 6 3 View Options Attributes of chlor a Attribute Name Owner Type Orbit Node Longitude lt GLOBAL gt fHoat32 116 097366 Sensor Characteristics lt GLOBAL gt charg Number of bands 8 number of ac Dat
27. by a single bit i e with possible values either O or 1 The bits have to be grouped into 8 bit bytes The 1 bit per pixel image here called overlay is converted to anormal 1 byte per pixel image buffer The pixels with their corresponding bit 1 will make a byte value 1 while zero bit will produce a byte value of 0 If instead of the value 1 you would like the pixels to have values 255 you can use Transf Binarize 7 to make the conversion You can then overlay the overlay image to your background image by using Multi Overlay Image NB When an image was compressed to an overlay file consecutive 8 bytes were compressed into one byte If the number of pixels in lines was not divisible by 8 the image size had to be increased It is therefore recommended to make overlays only from images with a multiple of 8 pixels per line A sample overlay image est256 0vI corresponds to the sample image est256 img It has to be noted that while 1 bit per pixel overlays provide 8 times compression for binarized images RLE encoding see File Open Compressed RLE and File Save As Compressed RLE may give even better compression rates while keeping all the 8 bits per pixel Subset Reads a subset from a 1 byte per pixel image file Sometimes images and the corresponding image files are very big and even if you have a lot of RAM in your PC the Windows functions that allocate RAM for WIM do not return enough RAM to accommodate both
28. current image dimensions from List of Images box or View Settings General Image Size If your image has different size then before reading the image specify the correct size in View Settings NB If WIM finds an info file for the selected image file the same name with the extension inf the image dimensions DX DY are always read from the info file disregarding the specified values It is convenient to make info files for your images if their size is different from the default size of 512 x 512 use File Save Info An info file contains at least two numbers DX and DY the number of pixels in a line and the number of lines and optionally the four geo conversion coefficients Problem You specify the correct image size before loading it in View Settings General Image Size but after reading the image from a file the size is changed to different values Hint You have an info file with the same image name and with the extension inf If WIM finds an info file for the selected image file the image size DX DY is taken from it Delete the info file or correct it with an editor When reading Erdas Lan file CoastWatch file or HDF file then the image dimensions are obtained directly from the file itself header thus overriding any default or specified values Problem While specifying a rectangular area on the image by mouse e g for cutting part of the image calculating statistics or histogram of an are
29. image Normally the coefficients are read from an info file inf but you can change the coefficients manually in the Settings Projection dialog box In case of geographic projections other than Linear up to 4 extra reference variables are used If any of the reference variables is different from zero they are shown near the bottom of the View Settings Projection dialog box ref1 is a reference latitude ref2 is a reference longitude ref3 is another latitude and ref4 is another longitude Some projections do not require all the reference variables The reference variables are read from the info file see File Save Info For more on projections see Geo Remap Proj WIM User s Manual 48 6 3 View Options HDF Options Current Settings Ed General Extensions Projection HDF Options Misc Show HDF global LUT Stretch image a attributes before loading E after loading Minimal witdh of Prompt for geolocation au image to be read M file if proj info not found Vv Select SDS buffers to Vv Mirror automatically read when loading acending orbit data JV Save as compressed Show HDF global attributes before loading if selected shows View Attributes before loading the images LUT Stretch HDF Images after loading checkbox if selected automatic LUT stretching between the minimum and maximum pixel values is performed after loading HDF SDS images see View LUT Stretch Minimal width of
30. image to be read WIM will discard arrays narrower than the specified number of pixels as images HDF files often fcontain various arrays of numbers that are not supposed to be images but to provide other auxiliary information e g calibration coefficients These arrays are usually narrow e g a certain number of calibration coefficients per scan line and can be discarded from being displayed by WIM Prompt fro geo location file if proj info not found if selected and WIM cannot recognize geo location projection of the image prompts for a separate geo location file name Select SDS buffers to read when loading when selected this option gives you a list of all image buffers Scientific Data Sets that can be selectively read into RAM When not selected all image buffers will be read into RAM Mirror automatically ascending orbit data when this option is NOT selected ascending orbit data will be upside down When selected and the image is detected as ascending orbit it is automatically mirrored Save as compressed if selected the image data inside the saved HDF file is compressed with the with the Lempel Ziv 77 dictionary coder that can reduce the file size by many times For typical images with large areas of uniform pixel values e g due to no data or clouds the compression ratio is WIM User s Manual 49 6 3 View Options 10 15 times For images with highly variable pixel values the compression ratio is
31. image with values over 255 then you need to first use Transf Convert to create either a Int16 2 bytes per pixel or Float 4 bytes per pixel image and then apply the linear transformation to that image For example when creating map overlays Geo Get Map Overlay the result is a Byte image with values up to 255 If you want the coastlines to have pixel values higher than 255 then you can use Transf Convert and Transf Linear Trans and multiply by e g 255 Mirror Makes an in place mirror transform of the current image over the horizontal vertical or both horizontal and vertical axes Repeating the same operation again restores the original image The image name is modified when the operation is first applied Note that in order to get the image what you see when you turn your head upside down you have to mirror it both over the horizontal and vertical axes Reduce Image Makes a reduced image from the current image Pixel values of the new reduced image are assigned by picking the nearest neighbor or by using the sum maximum minimum or average over the source window E g if the reduction is by a factor of 2 the operations will be done over non overlapping windows of 2 x 2 pixels When the sum over the window gets larger than 255 the new pixel value will be 255 Due to a requirement of the Windows memory management the width X of the new reduced image has to be divisible by 4 If Xis not divisible by 4 X will be increa
32. in case of a Byte image with the logarithmic Chlorophyll scaling Log Chi the Start and End pixel values of 20 and 200 correspond to Chl concentrations of 0 02 and 10 0 mg m see example on the left In case of 2 or 4 bytes per pixel integer or float the pixel values are first scaled with the Min and Max values from View Settings General Color Scaling to be between 0 and 255 After that the color coding is similar to the 1 byte per pixel images The Start and End values are saved as WIM Color Stretch attributes with Save as HDF Vector Objects Shows the table of Vector Objects The Vector Objects table consists of a number of the following vector objects Points Transects Rectangles Labels Drifter Tracks and Vectors and the following buttons Add Edit Delete Close Save Save Lat Lon Value Show Profiles Statistics The Add button allows a manual creation of a vector object Usually vector objects except the Label are imported with the Geo Get Vector Objects function from simple ASCII text files or with the Geo Read Vector Objects from HDF from HDF files recorded previously with the Save button in this function An ASCII text file with the longitudes and latitudes can be created with an editor such as Notepad exported from a spreadsheet program like MS Excel or created by the WIM function Geo Distance or with Examine Point Save An example of a Vector Objects table is shown below Each image has its ow
33. mean of the non negative pixel values The Max and Mean Pos filters can be used for example to smooth and fill gaps in sparse images generated from the Lat Lon Value ASCII input files Filter to Disk Performs filtering operations similar to the Filter option above but instead of working with image buffers in RAM transforms a disk file to another disk file This operation is needed when the image is very large and impossible to load into RAM For example a full scene image of the Landsat Thematic Mapper single channel has 7942 x 6800 pixels With each full image occupying about 50 MB it may be difficult to load the whole scene into RAM on a low end PC By first filtering the image file to another disk file and then sub sampling it at lower frequency File Open Subse it is possible to get a better representation of the large image WIM User s Manual 77 6 6 Image Transformations Gradients Calculates a gradient image of the current image using either a 3 by 3 Sobel filter summing absolute values of the vertical and horizontal convolutions or by picking maximum centered differences in all possible 4 directions Linear Trans Uses 2 float coefficients A B to calculate a new image as a linear transformation of the current image New pixel value A B old pixel value If the current image is a 1 byte per pixel image the resulting pixel values are capped between 0 and 255 If you want to create an
34. objects are shown with View Vector Objects By clicking on a vector object in the Vector Objects table the selected vector objects starts flickering by changing its color Geo Tracking Xx m Tracking type Coordinates Point Standard C Pixels XY Transect Geographic Lon Lat A Rectangle Float Lon Lat Drifter Track Jv Thick Line nfter Track V Thick Lir C Int Degree Float Minute Vector m Drawing Into C Bitmap Only Image Buffer Pixel Value 255 OK Cancel Drawing Into The selected vector objects can be marked in the screen bitmap or actually recorded with a selected pixel value into the image buffer WIM User s Manual 64 6 5 Geo Operations When marked only in the Bitmap they will disappear the next time the screen bitmap is rebuilt from the image buffer When recorded in the image buffer the actual pixel values in the image buffer are changed to the selected pixel value The changed image can then be saved Note that when you change pixel values in the image and look up the pixel values the pixel values that you select will not be the original ones but the values that you assigned yourself Therefore it does not makes sense to do statistics over the pixels that you have assigned yourself When reading text files with vector objects the first line in the list is usually regarded as a comment line and is skipped The coordinates can be specified either in the
35. of 705 kilometers in a near polar near circular Sun synchronous orbit imaging the same 183 km swath of the Earth s surface every 16 days The L1G product is a radiometrically and systematically corrected LOR image The format of individual bands is simple binary raster with 1 byte per pixel WIM Image In order to use File Open as Image you typically need to know the size DX DY of the image The size geo referencing and other information is available under the Description files link By clicking on that link you can download a compressed set of files tar gz that you can uncompress and untar into the same directory where the uncompressed image files are A particular file productdescription self has all the metadata about the particular Landsat scene You can get the size of the images from there but a better way is to read as HDF files the launch pad HDF file HDF L1G that has all the links to the individual band images Just make sure that HDF L1G is listed as one of the extensions of HDF files in Current Settings Extensions If not please add it by appending sHDF L1G into the HDF box Geo referencing of these files is currently done using the upper left and lower right coordinates in the productdescription self file corresponding to the particular Landsat scene A Linear projection with corresponding coefficients is automatically produced Due to various non linearities and errors this geo referencing is n
36. perform resolution sharpening of MODIS bands can be found in the Practical exercises with WIM and WAM WIM User s Manual 88 6 7 Multiple Image Operations Image Sharpening l x Full resolution Band 1 0 Interpolated Band 1 1 Interpolated Band 2 2 cns Shift Image Sometimes you may notice that when you create a coastline overlay for an image and lay it on top of the image the image is shifted relative to the coastline or vice versa With this function you can shift the coastline relative to the image so that it fits the image When you are satisfied with the fit the image is shifted by a discrete number of steps in X and or Y direction This is a very simple geometrical correction and is useful only when the shift is relatively small and uniform for the whole image In order to fit an image to a standard overlay e g coastline the overlay a contoured image is moved in the X and Y directions over the image until the user accepts the fit by pressing Enter The overlay contour is taken as the template according to which the image will be corrected It is inevitable that when shifting the image inside its frame some borders of the image will have no values and will be converted to zero This operation is used when you have a set of images of the same area that you want to make to match to each other as good as possible You need to have a standard overlay contour map of the area or you have to create
37. solar zenith angle needed for this function is calculated for WIM User s Manual 90 6 7 Multiple Image Operations the center of the image based on the timing of the image valid for relatively small images For the timing the Julian day GMT hour and minute as specified in View Attributes are used If the Julian day is O it will be calculated from the year month and day The year is specified with the last 2 digits e g 94 for 1994 The longitude and latitude range View Attributes is used to estimate the geographical coordinates of the center of the image The default optical depths due to the Rayleigh scattering t and gaseous absorption tg i e for ozone and water vapor are estimated by Stumpf and are the same as in the DECCON program A new 1 byte pixel image buffer is created with the turbidity index values The output is always a 1 byte per pixel buffer For 1 byte pixel input buffer the correction is not performed for values O and 255 i e for out of range values When calculating the output pixels are scaled according to the albedo minimum and maximum values View Attributes If no albedo range has been specified the range is assumed to be from 0 0 to 25 5 and the corrected albedo values are multiplied by 10 0 Vegetation Index Finds the Normalized Difference Vegetation Index NDVN from two images It is assumed that the current image is from AVHRR channel 1 0 58 0 68 um and another image is channel 2
38. some distortion Geo referencing of these images is possible by using the values in the productdescription self file but is currently not very convenient as it requires the user to calculate the Longitude and Latitude coefficients from the coordinates of the center and corner points A more convenient function for geo referencing may be added to WIM in the future WIM User s Manual 107 GOES SST Products 14 GOES SST Products The Sea Surface Temperatures have been derived from the series of NOAA s Geostationary Operational Environmental Satellites GOES Data is provided at a near real time rate as 1 hour 3 hour and 24 hour gridded files at a spatial resolution of 6km http podaac jpl nasa gov noaa goes Three hour and 24 hour files are averages of the 1 hour derived SSTs The Level 3 product includes data from both the GOES East GOES 10 and GOES West GOES 12 satellites Data is processed at the National Environmental Satellite Data and Information Service NESDIS and is available from ftp podaac jpl nasa gov sea surface temperature goes NOAA These files are a special version of the simple raster byte image with the upper left corner at 180W 60N and the lower right corner at 30W 45S These images are always 3000 columns by 2100 rows and in a simple Linear projection Longitude coefficients are 180 0 05 Latitude coefficients 60 0 05 The scaling of the SST is Linear with Slope 0 15 Intercept
39. the image buffer and the bitmap In these cases you can read only a part of the image and still be able to analyze and view it You can select the image area that you want to extract and select which pixels you want to pick i e every Nth pixel in X and every Mth pixel in Y direction Please note that both pixels in a line as well as lines in an image start from 0 E g If your big image is 4096 x 2048 in size and you want to subsample the entire image by picking every 4th pixel in both X and Y that produces a subsampled image of 1024 x 512 the default values of the ending range are 4095 and 2047 respectively Similarly if you would like to subsample the upper left corner of 1024 by 1024 your ending range is 1023 and 1023 respectively Due to a requirement of the MS Windows memory management it is recommended that the width X of the resulting image is a multiple of 4 If your final X dimension is not a multiple of 4 WIM will increase it or if it reaches the right bound of the image decrease it This operation is commonly used with images from the Landsat Thematic Mapper TM A full scene of TM data has 7 bands of 7942 times 6800 pixels With each full image occupying about 50 MB it is often not possible to view the whole scene or make a RGB composite from a selection of 3 bands on an average or low end PC A convenient approach is to first read the individual images at lower resolution e g every 10th column and every 10th ro
40. the image into geo physical values Simple raster images WIM User s Manual 16 6 1 File Operations with no additional information may need additional information e g the size of the image that can be read from an optional info file Before allocating a new memory buffer for the image WIM first checks if there is an info file in the same directory for the selected image file The info file is an ASCII text file with the same name as the image file but with an extension inf containing DX DY and optionally value scaling and parameters for geo registration i e converting between image X Y and Earth longitude latitude coordinate systems Note If an info file exists and contains data the number of pixels in a line DX and the number of lines DY are retrieved from it disregarding the current values see View Settings General Image Size With no info file in the same directory the current image dimensions are assumed for the new image Before actually loading the image the user can change the image dimensions The new image buffer of DY times DX bytes is allocated and the number of buffers in use is increased by one The default image size during start up is 512 x 512 You can change it with Edit Size or View Settings General Image Size If your images are of variable size it is convenient to save the info files for each of them File Save Info Info files can also be created manually e g
41. the name of the projection should be indicated on the following line Only the first letter of the projections name either capital or small is used The available projections are Mercator Polar Stereographic Transverse Mercator Conic Albers equal area Equidistant Cylindrical n3a n3b s3a s3b The last 4 are actually specific forms of the polar stereographic projection used for the SSM I products see Chapter 8 The next line should have the latitude and longitude of the center of the image and the latitude range in degrees decimals Two reference parameters are needed for the Mercator Polar Stereo and TV Mercator Transverse Mercator projections and four parameters for the Conic Albers and Equidistant cylindrical projections They should be on the line following the line with the center latitude and longitude A sample info file for an image in the Mercator projection and no Value Scaling Pixel value looks like that 256 256 119 248047 0 009727 34 853882 0 008081 Value Mercator 34 0 118 000000 2 500000 0 000000 0 000000 0 0 0 0 Sample info files for images in the Conic projection are given by s_calif1 inf and s_calif2 int WIM User s Manual 33 6 1 File Operations The geometric projections follow the information by the US Geological Survey Snyder 1982 and are compatible with the CCAR navigate program Baldwin and Emery 1993 Emery 1995 It has to be noted that with the increasing populari
42. the standard overlay image yourself from the contours of one of them using edge detection thresholding zero crossing etc This overlay will then be used as a reference according to which all the other images are shifted To use the operation make the image that you want to shift as the current one by clicking on it and have the overlay also in a memory buffer After you have specified the overlay index the number of its memory buffer the contours of the overlay will be drawn as white pixels on the current image By pressing the arrow keys the overlay is shifted by one step in the respective direction after each key action Combining the arrow key with the Shift or Control key makes a shift of respectively 5 or 20 pixels in the corresponding direction When pressing PageUp or PageDown keys the shift is by 10 pixels up or down When pressing Home the shift is 20 pixels up and 20 pixels left When pressing End the shift is 20 pixels down and 20 pixels right To end the positioning press Enter ESC or Space You may now choose to rebuild the image using the shifting information and save the corrected image for later use If you know already the shift that you want to make instead of the overlay index buffer number specify a negative number or a number that is larger than the last image buffer Positive number for a shift in the X directions means shifting to the right positive shift in the Y direction means shifting down WIM User s
43. v in the x and y directions respectively and the cross correlation coefficient of the calculation The value of the cross correlation coefficient gives an estimate of significance of the derived velocity vector If the value is high the vector is estimated with a high degree of confidence The format is the following x y u vV XCOrr The first 3 lines are not used and contain 1 a comment 2 the names of the parameter file and the two image files 3 the numbers of blocks where the maximum correlations were found image sizes and coefficients that can be used to convert the u and v speed values back to pixel separation in the x and y directions Here is an example Comment param par imagel pic image2 pic 20 20 256 256 0 7845 0 7845 180 100 12 5 12 5 89 190 110 11 0 10 0 90 WIM User s Manual 69 6 5 Geo Operations Note that the velocity values are given in relation to the mathematical coordinate system where positive u means direction to the right and positive v denotes upward direction on the image Note that the x and y coordinates are integers or may be floating point numbers meaning the longitude and latitude The velocity and cross correlation are float numbers Before the vectors are plotted you are given an option to change their length by specifying a scale factor to which the vectors are multiplied and to select a threshold value for the cross correlation coefficient float numbe
44. valid pixels is accumulated only for methods i and ii to create an image of valid pixel counts i e the number of valid values used in the averaging process of the composite When using the Last pixel method for compositing it is assumed that the images are sorted in increasing time For example if using images numbered 0 to 10 it is assumed that image 0 is the earliest and image 10 is the latest The Last pixel method is useful when you are interested in the current situation but some of the areas on the latest images are covered by clouds or unusable for other reasons The pixels missing on the last image are therefore retrieved from the nearest previous image Composite Series x Image Range First Image fo Last Image ft 0 m Composite Average C Maximum Last m Valid Range Min joo Pixel Values Max jes Geophysical Values OK Cancel You can specify the range of valid pixel values Valid Range Min and Max as often some pixel values e g 0 or higher end values near 255 are used to indicate missing or invalid data and should be excluded from the calculation of averages The valid pixel range can be specified either in pixel values or the real geophysical values e g pigment concentration or SST Images with different pixel size can be used The type of the Composite image is the WIM User s Manual 83 6 7 Multiple Image Operations same as the first image used in the series or f
45. ways The most common approach is to create a coastlines image with Geo Get Map Overlay and then use Edit Draw to create non zero areas Up to 255 different masks can be created on a single image using mask values from 1 to 255 The only restriction is that the different masks cannot overlap Please note that if you use pixel value 255 to create the coastlines then you should not try to use mask number 255 unless you really want to calculate statistics for the coastlines Time series Makes an time series plot by extracting the currently selected line or rectangular area from a set of images currently in memory The set of images have to be of the same size and form a sequence The vertical plot range is stretched between the respective minimum and maximum pixel value The respective means are connected by a line whereas individual pixel values are shown by x Optionally allows to save the pixel values in an ASCII file where the number of lines equals to the number of pixel values in the selected area of interest and the number of columns equals to the number of images selected You can then import the saved file into another program for further analysis or plotting X Y Scatter Makes an X Y scatter plot from the currently selected line or rectangular area of the current image versus respective pixel values of another image The pixel values of the current image are assumed to be on the X axis The image number of the pixel values put on the Y ax
46. with an editor by writing the respective DX and DY into them If you have specified a header length different from O View Settings Misc Image Header the specified number of bytes is skipped from the beginning of the image file In Windows bitmap rows must contain a multiple of four bytes If DX is not a multiple of four the bitmap rows are padded on the right to insure this It is therefore a good practice to use a multiple of 4 for DX The following file types can be selected Image Reads unformatted raster image with one unsigned byte per pixel Before actually reading the image buffer WIM checks if the buffer contains certain header information corresponding to the CCAR navigate software Baldwin and Emery 1993 Emery 1995 If that is true then the user may either accept these parameters image dimensions and projection parameters or cancel and discard the header information NEC Image A special version of the raster image with one unsigned byte per pixel corresponds to the North East coastline NEC of the United States These images are always 1024 x 1024 bytes and the sequence of bytes starts from the bottom row in contrast to a standard Image that starts from the top pixel row The files containing images in NEC format should have the extension nec NEC images are assumed to be in the Lambert Conic projection ASCII Reads an ASCII text file of pixel values into a Byte image with each pixel value from O to 255 re
47. 0 000000 0 000000 0 000000 Cancel Unify Geo coeff This function copies the projection parameters geo conversion coefficients see View Settings Projection of the selected image to the rest of the images in memory This function is useful if you have a number of images in the same projection and you want to transfer the projection parameters to other WIM images You can save the geo coefficients and other parameters of to as text in a corresponding info file with File Save Info Unify Vector Objects Is used to copy the vector objects of a selected image to all the other images in memory This function is useful if you have a set of vector objects that you want to apply to many images By using this function you can read the vector objects once and then apply them to all the images in memory 5 View in Google Maps This function opens a new window and displays the image boundaries in Google Maps see http www google com maps with a wealth of WIM User s Manual 72 6 6 Image Transformations geographical information You must have active internet connection for this function to work gt View in Google Earth This function opens Google Earth must be installed see http www earth google com and displays the image You must have active internet connection for this function to work This function creates a temporary KMZ file in your temp folder and runs loads it with Google Earth Note that you can als
48. 002 hh msla oer tp h 19015 nc Mapping errors in percentage of signal variance are provided on separate files More details are provided in the SSALTO DUACS CLS CNES near real time altimeter processing system handbook available at http www jason oceanobs com documents donnees duacs handbook duac s uk pdf The Jason 1 data is downloadable from ftp ftp cls fr pub oceano enact msla j1 A list of various altimetry data is available at http www aviso cis fr html donnees welcome uk html For example the 1 degree monthly average MSLA data from the merged TOPEX POSEIDON Jason ERS 1 2 dataset is available from the following FTP side ftp ftp cls fr oub oceano enact msla merged monthly average The 1 degree data are compatible with the WIM Global Equal Angle projection The netCDF files should be read with WIM as regular HDF files They have the nc extension You can select WIM as the application to open nc files so that you can read them by clicking double clicking The original float32 data are converted to Byte data with Linear scaling and no appreciable loss to the quality of the data As Sea Level Anomaly SLA values can be both negative and positive the full scale is set to 50 to 77 5 cm The time of the data file is specified in the attribute Date CNES JD that is the sequential Julian day starting with January 1 1950 http www aviso cls fr html donnees tools jjtocd_uk html provides a conversion utilit
49. 32 pixels A special type of unformatted raster images corresponding to the North East coastline of the US NEC in the Lambert Conic projection ASCII files Multi band images of band sequential line interleaved and pixel interleaved types CoastWatch NOAA NESDIS formats of compressed or uncompressed images that include ancillary information and may include overlays of coastlines and other features Compressed run length encoded images Erdas LAN format images NOAA AVHRR Level 1B format files with multiple bands images see http Awww saa noaa gov Images in HDF Hierarchical Data Format version 4 that include raster 8 images and Scientific Data Sets SDS with special functions for the SeaWiFS MODIS SSM I OCTS MOS and other satellite sensors HDF files transformed from the Terascan data sets HDF4 has become the de facto WIM User s Manual Introduction standard for storing satellite image data and WIM includes many specialized features for handling these images Starting in 2015 NASA Ocean Color Processing Group is gradually switching from HDF4 to netCDF version 4 as the standard file format WIM starting from version 8 has been upgraded to use those netCDF files Older versions 2 and 3 of netCDF files could be read in WIM as HDF files However it is better to read them as netCDF in WIM that is the only option with netCDF version 4 and later Sea surface temperature SST datasets in the
50. 4 741 744 Wessel P and W H F Smith 1996 A global self consistent hierarchical high resolution shoreline database J Geophys Res 101 8741 8743 WIM User s Manual 125
51. A AA 27 DING ABA AA NANA NANA ANA NAAN AA ANA 27 ROS PAPAANO AA NAGANA BAHA PAKANAN AGHAM PAA A PAPA BANONAAAA AHAHAHA AAHOHHARR 31 Lookup Tabla AA AA AA 31 Load LUT en AGA 31 Load LUN CW sc aaa ANA AA AA 32 SV LY ARAMA ANA RAP ARA ASAP NDN 32 S ve LUT KAW aaa NA see cee et ee eee cee eee 32 Save INO Se AN AA 32 Page SB aUpu AA NAIBA SAAN 34 a T T E E E ETS 34 Print PreVieW aa ieciacusie sancnntedacuase sanchnnttcacuakt sandunatedaduass saneuuaiseaduantsandeeatedauuastcaneuentaeiae 34 Print SEND ae ee ee eee eee ee eee ee een 34 AA ac ncn ia pg cc AA AA 34 6 2 Edit Edit Operations is ce a 35 CODY A AA AA E AA AA 35 BLG AA AA AA AA AA AA 35 io Rem ene aE eee ee ee ree ee eee eee 36 SCale tO Clipboard sss cciscacssesecsescecccesacnsndssceccectaxesancnsieccevsaceraacaneeacessantacananersceesexss 36 6 3 View VIEWODPLONS aaa ANAKAN ec 37 Gaba ANNA ANA AA AA 37 SEL Eg BINAN ANA ANG AA AA 37 Maga AA AA AA 37 POO es ANAKAN 37 ANO Jee ne ee GATA AA AA AA 37 UNG 39 Settings Aa AE AA NAA AA AA 40 BO Eeo ela AA AA AA AA AA 51 Vector OD GCUS Kama AKIRA ANA 52 LUT Medi i aaa RANA NANANA RIN NANANA 53 LUT SUE AA AA NAA AA AA AA AGA 54 Loop VMN ACS aaa AA AA AANI Saira aa 54 PRO GONGE SCAG AA PAA 54 6 4 Examine Examine Operations 11 nnna 55 8 vic AA AA AA 55 Contour Lines rissen AYA GANAN NANANG NAAN NAAN NAAN 55 1 Introduction MUSEUM PAA cee cd oa eed co eects 55 Line SN Since hoc lbp oc AA 56 PEEKEL aa
52. AN NANANA NN NARAYAN eel 104 Level 2 produCtS isin siina eienares PA 104 Level 3 ooo ob em iP re er ere 104 Land Products cis 22 Lan dk o ena a ccs he otek Ob aa gue nad Lababo na 104 Level 2 prodUCtS sa GAGANA AGA 104 Level 3 alen bA 104 MODIS Data at NSIDG ANA NANANG LANANANG TANGAN 104 Show Cover products sieran apd aiea aeieeiaii araida na naenin danara 104 S a IGG Products A AA AP T 104 PE c oe O E E E 105 13 kandsat Prod C Sara man nan NARR NENE NAR NANU ANN NANA 106 14 GOES SST Prod C S NAGANA aeaiia aatia aeaea 108 15 AtMmMe try P ao ce 0 73k AA PAA 109 16 SST data from the Institute Maurice Lamontagne 0eeee 110 17 AMSR E sea ice data from the University of Bremen 00 8 111 18 New Generation SST for Open OCeAN cscceseeeeeeeeeeeeeeeeeeeneeeneeeeeees 112 19 AMSR E data from Remote Sensing Systems mmmnnnmmmauuwwnnnans 115 20 GlobColour ocean Color Products cccceceeeeseeeeeeeeeeeeeeeeeeeeeeseeeeees 116 21 MERIS Level 3 product ccccccccessesseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeseeeeeeeeeeees 118 22 Cross Calibrated Multi Platform CCMP Ocean Surface Wind DKOGUCIS maa 119 23 Useful HINIS LU mama ccnsisveesevenstewaseansecnserweueeseanneneveneswesteeeenseeeeeseee 120 24 LISt OF Pil S AGANE 123 25 REIETENCOS stabs tecestiadeccsitctacsasadincwldedanassdatnnanitedaenalabanentevdaacbeasaaein titan 125 WIM User s Manual 6 Intro
53. An easy way to find the minimum and maximum values and stretch the color coding between them is to use View LUT Stretch Some image operations available for 1 byte per pixel image buffers are not implemented for 2 and 4 byte buffers The operations not supported are grayed in the menus You can convert between image buffers of different pixel size with Transf Convert The options in Transf 2 Byte To 1 are similar to the options when the 2 byte buffer is read and immediately converted to 1 byte buffer When converting the range of O to 65535 different values of an unsigned integer to the range of O to 255 possible for an unsigned byte a transformation of values is needed The representation of an unsigned integer depends whether the most significant byte MSB is the first or second The pixel value is formed as 256 MSB LSB least significant byte Besides the byte order you can choose many ways how to make the reduction from 16 bits to 8 bits The easiest would be to shift the bits to the right Shifting to the right by 1 is equal to division by 2 shifting by 2 corresponds to division by 4 etc In essence if you select shift 1 you will loose the least significant bit but retain the original values between 2 and 511 transformed to the range of 0 255 If you make a shift of 2 you will retain the original values between 4 and 1023 In WIM the bit shift operation is coupled with magnitude test If after the operation the value is still
54. DIS Products Atmosphere Products Land Products Level 2 products Aerosol products Water Vapor products Cloud products Atmosphere Profiles Cloud Mask products Joint products recent data can be downloaded from the Data Pool at http daac gsfc nasa gov data datapool MODIS 02 Atmosphere 01 Level 2 index html Level 3 products Only a small selection of products is listed here For example the Atmospheric Monthly Global Joint Product contains 739 statistical datasets that are derived from the Level 3 Daily Global Joint Product MOD08 M3 Recent data can be downloaded from the Data Pool at http daac gsfc nasa gov data datapool MODIS 02 Atmosphere 02 Level 3 index html Level 2 products Level 3 products Many products e g Vegetation Indices 16 Day L3 Global 1 km data MOD13A2 are available from http edcdaac usgs gov dataproducts asp MODIS Data at NSIDC The National Snow and Ice Data Center holds many MODIS products on snow and ice at http nsidc org data modis data html Snow Cover products For example MOD10A2 has 8 day Terra snow cover data at 500 m resolution in the Sinusoidal projection Sea Ice products For example MOD29P1D has daily Terra sea ice extent data at global 1 m resolution WIM User s Manual 104 GLI Products 12 A2GI A2GI A2GI A2GI A2GI GLI Products GLI Global Imager was an advanced sensor aboard the MIDORI II ADEOS 2 spacecraft of the Japanese Sp
55. Lon Value ASCII Loads individual triplets of Latitude Longitude and Value from a text file creates a Float32 image in Global Equal Angle projection The size of the image is taken from Settings Misc Default Navoceano Size Each triplet in the input file is assumed to be on a separate line The sequence can be either Longitude Latitude Value or Latitude Longitude Value depending on the Lat first checkbox in View Settings Misc Each line may have more data after the Value only the first columns are used and the rest are ignored For example the input file may have lines like 180 2500 78 2500 0 0094 180 2500 77 7500 0 0087 180 2500 77 2500 0 0108 180 2500 76 7500 0 0155 WIM User s Manual 22 6 1 File Operations 180 2500 76 2500 0 0180 However if the Latitude Longitude sequence is chosen incorrectly the resulting image may look like this The Latitude is assumed to range from 90 to 90 and Longitude from 180 to 180 If Longitude is over 180 then it will be normalized to the range 180 to 180 It is natural that depending on the specified size and the number of specified pixels the resulting image may be sparse i e have empty pixels between pixels with value NAVOCEANO L2 def Sea Surface Temperature SST datasets derived from the NOAA National Oceanic and Atmospheric Administration Polar Orbiting Advanced Very High Resolution Radiometer AVHRR using the multichannel sea s
56. M and WAM and follow all default Typical options If you are installing from DVD then the setup process should start automatically If you have older versions of WIM please uninstall the older version before installing the newer version The setup program will make a new directory for Wimsoft the default is C Program Files Wimsoft expand and copy all or the required files Note that in order to install WIM you need to have Administrator permissions The setup process creates some separate folders under the Wimsoft folder such as Maps Examples LUT that include e sample images a sample lookup table etc they help you get familiar with WIM functions and can be deleted to save disk space e Terascan format map projection files for North America e map files for generating overlays of coastlines country boundaries and rivers for any image with a projection information e Note that the data files for generating high resolution maps of global bathymetry coastlines political boundaries and rivers are not part of the downloadable WIM evaluation package e Please note that in addition to the Example files the WIM DVD has folders Images that includes sample real life satellite images and a Sat folder with large sets of satellite data The automated installation procedure will not install the mages or Sat folders You can copy all or part of them to your hard disk for faster access When you run WIM for the first tim
57. Mask w Image has a flexible masking that corresponds to the MODIS Quality image The default values O for mask value and Yes for smaller and equal of the Multi Mask w Image function are suitable for picking the good pixels of MODIS images MODIS data are delivered with a corresponding Quality flag image that has pixel values corresponding to the quality flag as follows 0 good 1 questionable 2 cloud 3 bad This meaning of the flags here is given as an example only and may depend on a particular convention and data type The default selection would selecting pixels with Mask value lt 0 i e good When selecting pixels with mask value gt 3 for example only the pixels of quality bad are returned For screening multiple images you can use the WAM program wam screen that processes all selected file in a folder picks only pixel values with 0 quality and saves the screened images in new files The program can be modified to perform other functions e g pick different quality pixels Radiances and Geo location Creating true color images Both MODIS Terra and MODIS Aqua instruments can produce great looking true color images WIM has several tools and functions for that As this involves operations on multiple bands it is better handled with the WIM Automation Module WAM Please see the WAM manual WAM pdf and the WIM and WAM exercises for examples WIM User s Manual 103 11 MO
58. NG Renee ree rao prime TOTS aa a eT OTe prim AA 93 Dale sc eeeeccc claret ds PA AA AA AA AA AP PAA 93 84 f 1 AA AA AA AA AHA AAO AA 93 840 111 101 SAMBA IPA PABAMPPEMABA NP PANAMAN A E PPABAPPPANAPAHIPE GP E ET 94 DISanG NANA hein E ti hl a ht hai ee ae 94 Kernel gS eee ee ere eer eee ANA ee ee ee ee ee a ee 94 Kernel Combine AAKALA 94 CLET ah ua E A E E A E TTT 95 PRON TRAC AA AA NAA AA 95 PUM 9 0 1 APAPAP PAPA PS 95 lio o NAA AA 95 6 10 Window Window Arrangements 1 00000000sassassamaaa a 96 Cascade iiiad PAA AKE NAGA eer ne eer PERE repr Se rs gt gt on gt peep ar pee Pap PAYER KANA PAN ee NAKE PANAN ANG 96 MN cect does AA AA AA AA AA AA AA 96 Arrange leone a ek has chee ins cs Secs oe ees ccc fesse Feeds 96 6 11 Fret WIM Helpi siiis ka 96 lalo 5 PAA AA AA AA 96 Using Halip AA AA AA 96 POUR Md 2 es core careers cused a cet se haaa NB T 96 ENG TNS aa 96 T SENUUPEOQUCIS a NAA tarred oath AGA NG 97 8 SeawiFS Products ancients iene deen 98 9 SOCTS Pro Ques 0GIG 100 10 MOS PloQue CA AA AA 101 WIM User s Manual 5 1 Introduction TT MODISPrOdUCIS nma nanana naRANRANARENANANANNNNN 102 OVEIVIEW pa R es nee ARR ee ie es 102 Ocean Products AGANG AGARANG 102 Radiances and Geo location 1 eeeena aaa a aa AA AA AA ANAN AA ANAN AA AA AASA AASA AA 103 Creating true color iMaGes cccccccssceeseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeneeeeeeeeeseeeeeeeeees 103 Atmosphere Prod NGI
59. Polyconic Lambert Azimuthal Lambert Conic Albers Conic The GCTP family of mapping functions includes most common projections includes as a separate dynamic link library gctp dll and derives from the WIM User s Manual 45 6 3 View Options General Cartographic Transformation Package GCTP from the USGS National Mapping Division Current Settings x General Extensions Projection HDF Options Misc WIM native projections C Unmapped Lat Lon arrays IngestL1B C Linear Equidist Glob Mercator Mercator C n3a C Glob eq angle C Pol Stereo C n3b Shift po C TY Mere C s3a C GlobColour ISIN C Conic C s3b Global ISIN ModisLand Swath Albers C HDF EOS C GCTP Grid Terascan Projections C Sensor Polar Stereo Polyconic C Stereo C Equidist Azim Lambert Azim Rectangular 1 Mercator Lambert Conic Orthographic C UTM C Albers Conic Longitude coefficients Latitude coefficients 0 000000 f0 000000 fo 000000 0 000000 OK Cancel Apply Help The Global equal angle projection is popular for Level 3 global images and is also known as the Plate Carr e grid It is a grid and can be used in WIM without any additional parameters The only issue is the meridian value corresponding to the left side of the image The default value is assumed to be 180 e g SeaWiFS Level 3 and Pathfinder Sea Surface Temperature images and the Shift value is equal to
60. Prompts for the size of the window where the test is made All isolated blobs of pixels that do not cross the window border will be converted to zero Dilate Dilates expands non zero pixels e g edges in all directions trying to connect broken or disconnected parts of an assumedly continuous contour edge Thin Lines Uses the sceletonization algorithm of Pavlidis 1980 to thin the dilated edges by peeling off the unnecessary layers of edge pixels SIED Uses the Single Image Edge Detection SIED method of Cayula and Cornillon 1992 with the variable window VW modification of Diehl et al 2002 It works with multi byte images but the unscaled pixel values must be between 0 and 255 It is recommended to use only with Byte images If the source data is not Byte then you should convert the to Byte using Transf Convert and a suitable scaling method e g SST Pathfinder for SST data The user can select a fixed window size for edge detection or accept 0 for the variable window size of Diehl et al 2002 Bigger window size detects larger scale fronts A typical fixed window size is 32 Cayula and Cornillon 1992 If the fronts look too thick then you can use the Edge Thin Lines after Edge SIED 92 6 9 Segmentation 6 9 Segm Segmentation This section contains routines for edge detection as well as segmentation classification of an image into different sub areas segments The segmentation algorithms have been dev
61. RSS under the DISCOVER project http www remss com The Level 3 5 data sets data are stored on a uniform grid with a resolution of 0 25 X0 25 near global 78 375 S to 78 375 N domain averaged over 5 day and monthly periods The data are in netCDF format and can be downloaded from ftp podaac jpl nasa gov ocean_wind ccmp L3 5a data After downloading you need to unzip the data gzip d gz WAM has a special converted program wam_convertr_ccmp that converts the netCDF files into HDF files that are much more convenient to work with WIM and WAM wam_convert_ccmp nc Each file has 5 products uwnd vwnd wspd upstr vpstr nobs see the user s guide at ftp podaac jpl nasa gov ocean_wind ccmp L3 5a doc ccmp users guide pdf wspdis the scalar wind speed magnitude uwnd is the eastward component vwnd is the northward component upstr is the eastward wind stress vpstr is the northward wind stress and nobs is the number of observations If you don t have WAM and wam convert ccmp then you need to do manually the following set of operations all done automatically with wam convert ccmp When loading the nc file cancel the selection of Geolocation file View LUT Stretch Transf Mirror Horizontal axis Settings Projection set Linear projection with Longitude coefficients 0 0 25 and Latitude coefficients 78 375 0 25 Check with Geo Get Map Overlay that you have the correct projection With WAM you can
62. Save Points option select a file point with the mouse exactly to each of the points and press the right button Find the longitude and latitude of the points from the map in decimal numbers enter them as new columns of data into a worksheet where you have loaded the ASCII file with the video coordinates Find the intercept and the slope of the linear regressions from the x coordinate to longitude and y coordinate to latitude What you should get is something like that 14 83356 0 016362 58 29435 0 00891 You can now use an editor to create an info file that contains first the image dimensions DX and DY and then the coefficient values each value separated with a space in one line The name of the file should correspond to the image file name with an extension inf and should be in the same directory as the image file Instead of using an editor you can also enter the coefficients in View Settings dialog boxes and use File Save Info Next time when you read the image WIM will automatically read the coefficients from the info file and use them You may want to check out the info files corresponding to some sample images e g calchl81 inf Problem The image data is in the GIF TIFF or some other format not directly readable by WIM Hint Use one of the many format converter programs Many are available in the public domain or commercially One of the most versatile image converters is Image Alchemy Handmade Software Inc th
63. TIFF tags will be added to the saved TIFF file according to the GeoTIFF format that is a common format in the Geographic Information Systems GIS Adding GeoTIFF tags preserves the geo registration of the image when read with GeoTIFF compatible software GeoTIFF tags for other projections may be added in the next versions of WIM Close Closes the current image buffer i e frees up the allocated memory and discards the corresponding bitmap from the screen The same can be completed by closing the image window or by selecting the image from the List of Images box and choosing Delete Lookup Table Allows to read load and write save lookup table LUT files LUT is used to associate certain colors with certain pixel values Load LUT Loads a custom look up table LUT for color coding from a file and sets the palette flag to Custom see View Settings General Palette The default custom LUT until another custom LUT is loaded is the grayscale The LUT file should be an ASCII file with the following structure 0 r0 go bO 1 ri g1 bl 2 r2 g2 b2 255 r255 g255 b255 Each line should end with a Carriage Return CR and Line Feed LF character which is standard in DOS Windows text files In UNIX ASCII text files end with a single LF character Here the first column is just the sequence number for your convenience and is not used by WIM The next columns specify respectively the amounts of red green and blue numbers
64. V Write Into Image Buffer Save C Deg Sec Decimal Value fo Close The tracks for which the distances are calculated can be recorded in the screen bitmap or they may be recorded with a selected pixel value in the image buffer When only the screen bitmap is changed the tracks disappear WIM User s Manual 70 6 5 Geo Operations when the bitmap is rebuilt When values in the image buffer are changed the image with the changed pixel values can be save as a new image with the assigned pixel values along the tracks Read Vector Objects from HDF Reads a HDE file with WIM vector objects such as points transects rectangles or vectors see Vector Objects Vector objects can be read from ASCII text files with Geo Get Vector Objects The vector objects of an image are show with View Vector Objects Remap Projection Remaps the current image from the current geographical projection to another projection and or to another size It may be difficult to generate a new projection from scratch it is much easier to pick a another projection from another image and remap the current image to that projection Projections other than Linear make use of additional reference variables see View Settings Projection that are read as HDF or CoastWatch attributes CCAR image header or from the separate info file The meaning of the reference variables depends upon the projection chosen In most cases ref1 and ref2 define the
65. a the rectangle borders and the numbers indicating its size are hardly visible on a fine grained image with highly variable colors Hint By using View Set Colors make the image look darker or lighter so that WIM can pick a color for drawing the borders and numbers that will be clearly visible Problem How to find the geo conversion coefficients to convert the screen video coordinates to the geographical longitude and latitude in Linear projection WIM User s Manual 120 22 Useful Hints Hint The conversion coefficients are two pairs A B C D of linear regression coefficients that convert the video coordinates 0 O in the upper left corner to the longitude and latitude Lon A B X Lat C D Y If you do have the coordinates latitude and longitude of the corner points then A equals the longitude of the upper left corner and C equals the latitude of the upper left corner To find B D subtract the longitude latitude of the lower right corner from the longitude latitude of the upper left corner and divide the result by the corresponding image dimension minus 1 If you don t know the coordinates of the corner points you have to find them the hard way To find the regression coefficients select around 10 points ground control points that are clearly distinguishable on the image and on your reference map e g islands headlands lakes and record their video coordinates to a file using Examine
66. a Type lt GLOBAL gt charg HRPT LAC Pixel Start Number lt GLOBAL gt int32 1 LAC Pixel Subsampling lt GLOBAL gt int32 1 Node Crossing Time lt GLOBAL gt charg 2001105041717758 long_name chlor a chars Chlorophyll Concentration OC4 Alge slope chlor_a float32 1 000000 intercept chlor_a float32 0 000000 units chlor_a chars mg m 3 Lit ai gt Add Edit Delete Copy To Close settings Allows to view and alter the current settings As this function is used very often it has a shortcut to it on the toolbar It has the following sections tabs General Extensions Projection HDF Options Misc General Image Size Width DX Height DY and type i e the number of bytes per pixel and either float or integer type 2 byte and 4 byte per pixel images can be either Signed or Unsigned Color Scaling Min value and Max value Up to 256 different colors are used by WIM to display images Color Scaling sets the Min and Max values of a range that is linearly stretched into 256 color bins Unsigned 1 byte per pixel images can have only 256 different values therefore this function is not needed for them It is needed for buffers of 2 or 4 byte integer and float numbers After you set the Min and Max values the subsequent color stretching View Set Colors or the Toolbar color icon only applies to the range between those Min and Max values These Min and Max values are saved as WIM Color Range attributes with Save
67. ace Agency http sharaku eorc nasda go jp GLI index html MIDORI II was launched on December 14 2002 and the initial data products looked very promising Unfortunately the spacecraft was lost on October 25 2003 Special functions for all levels from 1 to 4 are being developed for WIM The GLI Level 1b and Level 2 products are somewhat similar to MODIS respective products For example the Latitude Longitude arrays LLA are in a separate file Starting with version 6 14 WIM is capable of displaying and navigating GLI Level 1 and Level 2 products Selecting the matching file with the proper LLA can be quite a challenge as the filenames are very long and hard to memorize For example having the following files in the folder and trying to load a dataset from the first file we need to select the second file in this list for geo location LLA and not any other file in the list L103040847120D2 OCSFRO20000000040A000 00 lt getting data from L103040847120D2 PS1BC431030 00 lt corresponding LLA 1103041647110D1 PS1BC431030 00 L103041647120D2 OCSFR020000000040A000 00 L103041647120D2 ONLFRO200000000100000 00 This seems easy when we have just a few files in the folder but becomes overwhelming with many similar files with long names A smart algorithm is being used to help the user to locate the right file with the LLA The algorithm uses the longest common substring to suggest the best matching filename with the LLA The Level 3 binned a
68. alculating average and standard deviation Get the edge image that was used to generate the edge distance image you may dilate it to exclude even bigger surroundings of edges binarize it Transf Binarize with a negative value e g 128 turn back to the original source image and mask it Multi Mask w Image with the newly created negative edge image Now use this image as the source image when using Set Segment and you get the segments set to the values of averages and standard deviations over their interiors WIM User s Manual 95 6 11 Help 6 10 Window Window Arrangements Cascade Tile Arrange Icons 6 11 Help WIM Help Index Get an index of WIM help topics Using Help Instructions for using Windows Help Ka About WIM Displays info about WIM including the version number and the licensed user name License Allows to insert your final license number WIM User s Manual 96 MODIS Products 7 SSM I Products The National Snow and Ice Data Center NSIDC CIRES University of Colorado at Boulder Campus Box 449 Boulder CO 80309 0449 USA http www nsidc colorado edu has made available a huge data set of among others daily North and South polar brightness temperatures sea ice concentration grids and monthly averaged sea ice concentrations obtained from the SSM I sensor The SSM I is a seven channel four frequency linearly polarized passive microwave radiometric system and is part o
69. alette Other palette files included with WIM are chl lut anomaly lut and spectrum lut can be used instead You can create your palette with Edit LUT Edit save it with File Save LUT and then make it your Default palette WIM User s Manual 50 6 3 View Options Override LUT in HDF this allows to ignore the palette in a HDF file and use the current default palette Starting with WIM 6 28 an individual palette is saved with each image as an HDF attribute Sometimes you may want o use a consistent palette for all images This option allows to choose whether to use the palette from the HDF file or the default palette Coast Overlay for Default Images this coastlines file is used in the File New function to create and overlay coastlines The default is C Program Files Wimsoft maps coast_inter b Lat Lon Value Format Lat first this checkbox allows to choose if Latitude or Longitude comes first in triplets of Latitude Longitude Value in the format of pnt or csv files in Geo Get Vector Objects and in saving files as Lat Lon Value ASCII dat Format string this edit box allows to select the exact format of saving triplets of Latitude Longitude Value The default formatting string is 6 4f 6 4f 6 4fln and means that 6 characters including 4 characters after the decimal point are used for each of the three values The n character means end of line and a tab character could be inserted as 1t
70. as HDF gt Value Scaling Pixel Value LOG MINUS4 x10 C SST PATHF C x 100 C SST SMHI C CZCS Pigm Log Chl mg m3 Bitmask Linear Slope fo 100000 Logarithmic Intercept Fi 2 0000 e om Wa BN WIM User s Manual 40 6 3 View Options Value Scaling Real life image values geophysical values are seldom confined to integers provided by the 1 2 or 4 byte per pixel integers therefore value scaling is used to convert geophysical values into pixel values WIM has both predefined and generic scaling options Generic scaling can be either Linear or Logarithmic The Base in the Logarithmic scaling is always assumed to be 10 0 Both Linear and Logarithmic make use of the Slope and Intercept values These scaling type Linear or Logarithmic Slope and Intercept values are saved as attributes with Save as HDF A simple predefined scaling option is x 10 which simply means that the pixel values have been derived by multiplying the geophysical values with 10 and rounding to the nearest integer E g a temperature range 0 0 to 25 5 C is naturally converted to unsigned byte values by multiplying by 10 x 70 value scaling In that case a pixel value of 250 would correspond to temperature of 25 0 However in order to compress more information to an 8 bit pixel value and still use the simple format of 1 byte images more sophisticated coding can be used WIM uses 2 predefined nonlinear Logarithmic value scali
71. at is available both for MS DOS and UNIX For example to convert a GIF or TIFF 24 bit file use Alchemy with the BIF option that produces band interleaved by pixel BIP files then use File Open Pixel Interleaved Problem How to save print the image together with the corresponding color bar scale Hint Use View Annotate since WIM 5 11 In older versions of WIM you can use a more cumbersome procedure You can use Examine Scale to Clipboard and then use another program e g MS Paint to place the color scale into a copy of the image at the bitmap level To WIM User s Manual 121 22 Useful Hints do that load the program Paint included with every copy of Windows 95 Copy the image bitmap from WIM to Paint via the Clipboard Edit Copy then within Paint Edit Paste Then copy the color bar via Clipboard to Paint Examine Scale to Clipboard within Paint Edit Paste move the color bar to a suitable location save the bitmap or print within Paint If your pixel values are coded e g real values multiplied by 10 don t forget to select the Value Scaling option in View Settings General in order to get the correct scale The transfer of color images works fine in 16 24 bit color and when the bitmap is saved as a 24 bit bitmap With 8 bit color the colors may be distorted when Windows runs out of available colors WIM User s Manual 122 23 List of Files 24 List of Files
72. ave an image buffer in a file under a different name and then want to save the corresponding info file File Save Info WIM User s Manual 15 6 1 File Operations 6 Menu System The main menu system provides the following choices File Edit View Examine Geo Transf Multi Segm Window Help Note Some of the menu options are not available for WimLE and WimCW versions The most frequently used functions have shortcuts on the Toolbar De AO Bs SQ Be me Bw e 6 1 File File Operations E New Opens a new empty all pixel values set to zero image buffer named lt New Image gt with one byte per pixel and either with selected width DX and height DY in pixels or the approximate pixel size in meters The WIM Linear projection can be set with a specified Latitude and Longitude range Coastlines can be automatically created and overlaid with a coastlines file specified in Settings Misc New Image Xx Projection Unmapped Linear Image Size in pixels 800 x s0 Pixel Size in meters 5000 Latitude Range 64 1 57 Longitude Range 63 5 40 IV Apply Coastlines Overlay OK Cancel Filter to Disk see Transf Filter to Disk Open Opens a file on the disk reads one or more images to the corresponding image buffers and displays them on the screen More complex image formats CoastWatch HDF usually include all the necessary information to read and decode
73. buffer and writes it to a binary file The pixels different from zero will get a bit value of 1 while zero pixels remain zero NB In order to do the 8 byte to 1 byte compression the number of pixels in a row has to be a multiple of 8 It is recommended to make overlays only from images with 8N pixels per line It has to be noted that while 1 bit per pixel overlays provide 8 times compression for binarized images RLE encoding see File Save As Compressed RLE gives even better compression rates while keeping all the 8 bits per pixel Monochrome Bitmap 16 Color Bitmap 256 Color Bitmap 24 Bit Bitmap Saves the bitmap from the current image to a bitmap file with a specified number of colors type of bitmap 2 16 256 or 16 7 million 24 bits It is usually sufficient to have 256 colors in the bitmaps as 24 bit bitmaps take about 3 times the space of 256 color bitmaps ASCII Writes pixel values of the currently selected area rectangle or line or the whole image as ASCII numbers in a file The format of the recorded ASCII WIM User s Manual 29 6 1 File Operations numbers depends on the Value Scaling option see View Settings General Value Scaling i e it is a 3 character integer separated with a space for no value scaling and a floating point number for the scaled pixel values This function is useful for transporting data to other software packages e g for additional statistics plotting etc As images u
74. c Data Sets to a file in the HDF format see File Open HDF file Note the Save as compressed option in Settings HDF Options With compression the file size can be reduced many times while the file extension and usage remain the same Auxiliary data in the form of global or local attributes see View Attributes are also saved into the new HDF file This allows to save e g the scaling settings and projection information when saving into a new file You have to select one image buffers to be saved into the HDF file Please note that the image data is internally compressed if the checkbox Save as compressed is checked in Settings HDF Options Compression is handled internally by the software when reading the HDF file For typical images with large areas of empty or uniform data the compression ratio is often 10 15 times and savings in disk space are drastic For images with variable pixel values the compression is less than two times When creating new images the attributes are generally copied from the source image to the new image However sometimes some attributes need to be changes e g those related to Value Scaling if the scaling was changed with the Transf Convert operation Starting with WIM 5 37 this WIM User s Manual 28 6 1 File Operations has been fixed so that the appropriate new scaling attributes Scaling type Linear or Logarithmic Slope Intercept are saved with the new image The Col
75. ces by the Kawamura Lab in the Tohoku University Japan The main advantage compared to other traditional satellite products is that by merging data from different sources and modeling achieves high spatial resolution at high temporal frequency The data format is a simple raster binary Image in WIM terminology with a header of 200 bytes However this format is not a modern format like HDF and therefore the reading program has to be specifically modified to read it In order to use NGSST files with WIM do the following e Download the data from ftp www ocean caos tohoku ac jp pub mergedsst binary e Uncompress the compressed files e g with a command line program gzip d gz e Set the Image Header to 200 bytes in Settings Misc see below Current Settings x General Extensions Projection HDF Options Misc Vector format r Navoceano Size r Image Header x1 yl x2 y2 Width 720 fo bytes xy uv co Height 360 Default LUT MV Use Default LUT Override LUT in HDF C Program Files WimS of SLUT chil_white_ m Lat Lon Value Format Lat first Format string 26 41 6 4f 6 4f n Load an image with File Open with Files of type Image img The default extension for the raster Image files is img but these files have an extension raw therefore you have to either specify raw in the File name text box or add raw to Settings Extensions for Image file Separate the ra
76. ctly read with WIM but it is much more convenient to use a WAM program wam_convert_amsre to convert the binary files into HDF files and at the same time adding some important attributes The HDF files can then be easily analysed with as times series using WIM and WAM programs If you cannot use wam_convert_amsre to convert the binary files into HDF files you have to go through multiple steps in order to read a single file When reading a binary file AMSR E file directly with WIM you need to do the following e Open the files with File Open as Files of Type Band Sequential As these AMSR E data files do not have an extension you need to specify in the File name text box in order to see the available files e Then specify 5 for 3 day weekly and monthly files or 6 for daily files as the Total number of bands You can leave the Bands to read box open and read all the specified Total number of bands e Next specify 1440 as the Width and 720 as the Height of the images e Now the images should be in WIM but you can notice that they are upside down To fix that do Transf Mirror Horiz axis e To fix geo location set Settings Projection to Glob eq angle with a Shift of 180 You can confirm the correct geo location with Geo Get Map Overlay coast inter b Background Value 0 Foreground Value 1 Note to use 1 as the foreground value as choosing 255 would create white coastlines that may not be visible Click back on one of the i
77. d and Step contour values refer to image values after Value Scaling If you want to select the contour lines according to pixel values i e without Value Scaling set the Value Scaling equal to Pixel Value in View Settings You can also select the pixel value with which the contour lines are drawn The pixel value of 255 results in white contour lines for most color palettes LUTs with Byte images but not with 2 and 4 byte per pixel images The background pixel value of the new image if created will always be 0 lal Histogram Consists of several histogram operations WIM User s Manual 55 6 4 Examine Operations Histogram Calculate Calculates histogram for the selected rectangle or line or for the whole image and shows it In case of 2 or 4 bytes per pixel integer or float the pixel values are first scaled with the Min and Max values from View Settings General Color Scaling to be between O and 255 The following histogram calculation is then similar to the 1 byte per pixel images Show Shows the previously calculated histogram Optionally switches to Histogram Save or Histogram Values The vertical scale of the plot can be changed with the scrollbar Values of the histogram maximum the pixel value with highest frequency maximum histogram and the vertical scale are shown Save Saves the current histogram in an ASCII file Thresh Shows the pixel values of the distribution tails below wh
78. direction is coded as following 1 right 2 up right 3 up 4 up left 5 left 6 down left 7 down 8 down right the direction points to the higher side Kernel Combine Is used to combine closely located kernels maxima into the same class segment This routine uses the kernel images from the previous operation and the distance image to the nearest edge Edge Distance It tries to connect only kernels for which to sum of separation in rows and columns is less than 100 The criterion for linking two maxima m1 and m2 is dist m1 m2 lt dt m1 dt m2 a where dt is the distance transform from the nearest edge Edge Distance of the pixel and ais a coefficient initially 0 6 The final number of different kernels must be less than 255 value O is WIM User s Manual 94 6 9 Segmentation assumed to represent a non kernel pixel If the number of different kernels after combining is still gt 255 the routine increases a in order to connect more kernels Kernel Grow Uses the combined kernel image must be current and the gradient direction image of the edge distance Kernel Find to grow the kernels It should be used once Pixel Trace Should be used after Kernel Grow It uses the same source and gradient image as Kernel Grow and tries to convert non kernel pixels to kernel pixels by following each pixel along it s gradient direction If non classified zero or black parts are left after it
79. do time series analysis of the 21 year wind records in your area of interest Note that if you want to extract the wspd image with wam_series you need to select SDS number 2 the sequence is 0 1 2 3 4 Also note that wam_statist is always using the first Oth SDS in the file Therefore if you want to use wspd in wam_statist you need to extract it first with wam_series into a series of HDF files with a single SDS References Atlas R Ardizzone J Hoffman R N 2008 Application of satellite surface wind data to ocean wind analysis Proc SPIE Vol 7087 70870B 2008 DOI 10 1117 12 795371 Atlas R Hoffman R N Ardizzone J Leidner S M Jusem J C 2009 Development of a new cross calibrated multi platform CCMP ocean surface wind product AMS 13th Conference on Integrated Observing and Assimilation Systems for Atmosphere Oceans and Land Surface IOAS AOLS WIM User s Manual 119 22 Useful Hints 23 Useful Hints Problem After reading a new image from a file the image appears distorted on the screen Hint In order to display an image WIM has to know its dimensions most importantly the number of pixels in a line The number of lines is not so important as the missing data is assumed to be zero and excessive data is simply ignored WIM gets the image size automatically from its attributes e g from HDF or other complex file formats but plain raster formats do not have any attributes You can check the
80. duction 1 Introduction WIM is a general purpose image visualization and analysis program for the Microsoft Windows operating systems with special features for analyzing satellite images WIM has been designed to work with digital images In the WIM parlor an image is simply a two dimensional array matrix of numbers representing the values of its elements pixels Digital images of this kind are produced by satellite sensors medical imaging devices computer models etc There are many ways and formats to store digital images The simplest is a sequence of numbers e g row by row representing the pixels which is often referred to as unformatted raster image In this case the total number of pixels equals the number of rows times the number of columns an the file size equals the number of pixels times the number of bytes per pixel In the simplest case each pixel is represented by one byte Complex file formats e g HDF add a lot of additional information to the file e g the dimensions color palettes geometric projections and other attributes The image data may be internally compressed to reduce the file size The most important file type for satellite data is HDF see below However WIM supports many file formats Some of these formats are listed below Unformatted raster images of unsigned byte Byte one bit unsigned and signed integer 2 bytes Int16 long integer 4 bytes Int32 or floating point 4 bytes Float
81. e as in the Transect track trk option For each specified rectangle the Examine Statistics function is performed If no image is in memory a clean zero values image buffer with the current size is created and the points along the perimeters of the rectangles are assigned the value of 255 You can then store it in various image formats e g File Save As Image or overlay file with File Save As Overlay or use it as an overlay for another image In order to overlay the rectangles on another image first switch to the image and then use Multi Overlay Image Drifter track This is a special version of the Point object and is like a set of points treated together An arrow showing the direction of the drifter track is shown The file format is the same as of the Point vector objects Additionally the NOAA drifter data format DAT can be read directly These files are assumed to have either DAT or txt extension and always have the latitude value preceding the longitude value regardless of the setting of the View Settings Misc Lat first checkbox A section of the NOAA drifter track format is given below 4L 4 K 1 2003 07 26 00 25 15 31 000 240 000 0 000 401647450 42 4K 42 4K WIM User s Manual 68 6 5 Geo Operations This short section has only two points 31 N 240E and 26N 239E The rest of the lines have no data Here 31 and 26 are the latitudes and 240 and 239 are respectively
82. e bitmask flags can be viewed on the source image by right clicking on the image the flags bits that are ON are shown in the window header The names of the individual bit flags are taken from the attributes For example in case of SeaWiFS Level 2 flags attribute f01 flag is equal to ATMFAIL atmospheric correction failure attribute f02 name is LAND attribute 103 name is BADANC bad ancillary data etc The example below shows that a particular selected pixel has set flags COASTZ coastal zone TURBIDW turbid water ABSAER absorbing aerosols ATMWARN atmospheric correction warning and OCEAN In fact any BYTE Int16 or Int32 image can be interpreted as a bitmask If the name of a particular bit flag is not found it is designated as unnamed A special operation Transf Bitmask exists for Bitmask images that will select areas where specific bits are either ON or OFF E 242 100 114 584 31 408 Ox80420840 COASTZ TURBIDW ABSAER ATMWARN OCEAN Area of Interest Line or Rectangle The Area of Interest feature toggles the type of selectable area between Rectangle and Line If Line is the current type of area of interest you can drag a straight line across the image in any direction Histogram Profiles and Statistics are then performed on the selected line If Rectangle is selected the corresponding operations are performed on the selected rectangle Palette Rainbow Grayscale Custom palette type Palette is es
83. e initial extensions are img dat for most image file types ov for overlays rle for compressed files hdf for HDF files etc If modified the new default extensions will be saved in registry for later WIM sessions Typical extension settings are shown below WIM User s Manual 44 6 3 View Options Current Settings E x General Extensions Projection HDF Options Misc Image file img dat nec ASCII img dat ASCII Float roas Band Sequential img dat Line Interleaved e dat Pixel Interleaved mg dat Coast Watch owt Compressed rle Erdas Lan i lan Float img dat Float Bigendian img dat GOES SST sat HDF file hdt 4km 9km NetCDF nc LIB file HE Unsigned 2 byte int img dat 2 byte Int gt Byte img dat Overlay ovl Subset F img dat OK Cancel Apply Help Projection Satellite images of the Earth can be in various geometric projections WIM has options for a number of projections Projection type is one of WIM native projections Unmapped Linear Mercator Polar Stereo Transverse Mercator Conic Albers Equidistant Cylindrical n3a n3b s3a s3b Swath HDF EOS Ingest L1B Glob Mercator Global equal angle with a Longitude shift Lat Lon arrays or one of Terascan projections Stereo Rectangular Orthographic Polar Stereo Equidistal Azimuthal Mercator Universal Transverse Mercator UTM
84. e you are requested to type in your license number that you received with WIM If you change the hardware configuration of your computer or re install Windows you need to enter the license number again If you installed WIM with the evaluation license then you have to receive and enter the final license number in order to continue using WIM beyond the evaluation period WIM User s Manual 10 Getting Started 4 Getting Started Here we learn how to use the most basic operations in WIM More detailed instructions are give in the Practical exercises with WIM and WAM and other exercises under the Manual section at http www wimsoft com manual htm or in the Course folder in the Wimsoft folder e You can load an image with the standard File Open dialog an icon on the toolbar but a much faster way is to find the necessary image files with the Windows Explorer and load them by just clicking or double clicking depending on your Windows setup on the filename To be able to do that you need to associate WIM with the specific file extension Click or double click on any HDF image file e g composite hdf in Examples or ImagesihdflSeaWiFS baja 2000 april and select Wim exe typically in C Program Files WimSoft as the program to open it with Then select read as HDF for that file type in WIM You have to repeat that for HDF files with a different extension e g SeaWiFS global Level 3 files like S20000612000091 3m MO
85. ected albedo values are multiplied by 10 0 Square Opens a new float 4 bytes per pixel image and calculates the squares of pixel values of the current image Square root Opens a new float 4 bytes per pixel image and fills it with the square roots of the pixel values of the current image WIM User s Manual 79 6 6 Image Transformations Texture Calculates a new image with a texture indicator of the current image The standard deviation or variance in a window can be selected The image is scanned with a small window and the center point of the window is replaced either by the standard deviation or variance in the window The borders of the image will have zero values You have to select the window size side length in pixels Only uneven numbers between 3 and 31 are allowed In case of a small window the features have better resolution but higher variance You have keep in mind that only integer values between 0 and 255 can be stored in WIM images 8 bits per pixel As the interesting range of the standard deviation and variance can be too low for that you can choose an integer constant with which the real value is multiplied before converting to a byte value Different regions of the image may require a different multiplication constant The new image will be named lt TexSD_WxC gt or lt TexVar_WxC gt where W is the window size and C is the multiplication coefficient Zoom Opens a new image buffer and zooms using e
86. ed images Holben B N 1986 Int J Remote Sensing 7 1417 1434 Holyer R J and S H Peckinpaugh 1989 IEEE Trans Geosci Remote Sens GE27 1 46 56 Lee J S 1986 Optical Engineering 25 5 636 643 Kahru M B G Mitchell 2001 J Geophys Res 106 C2 2517 2529 Madsen S N 1986 Ph D Thesis Danish Electromagnetics Institute LD 62 McClain E P W G Pichel and C C Walton 1985 J Geophys Res C6 11587 11601 National Snow and Ice Data Center 1996 DMSP SSMI I brightness temperatures and sea ice concentration grids for the polar regions User s Guide 110 p O Reilly et al 1998 J Geophys Res 103 C11 24 937 24 953 Pavlidis T 1980 Computer graphics and image processing 13 142 157 Peckinpaugh S H 1991 CVGIP Graphical models and image processing 53 574 580 Prangsma G J and J N Roozekrans 1989 Int J Remote Sensing 10 4 5 811 818 Prentice G S 1987 Int J Imag Rem Sens IGS 1 1 53 55 Skriver H 1989 Ph D Thesis Danish Electromagnetics Institute LD 74 Snyder J P 1982 Map projections used by the U S Geological Survey USGS Bulletin 1532 313 p Stumpf R P 1992 First Thematic Conference on Remote Sensing for Marine and Coastal Environments New Orleans Louisiana June 1992 ERIM Ann Arbor Michigan 293 305 Sun Yan A Carlstr6m and J Askne XXXX Int J Remote Sensing Vesecky J F M P Smith and R Samadani 1990 IEEE Trans Geosci Remote Sensing 28
87. ed the value of 255 You can then WIM User s Manual 67 6 5 Geo Operations 06325 39096 3 06325 39096 3 06325 39096 3 06325 39096 2 4 L 2 2003 08 04 05 28 41 26 000 239 000 0 000 401647448 06325 39096 2 06325 39096 2 06325 39096 3 store the track as an image File Save As Image or overlay file File Save As Overlay or use it as an overlay for another image In order to overlay the track on another image first switch to the image and then use Multi Overlay Image Note Keep in mind that WIM usually shows longitude corresponding to X coordinate before latitude corresponding to Y coordinate and not vice versa Also be careful to use negative numbers for both western longitudes e g 118 30 0 or 118 5 and southern latitudes The total number of tracks in one file is limited to 100 If any of the track points appears to be outside of the image the respective X or Y value is assumed to be either 0 if negative or the maximum X or Y value Note The first line is always regarded as a header comment and skipped Please note that pixels are picked along the track from the beginning to end and then plotted from left to right If your track goes from east to west you will see a reversed graph on the screen Rectangle In the Rectangle option rectangular areas are specified by their upper left and lower right corners in either video or geo referenced coordinates The file formats are exactly the sam
88. edcdaac usgs gov glcc images gif igbp lgnd jpg for SST etc Global low resolution datasets the corresponding high resolution 1 km and 100 m datasets are installed from the WIM CD boundhigh dat boundhigh ndx boundlow dat boundlow ndx coast crude b coast full b coast high b coast inter b coast low b coasthigh dat country boundaries in high resolution country boundaries in low resolution GSHHS shorelines in crude resolution GSHHS shorelines in highest resolution 100 m GSHHS shorelines in high resolution GSHHS shorelines in intermediate resolution GSHHS shorelines in low resolution coastlines in high resolution using GSHHS is faster WIM User s Manual 123 23 List of Files coasthigh ndx coastlow dat coastlines in low resolution using GSHHS is faster coastlow ndx GEB 050 DI6 50 m isobath sample for the US east coast GEB 100 DI6 100 m isobath sample for the US east coast riverhigh dat rivers in high resolution riverhigh ndx riverlow dat rivers in low resolution riverlow ndx world bathy 5min dat world bathymetry for Geo Bathy Image at 5 min resolution world bathy 2min dat world bathymetry for Geo Bathy Image at 2 min resolution Terascan Map Files Sample Terascan projection files of North America in albersco hdf Albers conic equal area projection equidist hdf Equidistant azimuthal projection lamberta haf Lambert azimuthal projection lambertc hdf Lambert conic projection mercator hdf Merca
89. ee ee ee 57 PONTE Saye ANA NANA BANANA ANA ees 57 Hii AA AA AA AA AA AA AA AA aan 58 RGB LiCET AAAH AYA PARA APA AAP AAP PAPA PAPA PARA 58 SpeclrallPI0 am AA 59 211716 o AA AA AA AA AA 60 means ee SS ee Pe eee 61 X Y SOU cise cease encase pee AA 61 6 5 Geo Geo Operations aaa AA AAAAAN 62 Bathy Magpa ce cata deca eaeens ABER AA 62 Get Bathy OVEN AY ARA AA cteenece cine 62 T Get Map Overlay Ka AA 63 Get Vector DIEE ES LA 64 CLARA AA AA PAKA NA 70 DISUNCE NAA ee 70 Read Vector Objects from HDF nnmnnn 71 Remap PROC QUOI Naa AA AA 71 Unify Ge0 60E Maa BADA E IANG AAO 72 Unify Vector OU CCUG a AA AA NBA Aa 72 View in Google MaPEaaa He wv ne ee ee mre NAAN 72 View In Google Earth iis aaa AA AA tere eee 73 6 6 Transf Image Transformations cccsseeeeeeeeeeeeeeeeeeeneeeeeeeeeeeeeeeneneeeeeeeneeees 74 2 Byte TO css AANI NANA 74 LT gy 4 AAAH PAA 74 ES RAEN id atc ANAN BIN aa aaa aaraa Aaaa 75 AS a a a Ne eee er oT 75 Convert to 24bpp RGB ANA 76 Decimal EXp 1072 ga AA AA AA AA 77 Decimal Log LOG VOR sicsestsccixccixecsnccextcixecixecincbinrcinetivenisccastittcieiataseecinediecas 77 Filaments PAA AA AA AA AA 77 PUG AA AA AA AA 77 Filter 10 DISK induire aa inadai iaia aAA EAA AEEA NEA ATE 77 0 Te NAAN piece nals duc aA ARAARA ANARE PAPA KSAPA KREAKE PAA KAPA 78 Linear WANS Xa NAINA ANAN 78 UAROTA E AABANG 78 Reduce IMAGE ak ANAKAN iti ete ee 78 Replage Valys ia NAKAKABATANG NANANG 78 S lar Si AA
90. eloped for classifying a highly variable speckled synthetic aperture radar SAR image and are not designed to work effectively with other less variable images where simple filtering and thresholding could yield better results As these routines have not been updated for a long time please use with caution Typical classes for a SAR image are open water even ice uneven ice A sequence of segmentation routines is usually the following e g Skriver 1989 edge detection edge thinning edge linking edge distance calculation kernel finding kernel connecting and kernel growing pixel tracing The routines in this section are mostly following the ideas in the paper of Sun Yan et al They should be considered experimental and a lot of experimenting is needed on behalf of the user to get an acceptable result Find Edges Finds edges and edge directions by the gamma ratio method Madsen 1986 The method compares ratios of pixel values in all possible directions around each pixel and assigns the maximum ratio value to the edge image pixel The corresponding edge direction image is saved in the edge direction image with the following coding 1 E W 2 NE SW 3 N S 4 NW SE 5 W E 6 SW NE 7 S N 8 SE NW The first direction corresponds the higher side in the window Warning this is a compute intensive operation and may take a long time on a PC without a math coprocessor The best window size depends on the variability of the da
91. em then use Fill gaps with Forward mapping Anther way of eliminating gaps is to use the Max or Mean Pos filters in Transf Filter It is also possible to automatically remap the image to Linear projection This operation finds the full latitude longitude range of the image creates a respective Linear projection and remaps the image to that projection This is useful for images in LLA Latitude Longitude Arrays projections such as MODIS and GLI Level 2 images LLA projection is versatile but very slow It is therefore often useful to remap a LLA image to Linear projection using Automatic parameter setting and deselecting Use projection settings from WIM User s Manual 71 6 5 Geo Operations another image You can experiment with the Forward and Fill gaps settings Be careful when using Inverse mapping as it may be very slow Remap Image Projection From Projection x size 419 Y size 636 Proj Param Center Lat 0 000000 Center Lon 0 000000 Lat Range 0 000000 0 000000 Affine trans 0 000000 fo 000000 0 000000 0 000000 0 000000 0 000000 Create new projection M Automatic parameter setting Use projection settings from another image Zz M Forward mapping T Fill gaps To WIM native projections Terascan Projections X size ns Linear fe E f Y size 636 oj E E Center Lat pi fo fs 5 Center Lon g g fa Lat Range 10110 z C C f5 Co is fs Proj param 0 000000 po Affine tans 0 000000 0 000000 0 000000
92. en the last number format is used for all of them This file format was introduced for easy transfer of image data from WIM to various GIS packages Be warned that this format uses a lot of disk space For example if a BYTE image would be save in the Lat Lon Value format using the default formatting string the size of the file would increase 21 times as each BYTE will explode to 21 bytes 6 for each of the three values plus 3 separators The same function is used for a rectangular area of interest the Rectangle object when selecting a Rectangle in the Vector Objects table and clicking on the Save Lat Lon Value button Using Geo Get Vector Objects Rectangle makes it convenient to fix the corner points of one or more rectangular areas of interest and save the values of these rectangles as Latitude Longitude Value triplets from a series of images GIF saves the current image bitmap as GIF JPEG saves the current image bitmap as JPEG PNG saves the current image bitmap as PNG Portable Network Graphics file WIM User s Manual 30 6 1 File Operations TIFF Saves the current image buffer with the current color lookup table in a TIFF Tag Image File Format file that is portable between different hardware platforms and operating systems Only 1 byte per pixel image buffers are supported by transforming to TIFF 6 0 Class P format palette color images If the image is in the Global Equal Angle projection then Geo
93. er has made the Chl a image the current image by clicking on it However the actual sequence of images can be adjusted in the dialog box The function needs to calculate the day length for each pixel If the Chl a image has attributes Start Day and End Day as the standard SeaWiFS OCTS and MODIS images have the program uses the middle day of the start and end times to calculate day length If the Chl a image does not have those attributes the program prompts for the Julian day e g January 10 of any year is Julian day 10 of the calculation period The function creates a new image of int16 NPP values with the scaling parameters of Slope 1 0 and ntercept 0 0 Please check out the attributes on the Toolbar of the calculated image with View Attributes It is essential that all three input images are of the same size and projection The VGPM model was developed for monthly images but it is also possible to apply it to shorter time intervals e g 8 day composites The function produces results only for pixels that are valid for all three of the input WIM User s Manual 87 6 7 Multiple Image Operations images As the SST images may have large areas blocked due to cloud contamination when using short compositing period it may be advisable to use longer compositing e g monthly for SST even when using a shorter e g 8 day Chl a images SST usually does not usually change very rapidly and the influence of smal
94. es The pixel values are saved according to the Value Scaling option in View Settings General E g if Log Chl or any SST option is selected as Value Scaling the pixel values are converted to mg m or respectively Each point is regarded as the center of a small 3 x 3 pixel area and the following information is appended to each line Value N Valid N Invalid Min Max Mean Median Here Value is the pixel value at the specified point N_Valid is the number of valid pixels in the 3 x 3 area N Invalid is the number of pixels outside the valid range Min Max Mean and Median are statistics for the 3 x 3 area There is a sample point file calsta81 in the long geo referenced format corresponding to a US West coast CZCS image calchi81 img The extensions for point files can be either csv or pnt If a listed point happens to be outside of the image the value and statistics are recorded as 99 999 You can easily create a point file using any text editor or by Examine Point Save You can either pick individual isolated pixels or use free hand drawing to pick pixels along a curve The file saved with Point Save can be WIM User s Manual 66 6 5 Geo Operations read with Geo Getf Vector Objects Point depending on the selected geo coefficient options Transect is the option where the endpoints of straight lines are specified respectively in either video or geographic coordinates In the video coordinates 0 0 is in
95. es has been highlighted click on the Open button of the dialog Attributes of the composited image can be seen by using View Attributes NAVOCEANO L3 dat Starting from January 2002 JPL PO DAAC has made available gridded 18 km MCSST Level 3 NAVOCEANO images their product 144 as weekly composites As explained above you can easily composite NAVOCEANO data with WIM but the composites provided by PO DAAC save time and provide additional products The format that these Level 3 NAVOCEANO images are distributed is Band Sequential binary It is a simple format with 5 images appended to each other Each image has 2048 columns and 1024 rows and uses 1 byte per pixel The bands are Sea Surface Temperature the number of points per bin Sea Surface Temperature Anomaly Interpolated Sea Surface Temperature and Interpolated Sea Surface Temperature Anomaly The files are separated by satellite NOAA 16 of NOAA 14 and by pass daytime or nighttime The value scaling used for SST is SST Pathfinder i e a Linear scaling with a slope of 0 15 intercept of 3 0 The intercept for the SST anomaly is 20 0 while the slope is the same Starting with version 5 45 WIM has is a special file type for reading NAVOCEANO Level 3 datasets while previous versions of WIM can read these files as NAVOCEANO L3 or as generic band sequential binary images When read as NAVOCEANO L3 file type WIM automatically sets the scaling and projection parameters When read a
96. esponding flag groups Two bit can code 4 numbers from 0 to 3 These numbers have the following meaning for MODIS quality products 0 good 1 questionable 2 cloud or sun glint contaminated 3 bad The meaning of the different flags are usually given in the image attributes e5 Convert Makes transformations between 1 2 and 4 byte per pixel image buffers with various Value Scaling For example phytoplankton pigment chlorophyll concentrations in the ocean vary over several orders of magnitude The distribution is usually log normal with the accuracy of measurement approximately linearly related to the concentration In order to compress a wider range of concentration WIM User s Manual 75 6 6 Image Transformations values into one byte different logarithmic scalings can be used see Settings Value Scaling When using the Convert on an image and a transverse conversion on the result the original values should be restored However some discrepancies are inevitable due to the range limitations and the numeric rounding errors A separate option is available for overflow control For example when converting floating point numbers larger than about 66 to Byte values with Chl scaling the pixel values get bigger than 255 that is the highest value for a Byte image Overflow replacement can assure that these values will be replaced with the maximum value of 255 and not truncated to a small pixel value that will probably n
97. f Convert Byte SST PATHF WIM User s Manual 110 17 AMSR E from University of Bremen 17 AMSR E sea ice data from the University of Bremen Sea ice concentration from the AMSR E Advanced Microwave Scanning Radiometer sensor for both the Northern and Southern hemispheres as processed by the University of Bremen in Germany and is available from http iup physik uni bremen de 8084 amsredata asi daygrid swath l1a s6250 The geo referencing information has to be read from an external HDF file e g LongitudeLatitudeGrid s6250 Antarctic hdf for the Southern hemisphere The data is in the format of SDS in HDF4 and as such is readable by WIM However some adjustments were made to WIM in order make the use of these datasets easier Namely the color scaling Min and Max are automatically set to O and 100 respectively corresponding to ice concentration from 0 to 100 see View Settings In order to better visualize the data it is advised to convert the Float32 datasets to Byte with Pixel Value scaling with Transf Convert Byte Pixel Value The loss in accuracy less than 1 should probably be acceptable WIM User s Manual 111 18 New Generation SST 18 New Generation SST for Open Ocean New Generation Sea Surface Temperature NGSST for Open Ocean http www ocean caos tohoku ac jp merge sstbinary actvalbm cgi eng 1 is a merged SST product for the ocean around Japan that is being assembled from various sour
98. f the Defense Meteorological Satellite Program In order to facilitate the use of these data sets a few additions were made to the WIM program The NSIDC SSM I data products are in four specific grids in the polar stereographic projection n3a n3b s3a s3b respectively for the Northern and Southern polar regions These grids projections are automatically recognized by WIM when reading the corresponding HDF data file A specific operation Multi Polarization Ratio was added for the horizontally and vertically polarized microwave signals WIM User s Manual 97 SeaWiFS Products 8 SeaWiFS Products SeaWiFS standard products are delivered in HDF format Image data in HDF is delivered in raster 8 2 dimensional images with 1 byte per pixel and scientific data sets SDS WIM can read both raster 8 and SDS data types and display them as images Special enhancements are included to make it easier to work with SeaWiFS standard products from the Goddard DAAC http daac gsfc nasa gov and the SeaWiFS Project http seawifs gsfc nasa gov SEAWIFS html When reading Level 1A products WIM reads the radiance counts in selected bands out of the total of 8 SeaWiFS bands corresponding to center wavelengths of 412 443 490 510 555 and 670 nm and the attribute information The projection of Level 1 images is automatically set to Swath and the projection parameters are read from the HDF file When reading Level 2 standard
99. from O to 255 in the consecutive color In case of a grayscale LUT the red green and blue components are equal 256 rows are expected If less than 255 color triplets are read from the file the missing values are assumed to be white 255 255 255 In order to use the specified LUT in unmodified form set the Start and End pixel values to O and 255 respectively with View Set Colors WIM User s Manual 31 6 1 File Operations You can create a sample LUT file corresponding to the default Rainbow palette by using File Save LUT you may want to name it rainbow lut The LUT in spectrum lut is another smoothly changing color palette You can check how the red green blue components look like with Edit LUT Edit A popular LUT file for chlorophyll and other images chl lut is included Another sample LUT that is included with WIM is the one used by the IGBP Global Land Cover Characterization program It has specific colors for 17 different land cover types but the rest of the LUT colors 18 to 256 are unspecified and filled up by white 255 255 255 for WIM See http edcdaac usgs gov gicc images gif igop_lgnd jog for more information You can create your own LUT files by modifying the current LUT and then saving with File Save LUT Load LUT raw Loads a custom look up table LUT for color coding from a file and sets the palette flag to Custom see View Settings General Palette The default custom LUT until
100. ge unsigned integer or float respectively the operation finds the minimum and maximum values and then scales the pixel values between them by setting the Min value and Max value see View Settings General Color Scaling The Start and End values of color stretching are not changed Note that the SeaWiFS Level 2 Chlorophyll a and CZCS pigment products are signed 2 byte integers and their typical Min and Max suitable typical for color coding are around 32000 and 29000 corresponding to concentration range 0 3 mg m respectively an Loop Images Plays a movie loop of the images in memory It switches sequentially from the first image to the last and starts all over again Select a suitable delay period 1 60 s between switching to another image This operation is useful for detecting changes between different images Select it again to stop the loop Add Color Scale Adds a vertical color scale to the left side of the image A more versatile color scale horizontal can be added with View Annotate WIM User s Manual 54 6 4 Examine Operations 6 4 Examine Examine Operations Color Consists of the following options Table Shows the current color table with the corresponding pixel values If a specific Value Scaling option has been selected View Settings General Value Scaling the displayed values are the calculated geophysical values e g pigment concentration in mg m or temperatures in
101. ght file with the LLA The algorithm uses the longest common substring to suggest the best matching filename with the LLA For example having the following files in the folder and trying to load a dataset from the first file we need to select the third file in this list for geo location LLA and not the second file in the list The algorithm uses the longest common substring to suggest the best matching filename with the LLA MYDOCL2B A2003168 2115 003 2003179212054 hdf lt getting data from MYDOCL2B A2003181 2045 003 2003192152344 hdf lt similar data file WIM User s Manual 102 11 MODIS Products MYD03 A2003168 2115 003 2003170050731 hdf lt file with ILA This seems easy when we have just a few files in the folder but becomes overwhelming with many similar files with long names Starting with WIM version 6 17 this algorithm was improved by eliminating some file types that are known to not have the LLA Most MODIS Level 3 mapped products and Level 2 products are scaled as 16 bit signed integers with a linear scaling Slope 0 001 and Intercept 5 0 The default color scaling range for these provided by WIM is 5000 to 10000 in pixel values corresponding to 0 0 to 5 0 in the geophysical value The user can change these default values in the Current Settings dialog box The 16 bit integer images can be converted to 8 bit unsigned integer byte images with the Transf Convert function The function Multi
102. gs of an image are shown in the in the Current Settings box Settings icon on the toolbar or View Settings in the menu For example check that the scaling for the chlor a image in file composite hdf is Logarithmic with Slope 0 015 and Intercept 2 0 equivalent to the ChI scaling and the projection is Linear You can read multiple images at once by clicking on the Open icon on the toolbar menu File Open and selecting multiple files For example you can read all the HDF files from Images hdfiSeaWiFS baja_2000_april in one step Each file can contain multiple images of its own For example you can load 3 SeaWiFS Level 1A images from sw_ 1a_mapped_byte hdf in folder ImagesihdflSeaWiFS and then try Examine RGB Image and choose the 412 555 and 670 nm bands as the red green and blue components respectively You can generate almost unlimited color combinations by adjusting the Min down and Max up scrollbars for each component WIM User s Manual 12 4 Getting Started WIM User s Manual 13 WIM Basics 5 WIM Basics The basic data unit for WIM is a digital raster image An unformatted raster image shortly an image is simply a sequence of numbers representing picture elements pixels The number of pixel values has to be the number of rows times the number of columns The sequence of pixel values starts from the top left corner and continues row wise to the bottom right corner A pixel value is often
103. has its projection set you can specify the rectangle used in the statistical calculations either in pixel or geographical coordinates by selecting between X range Lon range and Y range Lat range E g if X range has been selected the longitude range will be calculated from the X coordinates whereas if the Lon range has been selected the X coordinates will be calculated from the Lon range Please observe that in the video coordinates the Y axis starts from the top e g the top left corner has coordinates 0 0 whereas in the geographical WIM User s Manual 60 6 4 Examine Operations coordinates latitude increases from bottom to top If an area of interest on the image has been deselected by clicking on the image the X Lon and Y Lat selection made on the Start Statistics Calculation dialog box will be preserved and can be used for another image To reset the area selection run Statistics without a selected range The described way of calculating statistics can be used for any rectangular area or transect For irregularly shaped areas there is another variant of the Statistics function that uses a second image a Mask to specify the areas of interest Masked areas can be of any shape To use that check the Use Mask check box and specify the mask image number and the mask number The mask image uses pixel values different from O to specify pixels for which statistics is to be calculated You can create a mask image in many
104. he maximal number of pixels that will be traced in both directions along the assumed edge direction If another edge pixel is found the intermediate pixels are converted to new edge pixels value 128 different from the original edge pixels 255 If image border or is reached or the maximum number of gaps traced has been reached without reaching another edge nothing is converted Quite often this routine does not produce acceptable results as the gradients may be followed in parallel to another edge You can then try Dilate and Thin Line operations from the Edge menu section or may wish to skip this operation at all Distance From an edge image creates a new image of distances to the nearest edge pixel using a method called chamfering Borgefors 1983 It includes thresholding where you can specify the threshold level between edge and not edge pixels Normally edge pixels are set at 255 and not edge pixels at O If the distances get larger than 255 the routine has to scale them in order to be able to store the values in an 8 bit image The distances are transformed by taking square root of them and then scaled from 0 to 255 Kernel Find Finds local maxima i e kernels for segmentation and the gradient directions Makes 2 new images the kernel and the gradient direction image The gradient direction image is later used for connecting adjacent kernels as well as for growing the kernels and tracing the pixels to kernels The gradient
105. he source image has more than 8 bits per pixel In case where the different bands can be used individually it might be more efficient to organize the data in the BSQ format This would allow accessing the bands with the simple File Open Image method possibly skipping the previous band s as a header CoastWatch Reads a NOAA NESDIS image file format see http www nodc noaa gov NCAAS ncaas home html that contains compressed or uncompressed image with ancillary information and potentially a graphic overlay with coastlines and latitude longitude grid Depending on the source of the CoastWatch file the graphics overlay buffer may be missing or useless The image dimensions as well as other ancillary data are automatically read from the file header Although the CoastWatch image dimensions are usually 512 x 512 pixels other sizes e g 1302 x 1401 for the Northeast US full resolution image can be handled as well As the CoastWatch pixel values are 11 bit values they have to be transformed Four options are presented Truncate Albedo Temperature and 2 byte pixel buffer You should normally use either Temperature for thermal bands and SST or Albedo for visible bands Albedo and Temperature transform the values into albedo percentages or temperatures respectively When Temperature scaling is selected the resulting image is scaled as X10 i e temperature in C multiplied by 10 or as SST Pathfinder Truncate with Shift 0 picks the leas
106. ich and above which the specified percentage of values are E g by specifying 10 you will see the corresponding pixel values above which the top 10 and respectively the bottom 10 of the values are Values Shows the numerical values of the previously calculated histogram in a table with eight values in a row Line Save Writes the X and Y coordinates and the pixel values of the currently specified line excluding the last pixel as ASCII numbers into a file each triplet on a new line This function is only accessible if Line is selected by View Settings General Area of Interest A similar routine File Save As ASCII only records the pixel values without the coordinates If image projection is set for Linear projection any of the geo conversion coefficients have to be WIM User s Manual 56 6 4 Examine Operations different from O see View Settings Projection the x and y coordinates recorded in the file are the Longitude and Latitude otherwise they are the screen video coordinates If Value Scaling x 10 Pigment Scaled or SST has been selected in View Settings General Value Scaling the recorded values are saved in their respective units otherwise they are the pixel values Peeker Allows you to view peek the pixel values in a 5 by 5 block centered at the currently selected pixel top left corner of the selected rectangle or the starting point of the line The 5 by 5 pixel windo
107. iew Settings Projection Current Settings PA General Extensions Projection Special WIM native projections C Unmapped TY Mere n3a Glob eq angle C Linear C Conic C n3b Shift 160 00 C Mercator Albers s3a C Swath Level 2 C PolStereo Equidist s3b In contrast to similar SeaWiFS maps the OCTS maps are shifted by 160 degrees You have to type it into the Shift field You can then use e g Geo Get Map Overlay to generate coastline and other overlays in high or low resolution A wealth of information about the image is included as attributes metadata and can be viewed with View Attributes The OCTS global GAC dataset was reprocessed in 2001 by NASA and is available at http seawifs gsfc nasa gov cgibrs octs browse pl The format of these files is identical to those of SeaWiFS e g no spatial shifting is necessary WIM User s Manual 100 MOS Products 10 MOS Products The Modular Optoelectronic Scanner MOS was a spaceborne imaging pushbroom spectrometer in the visible and near infrared range of optical spectra 400 1010 nm which was specially designed for remote sensing of the ocean atmosphere system MOS PRIRODA and MOS IRS instruments were basically identical providing 17 spectral channels with medium spatial resolution in the VIS NIR The advanced instrument built for the IRS spacecraft has one additional channel in the SWIR at 1 6um More information on MOS can be f
108. image starting at a specified position in the target image If you specify a target image index that is not allocated in memory any number larger than the last image buffer a blank image buffer with the specified size is allocated and used for the insertion Linear Comb Calculates a generalized linear combination of up to three images or image ratios pixel wise Linear Combination X n ling Result X cq ZEL WIM User s Manual 84 iz INE i Ci Img il Img i2 2 MB NP Ba S Ss Ka eo eee Set a Kai OK Cancel 6 7 Multiple Image Operations By either selecting or not selecting the second image each term is either a an image multiplied by a coefficient or a ratio of two images multiplied by a coefficient Images are selected by their sequence number see List of Images The resulting image is always a 4 byte per pixel float image Mask w Image This function allows selection of the pixels depending on the pixel value of another image i e a mask image A new image is created with only the selected pixels retaining their value while the other pixels unselected will be converted to zero pixel value zero The pixels that are selected on the mask image retain their values whereas those pixels whose counterparts on the mask are not selected will become zero It prompts for the mask image index the mask pixel value and whether to select all pixels s
109. ions DX and DY have to be used or the RLE decoding may not work A sample RLE encoded image s_califm rle contains the latitude longitude grid of the s calif1 img and s_calif_2 img images in Conic projection RLE encoding is very efficient for storing images with large constant areas e g overlays with latitude longitude grids and coastlines WIM User s Manual 19 6 1 File Operations You can produce RLE compressed files with File Save As Compressed RLE Erdas Lan Reads a special file format introduced by a popular image processing program by Erdas Inc The 128 byte header contains information on the number of bands band interleaved by line BIL the pixel size 4 8 or 16 bits and image dimensions The first 4 bytes of the header are assumed to be HEAD At present only 8 bit pixel images can be read The extension of Erdas Lan files is usually an The user is given the option to select the set of bands and the subset of the X and Y ranges to be read Subsetting is sometimes required if the Erdas Lan file has been produced on a Unix workstation and is too big to be handled by the PC The names of the image buffers will have the number of the band appended Please note that the bands are numbered from 7 to N but the lines and pixels are numbered from 0 to M 1 where Mis the Xor Y range of the image Float Reads a plain unformatted raster image with each pixel represented by a 4 byte float number in the
110. is has to be entered The plot ranges are initially between the respective minimum and maximum pixel values on both axes Subsequent refinement of the Min and Max values on both exes allows filtering out pixels affected by clouds missing data or examine the relationships for different clusters of pixels The selected clusters of pixels can be saved in an ASCII 2 column file This file can then imported into another program e g MS Excel spreadsheet for further analysis or plotting WIM User s Manual 61 6 5 Geo Operations 6 5 Geo Geo Operations Bathy Image Reads a bathymetry database file with ocean depth values and creates a Int16 image of bathymetric depths in meters The current source image must have a defined projection and the created bathymetry image will have the same size and projection as the source image You can easily create a depth image for any area of the world by first creating a new blank image in Linear projection File New and then creating a corresponding bathymetry image WIM includes two world bathymetry databases a lower resolution 5 minute latitude longitude grid world bathy 5min dat and a higher resolution 2 minute latitude longitude grid world bathy 2min dat both in the Maps folder The bathymetry image is created from the data in these files using bilinear interpolation After creating the bathymetry image you can create depth contours using the Examine Contour Lines function Fo
111. ither the Nearest Neighbor Bilinear or Bicubic method With the Nearest Neighbor method new image is double in X and Y size The Bilinear and Bicubic methods produce smoother looking images and also allow to select a floating point scaling factor When the scaling factor is less than 1 0 the new image is actually smaller than the source image Please note that a somewhat related function View Zoom zooms the viewable bitmap and does not create a new image buffer WIM User s Manual 80 6 7 Multiple Image Operations 6 7 Multi Multiple Image Operations Multi consists of operations that are used to transform several images to a new image It is usually assumed that the current image is the first operand and you have to specify the second operand image by its index number The resulting image is usually recorded in a newly allocated image buffer However with Mask w Image and Overlay Image you can also record the result in the current image by specifying the second image index as a negative number This option may be preferred if your images are very big and you want to save RAM Add 2 Images Adds two images pixel wise If the Value Scaling of an image is set see View Settings General Value Scaling the decoding operation is performed before the addition In case when both images pixel size of 1 byte if the sum is larger than 255 255 is assumed as the new pixel value In case of images with different pixel size the
112. l variations in SST on the calculated product is usually of the order of 10 that is well below the typical uncertainty in the NPP calculation Primary Production Images OK Chl fo PAR 1 Cancel Photoinhibition Euphotic zone depth Present VGPM standard Morel Berthon 1989 Not Present GPM modified C Morel Maritorena 2001 C SPG Southern Ocean PBopt temperature model Behrenfeld Falkowski 1997 Eppley 1972 Antoine 1996 C Constant 4 54 mg C mg Chl fh Sharpening Resolution sharpening uses one band with higher spatial resolution to increase the spatial resolution of another band For example the MODIS sensors have bands 1 and 2 620 670 nm and 841 876 nm at 250 m resolution bands 3 7 at 500 m resolution and bands 8 36 at 1000 m resolution Assuming consistent ratios between measured and interpolated bands it is possible to create fake 250 m resolution images from bands 3 36 For example in order to create a fake MODIS band 3 image at 250 m resolution we need band 1 image at 250 m resolution Full resolution Band 1 band 1 sampled at 500 m and interpolated with linear interpolation to 250 resolution Interpolated Band 1 band 3 sampled at 500 m and interpolated to 250 m resolution Interpolated Band 2 The simple algorithm used here assumes that the ratios of full resolution and interpolated i e smoothed bands are constant within each spectral band More details on how to
113. lags that are ON 1 can be viewed by right clicking on the image These flags are then listed in the window header Alternatively areas where selected flags bits are ON can be shown by creating mask images with Transf Bitmask SeaWiFS Level 2 flag bit numbers along with their associated masking keywords are 1 atmospheric correction algorithm failure ATMFAIL 2 land LAND 3 missing ancillary data BADANC 4 Sun glint HIGLINT 5 total radiance greater than the knee value HILT WIM User s Manual 98 SeaWiFS Products 6 large sensor zenith angle HISATZEN 7 shallow water COASTZ 8 negative water leaving radiance NEGLW 9 stray light STRAYLIGHT 10 cloud or ice CLDICE 11 coccolithophores COCCOLITH 12 Case 2 turbid water TURBIDW 13 large solar zenith angle HISOLZEN 14 high aerosol concentration HITAU 15 low water leaving radiance at band 5 LOWLW 16 chlorophyll algorithm failure CHLFAIL 17 questionable navigation NAVWARN 18 absorbing aerosol ABSAER 19 tricodesmium TRICHO 20 NIR algorithm exceeded maximum iteration MAXAERITER 21 moderate sun glint MODGLINT 22 chlorophyll lt 0 01 or gt 64 CHLWARN 23 epsilon out of range ATMWARN 24 dark pixel DARKPIXEL 26 spare SPARE 27 spare SPARE 28 spare SPARE 29 spare SPARE 30 spare SPARE 31 spare SPARE 32 ocean OCEAN When reading Level 3 Standard Mapped Images WIM automatically sets the projection to Global Eq
114. larger than 255 it will be set to 255 By specifying a negative shift number you simply select the case when the LSB is first but with the same absolute shift to the right Specifying 0 shift produces the second byte value only if the first byte is zero i e the integer is less or equal to 255 and 255 if the first byte is different from zero For the corresponding operation of 0 shift assuming that the LSB is first you have to select 16 By selecting shift 8 or 8 you will extract the first second byte for the new pixel value By selecting shift 17 17 the square root will be taken from the integer value to compress it into a byte assuming that MSB is the first second Special options are available for the case when the 2 byte value specifies temperature in K multiplied by 10 By selecting 18 or 18 when MSB is second the value is converted to C multiplied by 10 i e 2731 5 is subtracted By selecting 19 or 19 the value is converted to a coded temperature see Transf Convert that keeps temperature values between 3 WIM User s Manual 25 6 1 File Operations and 27 C In both cases temperatures outside the possible range get respectively the minimum or maximum value A related operation File Open 2 byte Integer reads a similar image file but keeps it as a 2 byte buffer instead of converting into a 1 byte buffer Overlay Reads a plain unformatted raster image with each pixel represented
115. latitude longitude of a plane tangent to the sphere The Conic and Albers projections may use all 4 reference variables In a conic projection the globe is projected onto a cone If the cone is tangent to the globe at one latitude only ref reference latitude and ref2 reference longitude are used The other variables ref3 and ref4 are needed only when the second reference plane is used Remapping is compute intensive and can be slow for large images On multi core and multi CPU systems the function is multi threaded and uses multiple cores CPUs concurrently Unless the target image is much smaller than the source image the fastest method to use is Forward or direct mapping that takes every source pixel and puts it into the target image in the right position The problem is that when the target image is larger than the source image then the result will be gappy i e there will be pixels which did not get a value from a pixel in the source image Gaps will not occur in case of inverse mapping as for each pixel in the new image the best matching pixel is located in the source image Inverse mapping usually generates better looking images However for large target files inverse mapping is very slow A compromise method is to first use Forward mapping and then inverse mapping Fill gaps only for those pixels that did not get a value with Forward mapping We advise to experiment start with Forward mapping and no Fill gaps If gaps are a probl
116. le coding CZCS Pigm The pixel values are assumed to represent phytoplankton pigment Chlorophyll a plus phaeopigments concentration mg m Pigment 10 0 012 x Pix Value 1 4 Pix Value log10 Pigment 1 4 0 012 This coding was often used with the Coastal Zone Color Scanner data The pixel values in the sample image calchi81 img are scaled as pigment concentrations You can view them and use in various operations when selecting CZCS Pigm for that image Note that the CZCS compatible pigment image of the SeaWiFS Level 2 product is actually a 2 byte per pixel buffer and uses Linear Value Scaling with Slope 0 001 and Intercept 32 0 Log Chl The pixel values are assumed to represent phytoplankton chlorophyll a concentration mg m chlor a 10 0 015 x Pix Value 2 0 With 1 byte per pixel pixel the approximate range covered is 0 01 67 mg m This is a special case of the generic Logarithmic scaling You can also set the scaling manually by selecting Logarithmic with the respective Slope and Intercept LOG MINUS4 This is similar to Log Chl but is used for geophysical values that are much too low for the Log Chi scaling val 10 0 015 x Pix Value 4 0 With 1 byte per pixel pixel the approximate range covered is 0 0001 0 668 This is a special case of the generic Logarithmic scaling With a generic Logarithmic Value Scaling the following power equation is used to calculate geophysical value
117. led Equirectangular or Plate Carr should be automatically set However these mapped datasets are available at low resolutions namely at 25 km L3m GLOB 25 nc and 100 km L3m_ _GLOB_100_ nc Note that the letter m after level designator L3 means that the products have been mapped A sample CHL1 mean image is shown below Global binned datasets are not mapped in any traditional projection but are binned on an Integerized Sinusoidal ISIN grid with a resolution of 1 24 degrees at the equator 4 63 km The number of columns is decreasing as the distance of a row of bins increases from the equator The data is also distributed in netCDF format but as 1 dimensional arrays WIM is transforming these arrays into a rectangular WIM image format using a specific projection GlobColour ISIN These global binned datasets at 4 63 km resolution have filenames like L3b_ _GLOB_4_ nc The files are rather big 180 MB A sample CHL1 mean dataset is shown below reduced 20X for visualization here WIM User s Manual 116 19 AMSR E from Remote Sensing Systems WIM User s Manual 117 21 MERIS Products 21 MERIS Level 3 products MERIS is an advanced sensor on the ESA Envisat satellite MERIS data are available from http envisat esa int level3 meris Global Level 3 datasets of normalized water leaving radiances at several wavelengths Chlorophyll a for Case 1 waters Chl1 and others are available at Daily
118. less than two Miscellaneous Misc Current Settings General Extensions Projection HDF Options Misc Vector format r Navoceano Size p Image Header x1 yl x2y2 width 1720 fo bytes C xyuy xcor Height 360 m Default LUT IV Use Default LUT lv Override LUT in HDF C Program Files x86 Wimsoft LUT chit_white_end h a Coast Overlay For New Images fc Program Files x86 Wimsoft M aps4coast inter b a m Lat Lon Value Format T Lat first Format string 26 4f 6 4f 26 4fhn Vector Format x1 y1 x2y2 or x y u v xcorr Vector Format toggles between 2 different formats how vectors are represented in a file x1 y1 x2 y2 is the default and x y u v xcorr is an adaptation to the format produced by the Multi Motion Detect function see File Get Vectors Default Navoceano Size The default Width 720 pixels and Height 360 pixels can be changed by the user The image width and height specified here are also use when opening files in the Lat Lon Value ASCII format Image Header In order to skip a certain number of bytes from the beginning of a plain raster image file e g those containing a header you can specify the header length in bytes that will be skipped when reading image files Default LUT Use Default LUT this allows to use a predefined palette file that will be read and used as the default palette for all images instead of the Rainbow p
119. lid Range i e usually clouds or land or otherwise failed You can choose the Valid Range either for the Geophysical values default or Pixel values When dealing with BYTE images e g SeaWiFS Standard Mapped Images it is easy to remember that values 0 and 255 are commonly used for land or clouds i e the Pixel Values should be e g from 1 to 254 WIM User s Manual 82 6 7 Multiple Image Operations Compare Compared two images pixel wise considering the Equality threshold Floating point images e g Float32 or Float64 can be considered equal if their respective pixels differ by a very small to a value The pixel wise comparison stops if a pixel is different between the two images A related operation Multi Difference creates the difference image but you then have to examine that image if any pixel values are different from O or from another small threshold Multi Compare makes this process easier Composite Calculates a generalized average composite of a series of images while trying to eliminate missing pixels e g due to clouds The images must be in a contiguous series of image buffers and with the same image dimensions The composite can be calculated in several ways i taking a pixel wise average of the series of images ii picking the pixel wise maximum over the series of images iii picking the last valid value for each pixel Invalid pixel values can be discarded from the calculation and the number of
120. like text the letter A for annotations pencil for free form line drawing line for straight lines brush for painting paint can for filling with color open and filled ellipse open and filled rectangle You can specify separate numerical values for the outline and for the filled area The rectangular tool in the top left corner disables the drawing action and gives the left mouse button back to specifying the Area of Interest either a rectangle or a line see View Settings General Area of Interest Drawing is limited to Byte and Integer image buffers the values will be wrong for Float image buffers Starting from version 6 41 you can undo the changes and moving back and forth up to 16 Kosa If the background image is a RGB image then you can edit the image with a set of 3 values for each of the Red Green and Blue components You can convert an image to RGB image with Transf Convert to 24bbp RGB The color selection is done with the dialogs shown below steps back using the following icons on the Toolbar WIM User s Manual 35 6 1 File Operations Color 21x Basic colors Ei Tel eee Si a DETE i Tee DE ETETE EEE eee EERE IU Custom colors if i i ip ip 5 ip BEEBE ees Define Custom Colors gt gt OK Cancel LUT Edit Performs interactive editing of the Red Green and Blue components of the LUT Lookup table in memory The default LUT is the grayscale palette
121. loat 4 bytes per pixel for images with no Value Scaling The Count image is always a 1 byte image as the number of valid pixels cannot exceed the number of images in the series lt 256 Value Scaling of the resulting Average image is set equal to the Value Scaling of the current image Value Scaling of the Count image is always Pixel Value Difference The operation is related to Multi Subtract Image It calculates an image of absolute differences between the two images The images must be of the same size Value Scaling see View Settings General is used for calculating the real value before the difference is taken The resulting image is created as a 4 byte Float image that is usually not the best option but is the most versatile In most cases is convenient to scale the resulting image into a Byte or Int16 image using the Transf Convert function Divide w Img Divides the current image with another image pixel wise If the pixel value of the second image is 0 the result of the division is 255 the maximum value for 1 byte per pixel image Value Scaling is used to calculate the real values before the division The images must be of the same size The Value Scaling of the resulting image is set equal to the Value Scaling of the current image and the pixel size is always set to float 4 bytes per pixel Insert Image Allows to insert one image the source to another image the target with the upper left corner of the inserted
122. lon and multiple consecutive separators are treated as one Before reading the file you can also set the color scaling parameters see View Settings General Value Scaling Examples of small 4 x 4 pixel ASCII images in space and comma separated formats are given below 255 200 180 111 100 90 8 122 8 100 100 234 199 212 231 123 167 0 176 0 22 0 24 10 0 133 0 188 0 211 166 0 0628 0 550 0 5 12 0 21 0 1666 0 18 DO O O Band Sequential Line Interleaved Pixel Interleaved Reads files that contain several bands channels of image data either in band sequential BSQ band interleaved by line BIL or band interleaved by pixel BIP formats BSQ is just a sequence of several images concatenated one after another In the BIL format a line row of one band follows the corresponding line row of another band In the BIP format a pixel of one band is followed by a corresponding pixel of another band You can specify any of the bands to be read or the whole set if you specify band 0 Either one or several new image buffers of DY times DX bytes are allocated A special file type Erdas Lan is included for reading ERDAS BIL or BIP files but the same can be accomplished by setting the header length to 128 with View Settings General Header Length A well known image transformation program Alchemy produces BIP files with the BIF option B WIM User s Manual 18 6 1 File Operations when t
123. lor 255 Top 24 Backgound color 255 Right 500 Transparent Font Bottom 44 C Opaque Fm L FAF j Ticmarks Auto generated Custom Comma sepated values L Image Name Date Annotation aaa Cancel 3 205 Annotated 512x512x1 chlor a l 1 T T 0 01 0 04 0 1 0 4 0 9 2 5 6 8 18 0 05 Annotated 512x512x1 chlor a 0 01 0 05 1 0 WIM User s Manual 38 6 3 View Options Attributes Allows to view edit delete add and copy to another image buffer of ancillary information describing the current image e g scaling parameters and global attributes describing the set of images e g the satellite pass date receiving station processing version Attributes are commonly used in HDF When creating your own HDF files with File Save as HDF it is very convenient to add additional information with this function These attributes are retrieved at a later time when reading the HDPF file Global attributes are common to all images in the same HDF file Local attributes describe individual images To access the local attributes of a particular image click on the image and the select View Attributes For NAVOCEANO images the attributes are generated by WIM from the ancillary data in the NAVOCEANO datasets For CoastWatch files the Attributes function is replaced with showing the CoastWatch header data that has similar information but is not modifiable
124. mages and select Multi Overlay Image select the sequence number of the coastlines overlay e Now set the correct scaling with Settings Value scaling For SST you can choose Value Scaling SST PATHF C For all others select Linear with the appropriate Slope e g 6 for TIME 0 2 for WSPD 0 3 for VAPOR 0 01 for CLOUD 0 1 for RAIN The Intercept must be O for all except SST that is 3 but it is fixed in the SST PATHF C scaling As you can see reading AMSR E data in the original format is rather cumbersome Therefore it is better to use the WAM program wam_convert_amsre and convert all available AMSR E binary files into standard HDF files that can be easily analyzed with WIM WAM or other HDF capable software WIM User s Manual 115 GlobColour ocean color products 20 GlobColour ocean color products GlobColour is an ESA project http www globcolour info that produces global and local ocean color datasets from individual sensors or merged from multiple sensors GlobColour global FPS Full Product Set data can be downloaded from http Awww globcolour info data_access full prod set html Local datasets for many diagnostic data sites are available from http www globcolour info data access dds list2 html Global mapped datasets in netCDF format can easily read into WIM as netCDF old version 2 and 3 netCDF files could be read also as HDF datasets Either the Grid or Global Equal Angle projection also cal
125. maller and equal or larger and equal to the mask value The default values 0 for mask value and Yes for smaller and equal are suitable for picking the good pixels of MODIS images MODIS data are delivered with a corresponding Quality flag image that has pixel values corresponding to the quality flag as follows 0 good 1 questionable 2 cloud 3 bad This meaning of the flags here is given as an example only and may depend on a particular convention and data type The default selection would selecting pixels with Mask value lt 0 i e good When selecting pixels with mask value gt 3 for example only the pixels of quality bad are returned Various masks can be easily generated with WIM For example sometimes it is necessary to mask all ocean pixels or all land pixels A simple way to do that is thresholding an appropriate image with Transf Binarize In order to save disk space masks are efficiently stored as compressed images Save As Compressed Motion Detect Calculates the apparent motion between two consecutive images using the maximum cross correlation method e g Emery et al 1991 Several parameters that are needed to tune the subroutine have to be previously recorded in a parameter file A sample parameter file 64 par as well as a pair of test images 64_1 img 64_2 img are provided As this is a compute intensive routine the test images are only 64 x 64 pixels in size in o
126. n set of vector objects When another image is selected the list of vector objects for that image is shown An individual object can be selected by clicking on it after which it starts blinking on the image This feature makes it easy to visualize and locate multiple objects on the image With the Edit button you can manually edit the selected vector object The Delete button deletes the selected vector object The Save button saves the selected object s in HDF file that can be read with the Geo Read Vector Objects from HDF function The Save Lat Lon Value button saves the selected object s in a text file The Show Profiles button shows the profile for a selected transect object The Statistics button shows statistics for each of the applicable object The statistics for a Point is calculated for the window of 3 by 3 pixels centered at the point pixel WIM User s Manual 52 6 3 View Options Vector Objects If you want to copy the set of vector objects from one image to all the other images in memory select the source image by clicking on it and apply Geo Unify Vector Objects Now all the images in memory will have the same vector objects This operation is used when you want the same vector operation to be performed on a set of images but don t want to read the same set of vector objects for each individual image For example you may want to pick the same transect on a series of images extract it get a series
127. nd mapped data are similar to the SeaWiFS and other SMI Standard Mapped Images products The main difference is that the left side of the image is at the 0 meridian and not at 180 WIM is automatically setting the Global Equal Angle projection with Shift 180 see Projections More specialized functions for GLI data will be added in the future as different data products become available WIM User s Manual 105 Landsat Products 13 Landsat Products The Landsat Program is the longest running enterprise for acquisition of imagery of the earth from space The first Landsat satellite was launched in 1972 the most recent Landsat 7 was launched on April 15 1999 These sensors typically provide high resolution 30 m imagery The resolution is enhanced 15 m for a wide band channel and reduced 60 m for the thermal infrared channel The main disadvantage of the Landsat imagery has been its high cost In cooperation with federal and provincial governments of Canada GeoGratis an initiative of the Canada Centre for Remote Sensing is distributing full resolution Landsat 7 imagery for all of Canada free of charge Landsat 7 data of selected scenes of Canada is freely available for downloading at http geogratis cgdi gc ca Landsat7 hdf New scenes will be accessible when they become available The Landsat 7 satellite carries the enhanced thematic mapper plus ETM sensor Landsat 7 data are collected from a nominal altitude
128. ngs CZCS Pigment and Chlorophyll are both suitable for phytoplankton pigment concentrations that often have log normal distribution A scaling of Pixel Value means no value scaling x10 means that the geophysical values have been multiplied by a factor of 10 to get pixel values x10 means that the geophysical values have been multiplied by a factor of 100 to get pixel values SST PATHF C This is the sea surface temperature SST coding used in the NOAA NASA Pathfinder Sea Surface Temperature MCSST data by the Physical Oceanography Distributed Active Archive Center at the Jet Propulsion Laboratory http podaac jpl nasa gov sst SST is calculated according to the following equation SST C 0 15 Pix Value 3 0 SST SMHI C This is a Sea Surface Temperature coding used by the Swedish Meteorological and Hydrological Institute http www smhi se The pixel values are assumed to be representing temperature in C with the following transformation Temp 255 Pix Value 20 7 255 3 Pix Value 255 Temp 3 255 30 This allows having values between 3 and 30 C and at the same time reverses the order of pixel values i e if cold was associated with low values before it has high values after the transformation The idea to reverse the WIM User s Manual 41 6 3 View Options order of pixel values is to have cold temperatures white and warm temperatures dark using a conventional graysca
129. notates with image name title etc The values on the color scale depend on the current View Settings General Value Scaling Please note that the easiest way of selecting a place for the annotation is by mouse just select a rectangular area with the left button where you want the color bar to be then select View Annotate and the color bar will be written into the selected area After you find a suitable size and location you can put the annotation into exactly the same location on other images by NOT selecting an area with the mouse before using View Annotate from the menu The last location of the color scale is always saved for future calls to the function This allows to apply consistent annotation to a series of images A copy of the image is automatically opened with the annotation You can easily make another try to find the most appealing size and location for the annotation A related operation adds a simple vertical color scale can be added to the left side of the image with Add Color Scale The tic marks are either generated automatically or can be specified one by one in the text field Comma separated values In the examples below the auto generated set of tic marks are 0 01 0 04 0 1 0 4 0 9 2 5 6 8 18 0 By specifying tic marks 0 01 0 05 1 0 5 0 and 20 0 only those will be used in the annotation WIM User s Manual 37 6 3 View Options Annotation Ka r Rectangle r Text Options Left 268 Text co
130. nt32 and float also 4 bytes per pixel These images are all read from the same SeaWiFS Level 2 file in HDF format List of Images Image Name FileName Width Height Type Sie Olongitude C Wimdev Images hdf SeaWwiFS Level245 20023 1746K td 1 latitude C Wwimdev lmages hd SeawiFS NLevel21520023 1746K td 2 chlor a C Wimdev Images hdf SeawiFS Level24S 20023 13915K C Wimdev Images hdf SeawiFS Level2 20023 13915 K C Wwimdev Images hdf SeawiFS Level2 S 20023 3478K C Wimdev Images hdf SeawiFS Level2 20023 6957 K Reload Delete Copy The List of Images can also be used for a number of operations on images in memory To switch between different images double clicking on the selected image Reload reloads the bitmap from the image buffer and recreates its bitmap You may sometimes need to run Reload on a selected image buffer when the image seems to have a problem with its colors Delete deletes selected image s from memory Copy makes an identical copy of the selected image to a new image buffer You can also Rename an image name by simply editing the name By selecting an image and then selecting another image while holding down the Shift key a whole range of image buffers can be selected for deleting from memory Multiple image buffers can be individually selected by selecting them while holding down the Control key Renaming an image may be useful if you want to s
131. o save your image in KML and KMZ files for later viewing in Google Earth using File Save as Google Earth KML or KMZ If you view many images then the temporary files accumulate in your temporary folder and you may want to delete from time to time the wim_temp kmz files from there You can do that with Windows Explorer or run the following command del temp wim_temp kmz More information on using Google Earth e g for making black areas transparent can be found in the exercises for ASTER images http www wimsoft com Exercises ASTER pdf WIM User s Manual 73 6 6 Image Transformations 6 6 Transf Image Transformations Transf consists of operations that are used to transform the current image to a new image Transformations that use more than one image as input are included in the Multi menu section multiple image operations While most of the operations record their result in a newly allocated image buffer some transform the current image buffer 2 Byte To 1 Transforms a 2 byte per pixel image buffer to a 1 byte per pixel buffer The options available are similar to those in File Open Int gt Byte This flow level function is not recommended for casual usage Please use the related function Transf Convert that adds much more functionality and is also much easier to use Binarize The idea of the binarize function is to select pixels that satisfy certain criteria for example larger or smaller than a
132. of these transects into MS Excel for plotting You can read the transect file once transfer the vector objects to other images and save each transect with the Save Lat Lon Value button to an ASCII text file You can then read that text file into MS Excel When selecting a Rectangle object all the pixel values within the rectangle can be saved with the Save Lat Lon Value button The result is similar to the menu function File Save as Lat Lon Value ASCII The formatting of this operation can be modified in View Settings Misc Lat Lon Value Format Vector Object Geo Coordinates lon lat Image Coordinates Label Drifter Track NMA Start 39 103 mypoints pnt Transect 120 0000 31 0000 117 0000 30 0000 73 120 183 1 My transect 1 Transect 121 0000 30 0000 112 5000 25 0000 37 150 347 3 My transect 2 Point 116 7830 31 6750 190 100 100 30 Point 117 1170 31 5080 178 105 100 35 Point 117 4500 31 3500 166 110 100 40 La PAT f NAT DEN DA AMIANAN NAGA AHAN ann AT ii Edit Delete Close Save Save Lat Lon Value Show Profiles Statistics Drifter Track is a Vector object that is similar to a sequence of Point objects but is treated as one set The file formats of the Drifter Track is similar to the Point files either the Point pnt or csv or NOAA drifter format DAT Each vector object needs to be in a separate file Text Label is a special type of vec
133. ogram icon on the toolbar or menu Examine Histogram Calculate Many functions Histogram Statistics Profiles etc operate on the selected area of interest either a rectangle or a line To create a coastline overlay for an image select menu Geo Get Map Overlay Then pick a map file with global coastlines e g coast inter b or coast low b All the map files are located in the Maps folder e g in WIM User s Manual 11 Getting Started C Program Files WimSoft Maps or in C Program Files WimSoft Examples WIM memorizes the locations and the parameters so that when you do it the next time you will be directed to that folder at once When creating an overlay make sure that the background pixel value is 0 and the foreground pixel value is either 255 white or 1 black To overlay coastlines or other features on an image click on the image in WIM then on the Overlay icon on the toolbar or Multi Overlay Image in the menu Make sure the image number to be overlaid is correct The current image is assumed when looking up values or performing operations e g Histogram Statistics You can make an image current by either left clicking on it or double clicking on it in the List of Images box The basic information about an image is shown in the List of Images box but more detailed image attributes are shown when clicking on the Attributes icon on the toolbar View Attributes in the menu Scaling and other settin
134. or Scaling attributes WIM Color Range Min Max are also saved Another problem is that when you manually set the projection type and then save the image WIM will not save the newly set projection type The idea is that WIM saves all the HDF attributes but manually setting projection does not change the attributes The solution is to manually create a projection attribute that will be saved and retrieved when the file is read the next time For example when setting the projection manually to Global Equal Angle you need to create a new attribute Map Projection and set it to Global Equal Angle HDF with Lat Lon arrays Writes an image into HDF file see previous section together with separate Latitude and Longitude arrays For each pixel a corresponding Latitude and Longitude values are recorded Having separate Latitude and Longitude arrays for every pixel is the most versatile form of projection however it is also the slowest to use in generating coastlines or other geo referenced objects Therefore if possible saving data in this file type should be avoided A single set of Latitude and Longitude arrays is saved when more than one image is saved to the file All the images to be saved into a file should be of the same size Note the Save as compressed option in Settings HDF Options With compression the file size can be reduced many times Overlay Creates a 1 bit per pixel overlay from the current 1 byte per pixel image
135. ot be desirable A useful conversion is for SST from 2 byte scaled Kelvin images to scaled byte in Celsius using the Kelvin to Celsius option x From To m Pixel Type Value Scaling Pixel Type Value Scaling BYTE C Pixel Value BYTE Pixel Value INTIE6 C x10 INTI6 10 FLOAT C x100 FLOAT C x100 C INT32 Kelvin to Celsius C INT32 Kelvin to Celsius C LOG MINUS4 C LOG MINUS4 J Signed C SST PATHF C P Signed SST PATHF C SST SMHI C Overflow C SST SMHI C C Log Chl mg m3 7 Replace with Log Chl mg m3 Linear Linear 255 Logarithmic ma C Logarithmic Slope Intercept untcaled value Slope Intercept 1 000000 0 0000 TU Issue warning OK Cancel Convert to 24bpp RGB Converts the current image to a 24 bit per pixel RGB image Typical WIM images use a color palette to visualize pixel values as colors but a RGB image has the red green and blue components specified with a 8 bit unsigned integer This function is useful to apply to a WIM image other than RGB to preserve the colors After converting to RGB you can overlay iton a RGB image and the colors will stay the same see Examine RGB Image WIM User s Manual 76 6 6 Image Transformations Decimal Exp 104x Decimal Log Log10 x Perform respectively pixel wise exponential or logarithm operation on the image buffer and create a new FLOAT image Filamen
136. ot very accurate and better methods may be introduced in the future WIM User s Manual 106 GOES SST Products Select Images p x Select buffers to read from L71009029_02920000618_HDF L1G Image Name width Height Type L71009029 02920000615 610 6480 6000 i L71009029_02920000618 5620 6480 6000 L71009029 02920000618 830 6480 6000 L 1009029_02920000618 B40 6480 6000 L71009029 02920000615 650 6480 6000 L 1009029_02920000618 B60 3240 3000 L 2009029_02920000618 B70 6480 6000 LTOANNNMAN MAIMANANAWNA oO MAN sanen 49000 Cancel Above is a typical dialog box for selecting bands from the launch pad HDF file As you can see the image bands are of different size due to the different resolution 30 m vs 60 m vs 15 m A commonly used option for these images is to create a RGB composite of a selected set of three bands When making a RGB composite the component bands have to of the same size You can reduce an image with Transf reduce Image and you can increase the size of an image by 2X with Transf Zoom The RGB compositing can be slow for large images like these therefore it is advisable to find the best color combination on a small subset of the images and then applying that to the larger image set When saving a RGB composite to a file please use Save as 24 bit Bitmap or Save as JPEG as they preserve the correct colors Save as TIFF may not produce the same colors as visible on the screen and Save as GIF may produce
137. ound at http www ba dlr de NE WS ws5 MOS data products of Level 1b are available in the HDF format Both MOS A and MOS B data files contain a 1 byte per pixel quick look image and radiometrically corrected and calibrated 2 byte per pixel band interleaved by line data from all the channels Special functions in WIM unpack the band interleaved format automatically into images Two new functions were added to WIM together with the support for reading MOS images Multi Band Ratio and Multi Linear Comb These functions are applicable to any images but may be useful for analyzing MOS data in particular WIM User s Manual 101 MODIS Products 11 MODIS Products Overview Moderate Resolution Imaging Spectroradiometer MODIS is a key instrument aboard the Terra EOS AM 1 and Aqua EOS PM 1 satellites MODIS is viewing the entire Earth s surface every 1 to 2 days acquiring data in 36 spectral bands For an overview of the instrument and science see About MODIS MODIS data is available at the Distributed Active Archive Centers DAAC s These data are divided into land ocean atmosphere products For information about data availability products and sources see the Data section MODIS has hundreds of different data products see http eosdata gsfc nasa gov CAMPAIGN DOCS MODIS index shtml As the default format of these data products is HDF WIM can read most of them Ocean Products The Mapped Level 3 products
138. ponent images point to the sequence number of the overlay image That makes all 3 components 255 and that produces a white RGB image Similarly you can create overlays of almost any color For example to create red coastlines pick the coastlines image for all the components but disable the green and blue components by making their Low and High scrollbar levels equal to 255 Now you can overlay the coastlines with desired colors on the target RGB image with Multi Overlay Image The Transf Convert to 24bpp RGB function is also useful for combining typical WIM images with RGB images For example you may want to combine a chlorophyll image for the ocean with the true color image for land You can do that by creating a mask for land and for ocean creating separate RGB images for land and ocean masking respectively the land or the ocean part with Multi Mask w Image and then the combining with Multi Overlay Image the masked RGB images Spectral Plot Interactively plots pixel values from multiple images versus image number If the images represent multiple co located spectral bands then the plot is a true spectral plot You should right click on the plot area and drag the mouse pointer over the image For each pixel the plot shows the values of the corresponding pixel as a function of the image number of each of the loaded images The pixel location x y and Longitude Latitude is also shown The images can represent anything and do no
139. presented by up to 3 characters and separated by one or more separators Separators can be a space a tab character a comma or a semicolon Multiple consecutive separators are treated as a single separator Negative values are not allowed are set to 0 and values larger than 255 are set to 255 The following is an example of an image of 4 rows and 4 columns an underscore stands for any separator WIM User s Manual 17 6 1 File Operations 255 200 180 111 100 90 8 122 8 100 100 234 199 212 231 123 Tab or comma separated CSV text files can be exported from many programs e g Excel The File Save as ASCII function in WIM also saves Byte images space separated in this format but ONLY if you set the Settings Value scaling to Pixel Value If Value Scaling is different from Pixel Value then the saved text file will probably have floating point numbers as text and be readable with the WIM function Read ASCII Float but not with Read ASCII as the text will have floating point numbers and not integers If the image is too small for viewing you can do multiple expansions of it with Transf Zoom ASCII Float Is similar to Read ASCII but reads a text file of pixel values into a Float32 4 bytes per pixel image buffer The pixel values may be either integers or float numbers without strict restrictions on their length As with Read ASCII the separators can be either a space a tab character a comma or a semico
140. products WIM sets the Value Scaling options according to the value scaling information in the file E g the water leaving radiances are scaled with Slope 0 001 and Intercept 0 0 chlor a is scaled with Slope 0 001 and Intercept 32 0 Please note that the pixel values are signed 2 byte integers The color scaling values for the Level 2 chlor a and CZCS pigment images are automatically set to 32000 and 29000 respectively see View Settings Color Scaling This corresponds to the concentration range of 0 3 mg m Anything above 3 mg m is white If your range of interest is higher than that you have to manually set a higher value to the Max value e g 22000 corresponding to the concentration of 10 mg m The projection of Level 2 unmapped products is automatically set to Swath and the parameter values are read from the HDF file This allows to create coastline overlays latitude longitude grids and other geo referenced images for the Level 2 files WIM allows to discard arrays narrower than the specified number of pixels View Settings HDF Options Minimal width of image to be read This prevents reading arrays of various calibration coefficients and other auxiliary information as images SeaWiFS Level 2 products include Level 2 flags Each of these flags indicates certain conditions The number of flags was increased from 16 to 32 in reprocessing number 3 spring 2000 Some of the flags are not used spare The values of the f
141. projection These files can also be read as generic raster images but then the user has to manually set the size scaling and projection HDF file Hierarchical Data Format HDF is a versatile data file format developed by the National Center for Supercomputing Applications at the University of WIM User s Manual 20 6 1 File Operations Illinois http hdf ncsa uiuc edu and adopted by NASA as the standard data exchange format for various data sets in their Earth Observing System EOS software HDF can store an almost unlimited number of data objects in a single file In the context of image analysis the basic object types are raster images 8 bit and 24 bit with the associated color lookup palettes for a raster 8 image and scientific data sets SDS i e multidimensional numerical arrays A multidimensional array in the SDS can contain 8 16 or 32 bit signed or unsigned integers or 32 or 64 bit floating point numbers int8 uint8 int16 uint16 int32 uint32 float32 float64 Ancillary information and non image data of the HDF file attributes can be examine with View Attributes When reading an HDF file WIM tries to find information on value scaling see View Settings General Value Scaling and projection If there is a palette in the HDF file the current lookup table is replaced with the data from the new palette and the palette flag is set to Custom see View Settings General Palette The las
142. r below which the corresponding vectors are ignored not plotted as they might be unreliable Grid Creates a copy of the current image file with the latitude and longitude grid lines in it The image has to have projection information in order to create a latitude longitude grid The numerical values of the grid lines are optionally written into the image A negative step for either the latitude or longitude grid disables the respective grid lines For example by using 10 as the latitude step and 1 as the longitude step produces only a latitude grid lines with 10 degree intervals Distance Calculates distance in kilometers and nautical miles along a selected line between the selected start and end points on the surface of the Earth The image must have a valid projection for distance calculation The points the distances and angles compass bearing between points can be saved in an ASCII text file that is similar to the trk track file used in Geo Transect The format of the file is always decimal degrees of Longitude and Latitude followed by distance in km and angles in degrees whereas on screen the user can select between 2 different formats of Latitude and Longitude The following example shows the screen layout Distance x m Start Point End Point Lat 36 560 Lon ji 23 124 Lat 36 526 Lon Aa 23101 Distance Compass Bearing 4 327 km 2 337 naut miles fi 73 01 o Display Format BESE AUG l
143. r visualization e g for overlays it is easier to use Byte 1bpp images instead of Int16 2 bpp images Therefore you may want to convert the Int16 depth image to a Byte image using some sort of scaling You can use Transf Convert and convert to Byte with Log scaling with a slope of 0 025 and intercept of 0 After that you can use View Set Colors and stretch the colors from 207 Start to 0 End Please note that Start is higher than end With most palettes that will create more intuitive colors You can then overlay the depth contours isobaths on top of the scaled depth image with Multi Overlay Image remember to select the image first and then select the overlay image number Get Bathy Overlay Reads an ASCII text file database file with coordinates of a certain depth isobath and creates an overlay file with FTP site In order to create a bathymetry overlay the image has to have projection information After creating the overlay you can merge it into the image by using Multi Overlay Image select the image first and then select the overlay image number The bathymetry contours are read from a dib file that has a sequence of Longitude and Latitude values of the contiguous line segments separated by an integer which states the number of records following that should be connected An example of a di 6 file 7 70 241949 42 650208 70 223761 42 663116 70 214960 42 715920 70 233148 42 735868 70 260137 42 728827 70
144. rder to reduce the calculation time The structure of the parameter file is the following 64 dx1 64 dy1 64 dx2 64 dy2 0 xlist WIM User s Manual 85 6 7 Multiple Image Operations 0 yist 64 dxlc 64 dylc 0 0 x2st y2st 10 xwin 10 ywin 8 maxmot 10 rangebeg 250 rangeend 5 overlap 24 0 timediff pixx pixy 0 6 thresh The numerical values must be first on the line the following names are just comments following after a whitespace Here dx7 dy and dx2 dy2 represent respectively the dimensions of the first and second image xfst yist and dxfc dyic respectively the start and dimensions of the area where correlation is to be done xwin ywin the correlation window sizes maxmot is the maximum displacement in pixels that is expected rangebeg and rangeend are the range of pixel values to be used overlap is the overlap between neighboring windows in pixels timediff is the time difference between the images in hours pixx and pixy are the pixel dimensions in km thresh is the cross correlation threshold below which the correlations are disregarded The apparent velocity vectors are saved in a file and can be overlaid on the current bitmap with File Vectors Note Do not forget to set the vector file format with View Settings Misc beforehand Warning As this operation is very compute intensive it is prohibitively slow on large images even on fast PC s Standalone
145. re several file formats for Point objects Header Xi Yi X2 Y2 X3 Y3 WIM User s Manual 65 6 5 Geo Operations Or Header lon lat lon lat lon lat Each point should be on a separate line ending with a CR LF sequence The longitudes and latitudes can be specified either as a single floating point value in degrees or as a degree value and a floating point minute value Please keep in mind to use negative numbers for both western longitudes and southern latitudes Please note that WIM uses Longitude before Latitude to be compatible with the sequence of x y and not vice versa Both 118 30 0 30 0 0 and 118 5 30 0 are valid lines specifying the same point of 30 0 0 N 118 0 0 W It is very convenient to have other information following the coordinates on the same line this information will be included in the output file For example you can use in situ values and additional information following the Lon Lat information The sequence of picked pixel values is shown as a profile The locations of the picked pixels are shown on the image bitmap as black or white rhomboids centered at the pixel or as connected lines if Line has been selected as Area of Interest in View Settings General You can store the picked points x y pixel value etc in a file In case of the Geographic Lon Lat option the Lon and Lat will be stored as floating point geographical coordinat
146. represented by one unsigned byte that can have values from O to 255 WIM can also handle images with pixel values of 1 bit 2 byte signed or unsigned integers 4 byte floats and integers Another type of raster object is called bitmap While an image is primarily a sequence of its pixel values a bitmap file follows a strict structure and includes all the information needed to display the visual color image When an intelligent program reads a bitmap it can immediately recognize it as a bitmap and has all the information to build the respective rendering on the screen In contrast when a program reads an image file it needs additional information e g the image dimensions colors corresponding to pixel values etc to build a picture That additional information is often stored in the image file together with the image data In the old days this additional information was often put into fixed headers while current formats such as HDF can store almost unlimited number of different attributes WIM uses headers when reading specialized image formats e g NOAA NESDIS CoastWatch Erdas Lan and it can be set to skip a predefined number of header bytes When reading plain unstructured raster image files WIM can make use of a small complementary info file that provides basic additional data describing the image size value scaling of pixel values geometric projection parameters If no info file is found WIM assumes the current or defaul
147. resulting image has the maximum pixel size of the operand images The images must be of the same size The Value Scaling of the resulting image is set equal to the Value Scaling of the current image Band Ratio Calculates various generalized band ratios for one or two images bands pixel wise Images are selected by their sequence number see List of Images x C One Image C Exponential Two Images Power Law Result 10 Coeffl Coeff2 Log Band1 B and2 Parameters Coeff1 0 393 Coeff2 0 872 Band 3 Band2 2 Valid Range Min fi Geophysical Values Max 253 Pixel Values OK Cancel WIM User s Manual 81 6 7 Multiple Image Operations E g for the Exponential case and one image selected the result is Result Coeff1 Bandl Coeff2 for two images selected the result is Result 10 Coeffl Bandl Band2 Coeff2 A common format is designated Power Law and is often used with a 2 band ratio in bio optical algorithms see O Reilly et al 1998 Result 10 Coeffl Coeff2 log Bandl Band2 For example the Morel 1 Chlorophyll algorithm O Reilly et al 1999 is Chla 104 0 2492 1 768 log Lwn443 Lwn555 where Band1 Lwn443 and Band2 Lwn555 Lwn normalized water leaving radiance the coefficients Coeff1 0 444 Coeff2 2 431 Band1 Lwn490 Band2 Lwn555 give the CalCOFI two band linear algorithm O Reilly et
148. ry file The size of the file will be the number of rows times the number of columns times the number of bytes per pixel The number of bytes per pixel in the saved image file will equal the number of bytes per pixel of the image buffer i e 1 2 or 4 The sequence of the pixels saved is from the top left corner row wise to the bottom right corner In case of a 2 byte per pixel image buffer the least significant byte is the first and the most significant byte the second Due to different floating point formats used on different computer platforms float images saved with WIM may not be directly readable on other platforms without conversion 2 Byte Int Writes the current image buffer either 1 or 2 bytes per pixel to a 2 byte per pixel file For each pixel the first byte is the least significant LSB and the second byte the most significant MSB byte If the current image buffer has 1 byte per pixel a zero byte will be added as the most significant byte to each pixel WIM User s Manual 27 6 1 File Operations Compressed RLE Saves the current 1 byte per pixel image with RLE run length encoding compression into a file RLE compression is not very effective for fine grained images However for images with large areas of constant values e g due to saturation by clouds errors etc it can save a considerable amount of disk space for storing large images RLE encoding is loss less i e after decoding the saved compres
149. s a 4 byte Float image that is usually not the best option but is the most versatile In most cases is convenient to scale the resulting image into a Byte or Int16 image using the Transf Convert function Turbidity chi ch2 Calculates the turbidity index from 2 visible band images according to the algorithm of Stumpf 1992 Remote sensing data is often contaminated by the interference of the atmosphere This function has been adapted from the program DECCON by Townsley and Stumpf and is intended to obtain an image of water reflectance corrected for some atmospheric effects As the water leaving reflectance is almost zero in the near infrared band 2 of AVHRR in all but the most turbid coastal waters it is possible to correct for some of the aerosol effects by subtracting a modified band 2 data from the modified band 1 data After the subtraction there remains a residual bias primarily from the Rayleigh radiance that can be removed by the user by subtracting a value from the entire scene so that the clear water in all scenes have the same value This operation is based on the assumption that the Rayleigh radiance is relatively stable over the scene For CoastWatch images all the necessary auxiliary data are retrieved from the image header The current image is assumed to be the channel 1 image and the user is prompted for the channel 2 image number If used for non CoastWatch images the user has to provide the necessary auxiliary data The
150. s from the scaled pixel values Real Value 10 Slope x Pix Value Intercept Obviously Linear scaling with Slope 0 1 and Intercept 0 0 corresponds to the predefined x 10 value scaling Likewise the Pigm mg m3 and Log Chl are special cases of the generic Logarithmic scaling While dragging the mouse with the right button down you can normally see the pixel values If various Value Scaling options have been selected you will also see the calculated values according to the scaling algorithm see also Transf Code Decode The pixel values used in several Examine operations Examine Color Table Examine Peeker Examine Profiles Examine Statistics as well as the values saved to a file Examine Line Save Examine Point Save File Save As ASCII will be converted to a floating point number depending on the Value Scaling option WIM User s Manual 42 6 3 View Options When performing operations with multiple image buffers e g Multi Add 2 Images Multi Composite Multi Difference Multi Divide w Image Multi Subtract Image the decoding to a floating point value is performed before the operation The Value Scaling of the resulting image buffer is set equal to the Value Scaling of the current image Bitmask is a specific form of scaling where each bit is representing certain preset flags or conditions It is used e g in the SeaWiFS and MODIS Level 2 data files Values of th
151. s generic band sequential images you have to manually sets Value scaling for the SST and SST anomaly images to SST Pathfinder projection to Global Equal Angle As the sequence of image data is different from the WIM convention Transf Mirror Horiz Axis has to be applied to prevent the images being upside down Unsigned Int Int gt Byte Read a plain unformatted raster image with each pixel represented by a 2 byte unsigned integer While Int gt Byte converts it immediately to a 1 byte WIM User s Manual 24 6 1 File Operations per pixel buffer Unsigned Int retains 2 bytes per pixel in the allocated image buffer An unsigned integer of 2 bytes 16 bits may have values from 0 to 65535 Depending on the convention either byte can be the most significant byte MSB Before loading the image from the file you have to specify if the MSB is the second default or first byte Before actually reading the image buffer WIM checks if the buffer contains certain header information corresponding to the CCAR navigate software Baldwin and Emery 1993 Emery 1995 If that is true then the user may either accept these parameters image dimensions and projection parameters or cancel and discard the header information For displaying an image of unsigned integer values i e from O to 65535 the pixel values are scaled linearly into 256 bins between Min value and Max value see View Settings General Color Scaling
152. se sample images were produced with the CCAR navigate program Baldwin and Emery 1993 Emery 1995 WIM User s Manual 91 6 8 Segmentation 6 8 Edge WIM User s Manual Shade Edge Finds edges according to the Cluster shade method of Peckinpaugh 1991 and Holyer and Peckinpaugh 1989 Contours Finds contours of pixels between low and high values zero crossing As Byte images are represented by unsigned bytes with values from 0 to 255 it finds crossings over 128 In order to find isolines corresponding to a certain pixel value you should first binarize the image with that value Transf Binarize Then use Contours with default values If the contours look too thick you can use Edge Thin Lines Prompts for the initial threshold Init Thr minimum threshold Min Thr and the maximum number of passes MaxPass If a pixel has a value that is at least Init Thr away from 128 and any of the 8 neighbors at least Init Thr away from 128 but on the other side the pixel is marked by 255 and considered an edge If MaxPass gt 0 additional passes are made attempting to extend the found edges Pixels that are not marked as edges but have edge pixels among one of their 8 neighbors are tested if they are crossing the 128 border with Min_Thr The operation is repeated until no more pixels are found or the maximum number of passes has been made Clean Edge Cleans an image of isolated small blobs of non zero pixels
153. sed as necessary Replace Values Opens a new image buffer of the same size and type bytes per pixel value scaling and replaces a range of pixel values with a new value If Value Scaling is set to Pixel Value then a range of pixel values is replaced by a new pixel value However if a different Value Scaling option is used see View Settings General Value Scaling WIM actually replaces values based on the current geophysical value and not the pixel value The replacement value that you enter is also treated as a coded value Only integer values can be entered If you want to replace values based on their pixel values set Value Scaling to Pixel Value WIM User s Manual 78 6 6 Image Transformations Solar Correction Does a simple correction for visible band images based on the calculated solar zenith angle and creates a new image buffer with the corrected values Using the geo conversion coefficients View Settings Projection or equivalently the longitude and latitude range to get the pixel geographical coordinates and the time of the satellite pass the solar zenith angle is computed for each pixel The sun zenith angle is calculated corresponding to the time of the image the Julian day and the GMT hour and minute If the Julian day is O it is calculated from the year last 2 digits e g 94 for 1994 month and day For a CoastWatch image all the necessary auxiliary data are retrieved from the image header
154. sed file you restore exactly the same image You can easily check it by saving an image as Compressed RLE reading it back with File Open Compressed RLE and calculating the difference Multi Difference You should get a black image consisting of plain zeros Check the maximum difference with View Set Colors as the Difference operation automatically includes View Lut Stretch Both the Start and End colors should be zero that prove that the maximum difference between the corresponding pixel values is zero Google Earth KML kml Google Earth KMZ kmz Writes the current image in the format compatible for viewing in Google Earth http Avww google com earth In the KML format the actual image in the PNG format is saved separately from the KML text file whereas in the KMZ format both the compressed image in the PNG format the KML file are embedded into a single KMZ file You can also view the current image directly in Google Earth with Geo View in Google Earth BS icon in the toolbar or in Google Maps with Geo View in Google Maps B icon in the toolbar Erdas Lan Writes a set of 1 byte per pixel image buffers to a file in the ERDAS LAN format ERDAS is a popular image processing program by Erdas Inc The ERDAS LAN format has a 128 byte header followed by image data in band interleaved by line BIL format The first 4 bytes of the header are HEAD HDF SDS Writes a set of image buffers as Scientifi
155. sentially a lookup table that matches a pixel value to a triplet of values for the Red Green and Blue components that are commonly used in computer color generation In 256 and higher color modes you can select between different palettes In 16 color mode not common these days you can normally use only the system palette The default palette before loading anything else is the built in Rainbow palette The Grayscale palette can be easily generated by the software All other palettes Custom have to be read from a file e g using the command File Load LUT The Rainbow palette has been adapted from Prentice 1987 with slight modifications It starts and ends with blue and shows a subjectively continuous color transformation WIM User s Manual 43 6 3 View Options You can set up WIM so that another default palette is read from a specified file on disk whenever WIM starts Use the Settings Misc Use Default LUT option for that By default WIM reads the default palette from a file chl lut in the Wimsoft folder This palette is good for many types of images and not just chlorophyll More palettes are included with the WIM package For example two tone red and blue palettes anomaly lut and anomaly7 lut are useful for showing anomalies Each image has its own palette unless it s the built in Rainbow type When an image or a set of images is saved to a HDF file then each image retains its individual palette as an HDF attribu
156. sually contain a large number of pixels ASCII files of integer or float numbers may become quite large approximately 4 times of the image file size If Line is currently selected Area of Interest see View Settings General Area of Interest the pixel values along the specified line are recorded but not the X and Y coordinates of the line cf Line Save Lat Lon Value ASCII dat Writes the image pixels as a series of triplets of Latitude Longitude Value in ASCII text characters If an area of interest Rectangle is selected then only pixels in that area of interest are saved You can choose whether to save the Latitude first followed by Longitude and the Value or Longitude first followed by Latitude and the Value depending on the Lat first checkbox in View Settings Misc Lat Lon Value Format The default formatting string is specified in View Settings Misc Lat Lon Value Format as 6 4f 6 4f 6 4f n and can be edited to suit your data The meaning of this string is to use 6 characters including 4 characters after the decimal point for each of the three values The n character means end of line and a tab character could be inserted as t Be careful when editing the string as WIM may become unstable with certain settings If the image has no projection information then the image coordinates X and Y will be used instead of the Longitude and Latitude If values from more than one image are saved th
157. t have to be spectral bands For example if you load images of Chl a and SST then the spectral plot would show the corresponding values of Chl a and SST for a selected pixel When six bands of normalized water leaving radiances nLw_412 nLw_443 nLw_488 nLw_531 nLw_551 nLw_667 are read from a MODIS Level 2 file you can have something like the plot below Note that the spectral shape shows a typical spectrum of water leaving radiance The minima in both the short and long wavelength parts are due to strong absorption by CDOM phytoplankton and water WIM User s Manual 59 6 4 Examine Operations Spectral Plot ai Righ click on this window drag mouse over the image Current point 1047 699 123 1698 35 7410 Value 4 5 Image Statistics Calculates a number of statistical variables for the currently selected rectangle transect or a selected area of the current image Statistics selected Transects or Rectangles read from a Track file trk can be calculated with Geo Get Vector Objects Statistics Min Max Mean StDev 5D Mean Median N in range N out range OK When calculating statistics it usually makes sense to exclude the pixel values represent no data or bad data To do that you can exclude the lower and upper range of pixel values from the calculations as values such as 0 or 255 are often used for indicating out of range values or for graphic overlays If the current image
158. t palette read replaces the previous Custom lookup table Due to the increasing popularity of HDF in storing and transferring satellite image data WIM includes many features for handling the special features of SSM I SeaWiFS MOS OCTS images and sea surface temperature data in various formats Some of these features are described in later sections of this document WIM can also save images with the accompanying auxiliary information in HDF files File Save as HDF that makes HDF the preferred file format when working with WIM In addition to images WIM can also store locations of Vector Objects such as points transects and rectangles but not text labels in a HDF file NOAA Level 1B L1B file NOAA Level 1B format is used by NOAA to store the Advanced Very High Resolution Radiometer AVHRR sensor data from NOAA s Polar Orbiting Environmental Satellites POES Current AVHRR sensors have 5 channels bands with bands 1 2 in the visible near infrared and bands 3 5 in thermal infrared NOAA Level 1B is a very complex format with a large number of modifications due to different sensors and options The basic types of Level 1B files are the GAC Global Area Coverage and the LAC HRPT versions The GAC data is produced by sampling the AVHRR 1 km data at 4 km reduced resolution The LAC Local Area Coverage Level 1B data are 1 km resolution data that are recorded on board the satellite and dumped at the receiving stations at a later
159. t significant 8 bits Truncate with Shift 1 discards the least significant bit equivalent to dividing by 2 and then uses the next 8 bits Truncate with Shift 2 discards 2 least significant bits and so on You have to choose a range between the lowest and highest value and the values in this range will be mapped to the pixel range of 0 255 The 2 byte option reads the image values directly into a 2 byte buffer without discarding any bits and the graphics overlay into a 1 byte pixel buffer The graphics overlay usually contains the coordinate grid and coastlines You can overlay it on the image with Multi Overlay Image If the match between the image and the overlay is not good you can use Multi Shift Image to shift the image relative to the overlay When reading a CoastWatch image WIM first reads the header and displays the header information Among others the projection parameters and the latitude and longitude range are read from the header Both Mercator and Linear projections are implemented right clicking or Examine Peeker shows the pixel coordinates You can check the Linear projection s latitude longitude coefficients with View Settings Projection or with View Attributes If the longitude latitude grid has been written into the image data by changing the pixel values to zero you can fill it in with Segm Fill Holes Compressed RLE Reads a run length encodes RLE 1 byte per pixel image format Correct image dimens
160. t values In order to read image formats containing an attached header other than CoastWatch and Erdas Lan it is possible to skip the header bytes from the beginning of the file A few programs e g CCAR navigate instead of appending an image to its header overwrite the beginning pixel values in the image with the header data An image i e a sequence of bytes representing the pixel values may reside both in a disk file or in a memory buffer image buffer In order to display an image from a file WIM allocates a memory buffer for it reads the pixel values from the file into the buffer converts the image buffer to a bitmap and then uses Windows functions to display the bitmap on the screen Both the image buffer and the corresponding bitmap occupy parts of the computer s RAM Up to 64 image buffers can reside in memory at the same time While reading images from disk files or generating new images itself WIM grabs additional memory from the computer s RAM for the image buffers and the corresponding bitmaps When RAM becomes scarce for loading new big images the user should free up RAM by deleting other image buffers from memory The names of all loaded images with their dimensions Width DX and Height DY Type and Size are listed in the List of Images dialog box WIM User s Manual 14 5 WIM Basics The following example shows 6 loaded images of 4 different types byte 2 byte integer int16 4 byte integer i
161. ta Small window size gives sharper edges but may be strongly influenced by the local noise Thinning Uses the edge value and direction images from the previous operation to create a new image with thinned edges This operation may be iterated several times to produce acceptable results Threshold Using the specified threshold value binarizes the image as 1 equal and larger than the threshold i e edge pixels and O less than the threshold This is the same operation as Transf Binarize that is simply duplicated here for the convenience of the user You have to experiment in order to find the best level of separating edges from noise Clean This operation eliminates transforms to zeros isolated edge pixels or small blobs of pixels It can be used as well with other binary images to eliminate WIM User s Manual 93 6 9 Segmentation disconnected pixels It is the same routine as Edge Edge Clean that is simply duplicated here for the convenience of the segmentation user Connect Allows to connect disconnected edge pixels It uses an image of edge pixels and looks for loose ends among edges edge pixels are different from zero and non edge pixels are zero It then reads tries to continue the loose ends along the gradient image actually perpendicularly to gradients You therefore need to have the edge gradient direction image as well in memory it was produced during edge finding Edge Find Prompts for t
162. te When reading an HDF file saved with WIM the user can choose whether to use the current default palette set in Settings Misc Default LUT and with the Settings HDF Options Override LUT in HDF checked or the palette from the HDF file unchecked Override LUT in HDF Settings General also shows the origin of the Custom palette obtained either from a palette file or from an image Confirm Settings If the Confirm Settings checkbox is checked WIM displays the View Settings dialog box each time when reading an image without a corresponding info file if disabled the dialog box is not shown When checked WIM prompts if projection parameters are to be taken from a CCAR image header Baldwin and Emery 1993 Emery 1995 otherwise they are accepted automatically if found Restore Defaults Clicking on this button deletes the WIM registry settings e g the WIM file extension settings window locations the locations of files used in the past parameter values used with various operations etc A typical case when you need this is if you accidentally associate a file extension with a wrong file type after double clicking on a file and selecting WIM as the application to open that file Extensions Has default image file extensions for all image file types You can change them to suit your file naming convention WIM always remembers the last file type that it loaded and suggest that as the default file type for a next load Th
163. ter b with Background Value of O and Foreground Value of 255 Click back to the original image and overlay the coastlines with Multi Overlay Image They should match perfectly You can now do File Save Info and that will save these settings for this file Next time when you load the same image you only have to do WIM User s Manual 113 New Generation SST Transf Mirror Horiz axis as all the other settings will be read from the corresponding inf file that you saved The geo location that was saved in the inf file is automatically applied only to the corresponding file and not to other NGSST files with the same projection However you can transfer the geo location info to all other images in WIM memory by selecting an image and then Geo Unify Geo coeff The information in the header 200 bytes is currently not interpreted by WIM However a WIM Automation Module WAM program wam convert ngsst converts all NGSST files to HDF files including all attributes that are essential when performing time series analysis WIM User s Manual 114 AMSR E from Remote Sensing Systems 19 AMSR E data from Remote Sensing Systems Various data products from AMSR E Advanced Microwave Scanning Radiometer are available from Remote Sensing Systems Inc http AWwww ssmi com amsr amsr_data_description html amsre data The gridded data files are available at 0 25 degree resolution in binary format These files can be dire
164. the top left corner The format is the following Header X11 Y11 X12 Y 2 X21 Y21 X22 Y22 Xn Yni Xn2 Vn2 Here x 1 and yii i 1 2 are the start coordinates and x 1 and yi2 are the end point coordinates of track The transect track files are assumed to have extension trk In case of the geo referenced option the longitudes and latitudes can either be specified in the long degrees and minutes or short floating point degrees format e g Track geo CalCOFI lines 87 and 93 123 35 6 29 50 9 117 18 8 32 57 5 line 93 123 45 3 31 19 8 118 30 2 33 53 2 line 87 or Track geo short form CalCOFI lines 87 and 93 123 593 29 848 117 313 32 958 123 755 31 330 118 503 33 887 The examples are from sample files calgeo trk and calgeosh trk respectively corresponding to a sample image calchi81 img The same tracks in video format from a sample file calvideo trk look like this Track video CalCOFI lines 87 and 93 234 359 324 315 232 338 307 302 If an image is loaded the pixels are picked along the track displayed on the screen as a profile and optionally stored in an ASCII text file with a format X coordinate Y coordinate Pixel value The Value Scaling settings in View Settings General are used to convert the pixel values to geophysical values If no image is in memory a clean image buffer with the current size is created and the track points are assign
165. this operation can be repeated several times in a row Fill Holes Replaces zero valued pixels with the mean value of neighboring non zero pixels Invalid pixels are most often assigned a value of O If the missing pixels have a value different from O e g 255 then they need to be converted to zero before Fill Holes can be used This function allows to fill small black holes due to small clouds image speckle or other noise Can be run iteratively to fill bigger areas Although there is no real substitute for missing data this function can be used to fill small areas of no data Be careful when using with floating point pixel values as very small positive or negative values are still different from O As an unwanted consequence the function expands the valid pixels into areas where there should not be valid pixels e g ocean values over land To fix the latter problem one can use Multi Mask w Image to put a proper land or another mask on top of the modified image Set Segments Uses the segmented image from Pixel trace and the original source image to produce two new images where pixel values are set respectively to the average and standard deviation for each segment If there is memory enough to open only one new memory buffer it skips the standard deviation and produces only the segment average image If you wish to exclude the edge pixels or any other set of pixels you can do that as the routine only counts not zero pixels when c
166. threshold It opens a new unscaled Byte image buffer and sets the yes pixels to a selected pixel value and the no pixels to zero Regardless of the source image 1 2 or 4 bytes per pixel this operation returns a 1 byte per pixel unscaled image This operation is useful for creating masks to be used in operations like Multi Mask w Image Starting from WIM version 6 11 the binarize function is generalized to include criteria involving more than one image Now you can set criteria on up to three images and include both lower and upper limits For example you may want to select all pixels that are lower than a threshold in image 1 above a threshold in image 2 and inside a range of values in image 3 The example below finds pixels where image 1 has values above a 60 and less than 255 1 image 2 has values below 201 and image 3 has values between 200 and 256 The new image will have unscaled pixel values 100 for pixels that satisfy the above criterion and zero for the other pixels The dialog will remain open and the user can interactively modify the criteria and get a new visual of the created image by clicking Create again Clicking Close closes the dialog Binarize ist image 9 Range eo lt 2nd image 0 Range 0 lt 3rd image 0 Range 200 lt Fill value unscaled 100 Close WIM User s Manual 74 6 6 Image Transformations Bitmask Opens a new 1 byte per pixel image buffer and pain
167. time The HRPT Level 1B data are 1 km direct readout data obtained as the satellite passes over receiving stations The Level 1B data normally include all 5 channels at 10 bit precision with time tags Earth location and calibration information It is also possible to unpack the 10 bit data into 16 or 8 bits The WIM adaptation of the Level 1B format is based on the Land Analysis System LAS AVHRR Data Acquisition and Processing System ADAPS of the US Geological Survey see http edcwww cr usgs gov programs sddm lasdist info index html WIM User s Manual 21 6 1 File Operations WIM includes functions for reading the Level 1B files Level 1B data can be ordered from the NOAA NESDIS Satellite Active Archive http Avwww saa noaa gov and from other satellite receiving stations In spite of being a standard format many varieties exist due to differences between sensors and file creation procedures Not all versions have been tested with WIM The default options of 5 bands at 10 bit sampling are supported Earth locating and calibration of the Level 1B data from different AVHRR sensors depends on a large number coefficients included in a number of tables These tables are included in a directory ADAPSTABLES that the user should copy manually from the WIM CD to your hard disk the default target location being C Program Files WimSoft ADAPSTABLES Before these tables become available to WIM the user has to create an environmental
168. ting when dealing with global or large scale images Therefore the most appropriate map file depends on the resolution and scale of your image Only the crude resolution file coast_crude b is included with the downloadable WimLE package Show GSHHS Polygons Background Value fo M Land T Lake Island in Lake Pond in Island in Lake OK Cancel In order to create a Map Overlay the image has to have projection information After creating the map overlay you can merge it into the image using Multi Overlay Image select the image first and then click on the Overlay icon on toolbar Both the background and foreground pixel values of the overlay images can be selected If the selected Foreground Value is lt 255 then the resulting image is a Byte image whereas if the Foreground Value is 5 255 then the resulting image is a Int16 image Please note that the maximum value of Int16 is 32767 and you probably don t want to select a bigger value for the Foreground as the value will be wrapped into a smaller value The background value should always be 0 if you want to overlay this image on top of another image A longer way of making the coastlines to have pixel values higher than 255 is first to create a regular Byte image with the coastlines of pixel value 255 then use Transf Convert to make it into an Int16 or a Float image and then Transf Linear Trans to multiply the pixel values by e g 255 Foreground Value 255
169. to 2 bpp and 4 bpp images the Current Settings Color Scaling Min and Max need to be suitably set You can try to use View LUT Stretch on each band before applying this function High color graphics modes i e 15 16 or higher are required for this function to work properly The RGB image can also be saved as HDPF In this form the HDF image can be read again as a RGB single Int32 image Finding the best color combination by shifting the Low and High scrollbars can be very slow for large images You can first find the best color combination on a small rectangle Just select a rectangle and select Examine RGB Image from the menu The RGB image is created quickly in the small selected rectangle You can now adjust the range values of the components After clicking OK the selected range values will be propagated to the whole image This is a very useful trick when working with large images WIM User s Manual 58 6 4 Examine Operations Another useful tip concerns making coastlines and other overlays A typical coastline overlay has pixel value 255 and that produces a blue coastline on a RGB image think of it as the first Blue component of a RGB image To create a white overlay you need to create a RGB image from the single overlay A special function Transf Convert to 24bpp RGB does exactly that A more difficult manual way of doing that is to select the overlay image and then Examine RGB Image Make sure that all 3 com
170. to use the best and latest system components there may be some backward compatibility problems running WIM on older Windows systems We therefore strongly recommend to use Windows XP or Windows 2000 because of their superior memory management Please note that in order to install WIM under Windows NT 2000 XP and later the user installing WIM needs to have the Administrator permissions As other Windows applications WIM uses the display and printer drivers that have been installed in Windows It is therefore relatively independent of the actual hardware assuming that it is compatible with Microsoft Windows Depending on your display facilities graphics adapter monitor and the active display driver the number of colors and the pixel resolution are different The now almost extinct standard VGA systems showed 16 different pure undithered colors at the resolution of 640 pixels times 480 lines You need at least 256 color graphics mode 8 bit color to use WIM while the higher bit depths 16 24 and 32 bits should be used if possible Most current graphics cards have 32 bit color capability The preferred screen resolution depends on the size of your monitor For viewing large images you should use the highest possible resolution at least 1024 x 768 WIM User s Manual Installing WIM 3 Installing WIM Installing WIM is simple You can run the setup file wimle msi evaluation version or wim msi WIM or wam msi WI
171. tor object It s primary purpose is to provide flexible annotation of the images Using the Add function in the Vector Objects table it is possible to write text labels anywhere in the image so that these labels can be edited or removed without affecting the image A related function Edit Draw Textis also available LUT Median Stretches a linear color look up table in the middle of the histogram leaving out the bottom and top distribution tails This operation is useful when you want to exclude a few very low and very high outliers that would otherwise make the LUT Stretch operation insensitive to most of the pixels in the middle of the histogram It is also useful if you want to focus the look up table on a certain area make a histogram of the area and run LUT Median with 0 exclusion You have to run Examine Histogram Calculate before using this WIM User s Manual 53 6 3 View Options routine Remember that the last histogram calculated is assumed when performing this operation LUT Stretch In case of a 1 byte per pixel image finds the minimum Start and maximum End pixel values on the current image and makes a new linearly stretched color look up table between them This function has another useful application finding the maximum and minimum values of a 1 byte per pixel image Use View Set Colors after it to check the minimum Start and maximum End values of the image In case of a 2 or 4 byte per pixel ima
172. tor projection orthogra hdf Orthographic projection polarste hdf Polar stereographic projection polyconi hdf Polyconic projection rectangu haf Rectangular projection stereogr hdf Stereographic projection utm hdf Universal Transverse Mercator projection ESRI Shapefiles USstates shp USstates dbf USstates shx ESRI shapefiles with US state boundaries WIM User s Manual 124 24 References 25 References Abbott M R and P M Zion 1985 Cont Shelf Res 4 661 680 Baldwin D G and Emery W J 1993 Ann Glaciol 17 414 420 Behrenfeld M J Falkowski P G 1997 Photosynthetic rates derived from satellite based chlorophyll concentration Limnology and Oceanography 42 1 20 Borgefors G 1983 Scandinavian Conference on Image Analysis 3rd pp 250 255 Cayula J F and P Cornillon 1992 J Atmos Oceanic Technol 9 67 80 Coll C V Caselles and J A Sobrino 1991 In lan Dowman ed SPATIAL DATA 2000 Christ Church Oxford 149 157 Diehl S F J W Budd D Ullman J F Cayula 2002 J Atmos Oceanic Technol 19 1105 1113 Emery W J et al 1991 J Geophys Res 96 C3 4751 4768 Emery W J 1995 personal communication An on line archive of NOAA AVHRR and other satellite data has been set up at http jester colorado edu EOSDIS html For details please contact Tim Kelley at kelley jester colorado edu The user can remotely run the CCAR navigate program and then download the produc
173. ts Finds filaments linear structures within the current image by using the Kasvand filter This operation is used e g for finding ice ridges on ice images Vesecky et al 1989 If ridges are defined as bright linear structures then we can find ridges with the following sequence of operations 1 Find the filaments 2 switch back to the original image 3 find the histogram 4 find the top 20 level Histogram Thresh 5 binarize the original image at the top 20 value only pixels at that value and above are marked as different from zero 6 switch to the filament image 7 mask logically AND with the binarized image Multi Mask w Image The final composite image is the image of bright AND linear structures Filter Performs filtering of the current image The selection of different filters is Median Sigma Mean Max Mean Pos The image is scanned with a small window and the center point of the window is replaced by the result of the respective operation e g mean sigma or median The borders of the image are directly copied from the source to the filtered image The sigma filter Lee 1986 is often used to suppress speckle noise in synthetic aperture radar images You have to choose the standard deviation of the speckle noise e g for a 3 look amplitude averaged image sigma 0 3017 Median filter replaces each small window with its median Mean filter with its mean Max filter with its maximum Mean Pos with the
174. ts areas where the selected bits are ON and other selected bits are OFF with the selected pixel value All others pixels are set to O The selected pixel Value should be from 1 to 255 The created mask images can then be used in operations like Multi Mask w Image Bits can be selected with the mouse multiple bits are selected with the mouse and the Control or Shift key pressed This operation is useful for creating masks corresponding to the values of certain flags e g for SeaWiFS and MODIS Level 2 images as shown in the following example Values of the bitmask flags can also be viewed directly on the source image by right clicking on the image the flags bits that are ON are shown in the window header Select ON bits k Select OFF bits 0 ATMFAIL A OK 0 ATMFAIL A ETHE a 7 Cancel Cancel 11 TURBIDW 12 HISOLZEN 13 HITAU 14 LOWLW 15 CHLFAIL 16 NAYWARN 16 NAVWARN First the user has to select the ON bits then the OFF bits For example by selecting OCEAN as the ON bit and TURBIDW as the OFF bit the user will create a mask image with non turbid ocean pixels having values 255 and all others with pixel values zero MODIS Level 2 products have all quality flags that are not individually marked but are in groups of two For example Bits 1 2 Quality for all of Gordon s nLw bands Bits 3 4 Quality for Carder s clear water epsilon band etc The user has to check the image attributes to see the corr
175. ty of HDF file format the need for info files has greatly diminished and these functions are not developed further Page Setup Allows to adjust the size of the image to be printed as well as to add a header and footer to the printout Header 51997252201 250 12 GAC a Footer wind 5 13 28Jan 1999 Margins Let fr fm Right p1 Bottom im im OK Cancel H Print Prints the current image bitmap to one of the installed printers See also Page Setup and Print Preview A Print Preview Shows the print preview of the current image Print Setup Invokes the Windows printer setup dialog box Exit Exits WIM WIM User s Manual 34 6 2 Edit Operations 6 2 Edit Edit Operations ZINI Lja ocjel ojm Outline pe Fill p Copy Copies the currently selected image bitmap area or the whole current bitmap if no area is selected to the Clipboard Clipboard is a common memory pool in Windows used for transferring data between applications Other Windows programs e g Microsoft Paint can then paste the copied area from the Clipboard Draw Allows to edit the image buffer with a versatile set of tools familiar from popular drawing programs In contrast to the drawing programs which can only edit bitmaps the Edit Draw functions actually edit the image buffer and then update the bitmap for viewing The set of tools includes the familiar tools
176. ual Angle and appropriate Value scaling Both the browse and the 9 km resolution full files can be viewed The Value Scaling used for both chlor_a and CZCS pigment data is Log Chl You can then use Geo Get Map Overlay to generate coastline and other overlays in high or low resolution Among the mapped SeaWiFS images other than the Global Equal Angle only the SeaDAS Cylindrical projection with the parameters Latitude Center 0 Rotation 0 Position 0 O 1 1 Isotropic 0 Scale 0 can be navigated in WIM This SeaDAS projection is automatically converted to the WIM Linear projection with correct parameters A wealth of information on SeaWiFS images is included as global or local attributes you can view this information with View Attributes WIM User s Manual 99 OCTS Products 9 OCTS Products OCTS standard products are delivered in HDF and are very similar to the respective SeaWiFS products In Level 3 mapped chlor a images both the image and the palette are delivered as scientific data sets SDS Select Images X Select SDS buffers to read from 019970613bmoccm hdt 13bm chlor a 4096 2048 uint8 palette chlor a 256 3 uint8 Cancel You can skip the palette as it is only 3 pixels high When reading Level 3 mapped images Value Scaling for both chlor_a and CZCS pigment data is set to Log Chl For global images in Standard Mapped Image format set the Projection type to Global Equal Angle Glob eq angle in V
177. urface temperature algorithm MCSST are being generated by the US Naval Oceanographic Office NAVOCEANO and distributed at a near real time WIM User s Manual 23 6 1 File Operations rate by NASA s Physical Oceanography Distributed Active Archive Center PO DAAC The individual Level 2 NAVOCEANO format datasets def are not images but Latitude Longitude dataset derived from the GAC Global Area Coverage data of individual orbits These datasets are converted to images in the Global Equal Angle projection by WIM When reading a single Level 2 NAVOCEANO dataset WIM creates a single global image The default image size is 720 x 360 pixels but this can be changed in the Settings Misc Default Navoceano Size As individual datasets provide a relatively narrow swath of data along the cloud free portion of the satellite track these datasets are usually composited over many passes in order to get good global coverage When selecting multiple Level 2 NAVOCEANO images simultaneously WIM creates 2 images a composited global image of average SST and another image with the number of pixels used in the composite This operation is analogous to Multi Composite It may be tricky initially to select multiple images with WIM In order to do that you should highlight one image e g by right clicking on it and then selecting a range of images with Shift left click or selecting individual images with Control left click After a set of imag
178. variable ADPAPSTABLES and set it to the location of the directory e g to C Program Files WimSoft ADAPSTABLES On Windows 2000 XP systems environmental variables can be set by right clicking on My Computer and selecting Properties Advanced Environment Variables New Please note that the tables must be in the Unix text mode i e line ends with a single LF character and not with a CR character Updated tables can be downloaded from ftp edcftp cr usgs gov orders g1k Please download the tables as binary and not as ASCII text AVHRR channels 1 2 are calibrated into percent reflectance values from 0 to 100 and channels 3 5 into temperature in degrees C In case of errors a common case being that a required table is missing or not accessible from the ADAPSTABLES directory a log is shown The full log of processing each Level 1B file can be viewed by using the Attribute icon on the toolbar or menu View Attributes By right clicking on the image the X Y and the Longitude and Latitude of each pixel are shown along with the pixel value and geophysical value percent reflectance or temperature in degrees C Coastlines and other overlies can be created by using Geo Get Map Overlay Be warned that this operation is very slow for Level 1B images The generated overlies can be inserted into the image by using Multi Overlay Image Sometimes the geo location has a significant error that can be manually corrected with Multi Shift Image Lat
179. versions of the program see Emery et al 1991 should be used for more extensive calculations Multiply Multiplies the current image to another image pixel wise The resulting image is always a float image 4 bytes per pixel The Value Scaling options see View Settings General Value Scaling are used to calculate the real pixel value before the multiplication i Overlay Image Overlays another image on the current image Pixels assume values of the current image if the corresponding pixel on the second image is zero and values of the second image if the corresponding pixel on the second image is not zero Prompts for the overlay image index If the index number is non negative a new image buffer is allocated and the result is put there If the index number is negative overlaying will be done in place and the absolute WIM User s Manual 86 6 7 Multiple Image Operations number is used as the overlay index This operation could be used for example to overlay coastlines tracks etc on the current image As an example load File Open Image a sample image est256 img then load the overlay image File Open Overlay est256 0vi switch back to est256 img and then run Overlay Image by giving the sequence number of the est256 ov image buffer Polarization Ratio Performs the TBH TBV TBH TBV operation on pixel values of 2 images TBH and TBV respectively and saves it in a new image buffer The
180. video format x pixel column number y line number counting from the upper left corner or in the geo referenced format longitudes and latitudes The geo referenced coordinates can be in the short format decimal Longitude and decimal Latitude or in the long format int Lon degree float Lon minute int Lat degree float Lat minute Please note that the Point csv or pnt file format can be either Lon Lat Value or Lat Lon Value i e starting with either Latitude or Longitude and is specified in the Settings Misc Lat Lon Value format options Another special format for point data is the NOAA drifter file format DAT see Drifter track below which is always in the Lat first format Point You can select the symbol used to mark each point in the image The default symbol Standard is a filled circle An extensible list of available symbols is stored in the file PointTypes xml and new symbols can easily be created by editing that file For example a simple rectangle pointing up is defined by the following series of zeros and ones lt name gt Triangle lt name gt lt data gt lt row gt 0001000 lt row gt lt row gt 0010100 lt row gt lt row gt 0100010 lt row gt lt row gt 1111111 lt row gt lt data gt The file PointTypes xml must be in the same directory with the WIM executable typically in C Program Files WimSoft Selecting different symbols for different data sets the user can distinguish There a
181. w make the RGB composites of those and then read sub areas of the WIM User s Manual 26 6 1 File Operations scene at full resolution A set of individual bands can be conveniently saved in one file with File Save As Erdas Lan Copy Image Makes a copy of the current image buffer into a new image buffer under the same name Cut Image Cuts the specified rectangle of the current image into a new image buffer that is then displayed You will see the dimensions of the selected area continuously as you drag the mouse If no rectangle has been specified by mouse dragging it prompts for the upper left and lower right corner coordinates The coordinates are bound between 0 the minimum X and Y at the top left corner and the size of the current image Due to a requirement of the MS Windows memory management it is recommended that the width X of the selected rectangular area is a multiple of 4 If your final X dimension is not a multiple of 4 WIM will increase it or if it reaches the right bound of the image decrease it If the current image has geo coefficients different from zero they are transformed for the new image named lt Cutlmg gt If the original image is in Mercator or other geographic projection the lt Cutlmg gt will be in a recalculated Linear projection E save as Writes the current image buffer to a file You can select between a number of file types Image Saves a plain unformatted bina
182. w from the other extensions with a semicolon e g img dat nec raw After selecting the file you get to another screen of the Current Settings with an option to set Image Size Set Width 1000 Height 1000 Set the Value Scaling in the same screen to SST PATHF C WIM User s Manual 112 New Generation SST After loading the image you can see that it is upside down To fix that use Transf Mirror Horiz axis Now the image should look OK but it has no geo location To set the geo location open Settings Projection and set the WIM native projections to Linear the Longitude coefficients to 116 and 0 05 and the Latitude coefficients to 63 and 0 05 see below Current Settings x General Extensions Projection Special gt WIM native projections Unmapped Albers C Swath Linear Equidist Lat Lon arrays Mercator C n3a C Ingest LIB PolStereo n3b C Glob Mercator C TV Mere C 33a C Glob eq angle C Conic C s3b Shift po C ModisLand HDF EOS C GCTP Terascan Projections Sensor C Polar Stereo Polyconic C Stereo C Equidist Azim Lambert Azim Rectangular Mercator Lambert Conic Orthographic C UTM C Albers Conic Longitude coefficients Latitude coefficients ft 16 00000 f0 050000 63 000000 fo 050000 OK Cancel Apply Help Now you can verify that you have the correct geo location by creating coastlines with Geo Get Map Overlay coast in
183. w is adjusted so that it would not get outside the image limits If any of the longitude latitude coefficients View Settings Projection is different from zero the longitude and latitude values of the center pixel are also shown The values displayed for each pixel depend on the Value Scaling parameter chosen for the current image see View Settings General Value Scaling By pressing and holding down the right button of the mouse it is possible to see the values continuously updated while you move the mouse around the image Note You can do this if you start dragging from the Peeker window and move over to the image window without releasing the right button If you start on the image window you will get the normal current pixel position X Y and value on the window top frame You can even use more than one Peeker window by opening a new one from the menu while keeping the previous one However do not start too many Peekers as you may finally crash the program In case of a float 4 bytes per pixel image buffer only 3 pixel values are shown per line View Zoom can also be used for detailed examination of a small image area Point Save Allows you to save pixel coordinates and values to a file Clicking with the Right button of the mouse on an image pixel normally shows the pixel coordinates and value on the window s top frame If Point Save has been selected the coordinates and values are also recorded in an ASCII file each
184. y For example CNES day 19750 corresponds to January 28 2004 WIM converts the CNES day to attributes Start Year and End Year WIM User s Manual 109 Institute Maurice Lamontagne SST 16 SST data from the Institute Maurice Lamontagne Sea surface temperature data for various regions of Canada is provided by the Department of Fisheries and Oceans Quebec The data is processed by the Institute Maurice Lamontagne and is available from ftp calypso qc dfo mpo gc ca period msst hdf The data is in the format of SDS in HDF4 and as such is readable by WIM However some adjustments were made to WIM in order make the use of these datasets easier Namely the color scaling Min and Max are automatically set to 1 8 and 27 97 respectively when reading those Float32 datasets see View Settings When using WAM utilities e g wam match and wam statist WAM will recognize the period start and period end attributes and convert those to the respective attributes that are needed to identify the time period of the dataset The time period of a dataset is needed when doing match ups with in situ data or when building time series of satellite data Geo referencing information projection is automatically read from the HDF file attributes and operations like Geo Get Map Overlay work without problems However in order to better visualize the data it is advised to convert the Float32 datasets to Byte with SST PATHF scaling with Trans
185. y attributes in the image file itself CoastWatch HDF If the info file exists its contents will be overwritten otherwise a new file will be created The name of the info file is formed from the current image name with the extension of inf Please observe that during image operations image buffers in memory get different WIM User s Manual 32 6 1 File Operations names some of these are not acceptable as DOS Windows filenames An error message will appear if you try to save a corresponding info file If you want to save the info file for one of these buffers first rename the buffer or write the image buffer to a file File Save As reload that file File Open and then save the info file The simplest format of an info file is the following Xsize ysize In case of Linear rectangular latitude longitude projection the first line has it s geo conversion coefficients Xsize ysize lon top left dlon lat top left lat e g the US West coast CZCS image calch 81 img has in its info file 512 512 140 0 0 07 55 0 0 07 Pigment The top left corner has coordinates 140 W negative and 55 N The pixel increments of longitude and latitude are 0 07 degree pixel Please note the that longitude parameters precede the latitude parameters On the next line the current pixel value scaling will be saved see View Settings Value Scaling x10 x100 Pigment In case of projections other than the linear

Download Pdf Manuals

Related Search

Related Contents

20326A_CycleOps Carbon Wheels IS  Manual de Instrucciones  SPSV Initial Suitability Inspection Manual for vehicles seeking first  Bedienungsanleitung Instruction manual  Istruzioni per l`uso    

Copyright © All rights reserved. Failed to retrieve file