Home

HydroLight Technical Documentation

1. 440 540 640 L L r 2 5 15 1 99809 300 12 14 16 18 L 20 data H20abDefaults_S 1 C H dumm C H C H dumm dumm dumm C H Dumm FAwater txt E52files data examples ac9 data txt yFilteredAc9 txt E52files data examples Hydroscat6 withH20 txt E52files data examples Chlzdata txt yCDOMdata txt yR bot ydata txt E52files data examples Chlzdata txt yirrad txt data MyBiolumData txt 102 APPENDIX B DESCRIPTION OF ARRAYS AND I O UNIT NUMBERS Within HE52 there are arrays and reserved I O file numbers that advanced users should be aware of before modifying existing or writing new HE52 routines Arrays imisc and fmisc are used to store miscellaneous integer and floating point variables respectively that are used throughout the code The array datafiles stores the names of data files expected for certain components The elements of these arrays as well as a listing of the file unit numbers assigned for HE52 are given here B 1 The IMISC Array Common block array imisc holds integer values as follows some values such as nphi and nL are not used by EcoLight and some values are not used in the HE52 standard routines imisc variable name description index the number of phi bands in phi 0 to 2 pi a nz the number of depths where output is saved including closely spaced pairs for K functions ncomp number of components in IOP model in ab routine 5 8
2. nwave isource flag for presence of internal sources or inelastic scattering 1 present 0 not present TATT AT idbug flag for additional printout for debugging most such code was removed in version 4 0 10 dimension for spectral amplitude arrays 11 jwave index of current wavelength band 12 ibotm flag for bottom type finite depth or infinitely deep flag for sweep direction in Riccati integration 13 isweep 14 current Fourier L value 103 number of initial rays started in surfwind ray tracing code 17 numrays 18 nuscr1 Fortran unit number for scratch file 1 Fortran unit number for scratch file 2 20 iwrtss1 flag for writing single wavelength spreadsheet data files written iff 1 19 nuscr2 lit 22 iwrtdf Digital output file option file written iff 1 23 nconc the number of component concentrations needed this is only relevant if CDOM or fluorescence is present 24 iOptRad flag for writing full radiance printout file Lroot txt files written iff 1 26 NwsSkip number of wavelengths to skip is skip options is selected 27 iDynZ flags if dynamic depth solutions are allowed when sources are included and bottom is ininitely deep B 2 The FMISC Array Common block array fmisc holds floating point values as follows fmisc variable name description pi 3 141952654 degrad 0 0174533 converts degrees to radians radeg 57 2958 converts radians to degrees ref
3. The default HydroLight quad layout shown in Fig 29a partitions the set of all directions into 10 deg polar angle 0 by 15 deg azimuthal angle The corresponding EcoLight band layout in Fig 29b has 10 deg polar angle bands Each partition has polar caps with a 5 deg half angle HE52 computes the radiances directionally averaged over these quads and bands This angular resolution is a balance between the conflicting needs of having sufficiently high angular resolution in the radiance distribution and keeping the run times acceptably small Numerical studies show that a finer angular resolution does not change computed quantities such as irradiances or AOPs by more than roughly one percent whereas accuracy starts to degrade for a coarser resolution see HydroLight Technical Note 2 However the run time is proportional the square of the number of quads or bands so finer angular resolution comes at a high computational cost with almost no improvement in numerical accuracy of the quantities of interest to most users We therefore strongly recommend against altering the default quad and band resolutions which are adequate for almost every problem of practical interest 7 ra S N N SS AX Eee Se S S Fig 29a The standard HydroLight quad Fig 29b The standard EcoLight band partitioning of all directions partitioning of all directions 72 Previous versions of HydroLight included the code needed for defining n
4. the standard code reads 300 1000 1 To enable HE to run from say 250 to 1500 nm change this line to 250 1500 1 and save the file It may also be necessary to increase the default maximum number of wavelengths on the UI change limits form Figure 23 shows the total direct and diffuse spectral plane irradiances incident onto the sea surface as computed by the MODTRAN version 4 0 Acharya et al 1998 atmospheric radiative transfer code and by RADTRANX for roughly the same input atmospheric 46 16 4 Modtran Ed_total m 14 Modtran Ed_direct E Modtran Ed_diffuse N 12 4 Radtran Ed_total gt 1 0 Radtran Ed_direct m 0 8 4 Radtran Ed_diffuse o amp 0 6 E 0 4 0 2 0 0 350 450 550 650 750 850 950 1050 1150 1250 1350 1450 wavelength nm Fig 23 Comparison of MODTRAN and RADTRANX sea level radiances for roughly corresponding atmospheric conditions Note that the total irradiances are close but that the component direct and diffuse irradiances are much different for the two models conditions MODTRAN was run from 350 to 1500 nm by 10 nm RADTRANX can be run only to 1000 nm One of the standard MODTRAN outputs is a file 1x of computed irradiances at all levels in the atmosphere The IDL program Modtran Ed to Hydrolight_Ed pro reads MODTRAN f1x files extracts the sea level irradiances and saves them on a HSF forma
5. 8 1 Filenames that are not needed for the run components for which ibbopt 0 may be stored with the name pfdummy dpf as a place holder Names of variables pfname 1 to pfname ncomp Example pureh2o dpf avgpart dpf othgg90 dpf RECORD GROUP 6 Wavelengths These records define the wavelengths to be used in the run The first record gives nwave the number of wavelength BANDS at which the model is being run Name of variable nwave Example 5 The format of the next record is determined by the value of nwave if nwave 0 the run is to be made at just one wavelength In this case the next record gives Names of variables wavel areset breset Example 532 0 0 1 0 25 where wavel is the wavelength in nm and areset and breset are the values of a and b to be used in the abconst IOP model if it is being used For all other IOP models areset and breset have values of 1 0 The sky spectral radiance at wavelength wavel will be used 1 nm resolution in RADTRANX 94 if nwave gt 1 the run is to be made with one or more finite width wavelength bands gt 1 nm band width with 1 nm resolution in the sky irradiances In this case the next record gives Names of variables waveb 1 waveb 2 waveb nwave 1 Example 400 0 420 0 430 0 440 0 450 0 475 0 where the values of waveb j give the nwave 1 WAVELENGTH BAND BOUNDARIES in nm for which the model is to be run More than one line can be used if nee
6. HYDROLIGHT HYDROLIGHT 5 2 ECOLIGHT 5 2 TECHNICAL DOCUMENTATION Curtis D Mobley Lydia K Sundman Sequoia Scientific Inc sEMUOIA First Printing October 2013 Update Note This version of the the HydroLight EcoLight Technical Documentation updates the previous version of August 2012 for HydroLight EcoLight version 5 1 to cover changes up to and including code versions 5 2 0 A pdf of this document with color figures is in the HE52 Documents directory Copyright HydroLight and EcoLight versions 5 2 are copyright 2013 by Curtis D Mobley All rights reserved worldwide This Technical Documentation is protected by United States copyright law No part of this document may be copied or distributed transmitted transcribed stored in a retrieval system or translated into any human or computer language in any form or by any means electronic mechanical magnetic manual or otherwise except by prior written permission of Curtis D Mobley Licensees of HydroLight and EcoLight are given limited permission to reproduce this document provided such copies are for their use only and are not sold or distributed to third parties All such copies must contain the title page and this notice page in their entirety Trademarks Microsoft MS DOS Visual Basic Excel and Windows are registered trademarks of Microsoft Corporation Lahey is a registered trademark of Lahey Computer Systems Inc IDL is a registered trademark of Exelis Vis
7. j Common block array datafiles holds the filenames described in Record Group 12 of the lroot txt file as follows array variable name contents of file in Hydrolight standard format index 0 pureH20abc pure water a b and c values to be used 1 abac9data user supplied UNFILTERED ac data ac 9 or similar for IOPdata model abac9Filtered user supplied FILTERED ac data ac 9 or similar for IOPdata model hscatdata user supplied b data bb 9 or similar for IOPdata model 2 3 4 chizdatafile user supplied chlorophyll vs depth profile 5 CDOM data user supplied CDOM absorption profile 6 Rbottomdata user supplied bottom reflectance vs wavelength data 7 and datacomp i for i 7 to concentration data corresponding to each component on ncomp 6 105 B 4 Reserved I O Unit Numbers Whenever possible files are opened read or written closed and appended if need be within a single call to a subroutine The user is encouraged to follow this example There are a few exceptions where the unit must remain connected throughout the run To prevent any potential conflict the unit numbers are assigned as follows HE52 basic IO scratch file 1 scratch file 2 Iroot standard input 3 4 5 6 Standard output to screen for run progress report etc 7 surfwind file 8 gcirrad file sky model data file 9 pureH2o pure water datafile HE52 Output and special project Input 10 Proot txt output file 11 Droot txt outpu
8. which do not represent the likely physical behavior of the concentration data and in some cases especially near sharp gradients may make the spline defined concentration go negative It must be remembered that an interpolation scheme e g cubic splines is guaranteed to fit the data exactly at the measured points but the values generated in between the measured values can be almost anything depending on the interpolation scheme and the smoothness of the data 61 255 30 f L fi 1 I I 0 10 0 20 0 30 0 40 0 56 0 10 0 20 0 30 0 40 0 50 concentration concentration Fig Fig 27 Hypothetical discrete depth vs concentration data symbols as fit by linear interpolation red lines and cubic spline interpolation green lines In the left panel the concentration varies smoothly with depth In the right panel there is a strong gradient in concentration near a depth of 15 and the cubic spline overshoots above and below this depth Because of the potential difficulties with cubic spline interpolation HE52 by default uses linear interpolation for all data Cubic spline interpolation is still available as an option for use by users who understand the smoothness requirements on the data to be fit The cubic spline code is commented out in the HES2 source code but can be reconstituted if desired Linear interpolation never overshoots However the numerical algorithms on rare occasions can have problems
9. 42 Figure 22 The Harrison and Coombes clear sky radiance angular pattern for the sun at a 30 deg zenith angle solid lines the sun is at the upper left of the figure Values are relative to one near the sun The red dotted lines show the default HydroLight quads HydroLight uses the direct sky irradiance in setting the magnitude of the quad averaged sky radiance for the quad containing the sun The relative sky radiance pattern is integrated over all quads and the diffuse sky irradiance is used to rescale this result and thereby set the magnitude of the quad averaged diffuse sky radiances in all quads This is why HE52 must partition the total incident irradiance into diffuse and direct parts The resulting quad averaged sun and sky radiance distribution then reproduces the direct and diffuse irradiances computed by RADTRAN X or as read in from a file of measured irradiances The irradiance and radiance routines just described which are used when the SEMI ANALYTICAL SKY MODEL option is selected in the UI do a sufficiently good job of modeling the sky radiance for most applications of HES2 The use of this sky option is therefore recommended for use in general oceanographic studies If your application depends critically on the sky radiance then you should use a measured sky radiance distribution as input to HE52 or you should use a sophisticated atmospheric model such as MODTRAN Acharya et al 1998 to generate the sky radiance input
10. CLASSIC CASE 1 This model is based on a recent reformulation Morel and Maritorena 2001 of the Case 1 water model seen in Light and Water Eqs 3 27 and 3 40 The absorption coefficient is modeled as the sum of three components Arid aA aA aza 1 where a is absorption by pure water a is absorption by chlorophyll bearing particles and a is absorption by covarying yellow matter CDOM The particle absorption is given by a ZA 0 06 A nA ICAS 2 where Chl z is the user supplied chlorophyll concentration profile in mg Chl m and 4 A is the non dimensional chlorophyll specific absorption coefficient given in Prieur and Sathyendranath 1981 their a and Morel 1988 his Fig 10c as extrapolated to 300 and 1000 nm This 4 A is shown in Fig 2 It should be noted that this 4 A is independent of the chlorophyll concentration Thus only the magnitude of the particle absorption coefficient depends on Chi the shape of the a spectrum is the same for all Chl values Our extrapolation of the original A A to 300 and 1000 nm is somewhat uncertain especially at UV wavelengths see 2 4 1 below We therefore recommend the new Case 1 IOP model for general use especially near 300 nm The classic Case 1 IOP model is retained in HE52 for ease of comparison with the new Case 1 IOP model 1 05 T T T O E solid is measured i a z OBE dashed is extrapolated 4 E U OF J L x
11. Edit the file HE52 SpecialRuns Input Input txt which has two or more records that are read by the discretization program The first record gives a descriptive title which is copied to the first record of the discretized phase function file The second record is the root name used to generate the name of the file containing the discretized phase function An example of these two records is Measured pf from Station 4 550 nm standard quad partition Sta4_ 550 If your phase function routine requires additional input you can add additional data records after line two of the Input txt file and write your phase function routine to read the needed data see file pfff f for an example of reading a third line of input The root name Sta4_550 in the above example will be used to generate the name of the file containing the discretized phase function this would be Sta4_550 dpf in the present example The file extension dpf identifies the file as containing a discretized phase function The dpf extension is only a convenient reminder of what is in the file HE52 does not actually require this naming convention for phase function files Edit the Input txt file to change the title and file name records as desired and re save the file 76 Step 5 Run the script HE52 SpecialRuns PhaseFunction discpf bat This can be done by double clicking on the file name in Windows Explorer or entering the command discpf bat in a command window T
12. of the Users Guide that a small increment will be added to each nominal output depth for the purpose of computing depth derivatives for K functions RECORD GROUP 12 Data Files In the last group of input the names of all of the remaining data files are specified in the following order PureWaterDataFile nac9Files 3 ac9DataFile 4 Ac9FilteredDataFile 5 HydroScatDataFile 6 ChilzDataFile 7 CDOMDataFile 8 RbottomFile Pure water data file e g H2OabDefaults txt The number of ac9 files to be read either 1 or 2 Standard format file with unfiltered a and c data e g a file of ac9 data Used only with the IOPdata model Standard format file with filtered a and c data which may be used when CDOM fluorescence is included Standard format data file of backscatter data e g HydroScat or bb 9 Used only with IOPdata model Standard format chlorophyll profile data file read by routine chlzdata Standard format data file containing values of CDOM absorption at a given reference wavelength specified above Standard format bottom reflectance data file read by routine rbottom 9 TxtDataFile i for i 1 to ncomp Concentration profile data files for component i ncomp lines of input 100 10 IrradDataFile Standard format data file containing sea surface total sun sky values to be used instead of RADTRAN X values 11 SObiolumFile Standard format data file containing biolumines
13. the a can be entered in the file as A and E can be set to 1 The Bricaud model of Eq 6 then reduces to the classic IOP phytoplankton absorption model with the form a ZA a Chile Regardless of which set of A and E spectra is chosen the A and E spectra are used in the same manner in Eq 6 to define a A for any Chl value Figure 6 shows the resulting particle absorption spectra for low medium and high UV absorptions and for Chl 0 01 0 1 1 0 and 10 0 mg m The corresponding absorption coefficients as computed by the classic Case 1 IOP model are shown for comparison as is absorption by pure water There are significant differences in the classic and new models which will lead to significant differences in computed radiances irradiances and AOPs when used in HE52 Note in particular that the shape of the particle absorption spectrum now changes with the chlorophyll value Presumably the new model gives a more realistic description on average of particle absorption in Case 1 waters than does the classic model of Eq 2 13 Chl 0 01 Chi 0 1 0 003 0 020 T E 0 015 0 002 a 0 010F 3 8 0 001 a 0 005 L a H o 0 000 i i J 0 000 i 300 400 500 600 700 300 400 500 600 700 Chl 10 0 1 0 r r ie 7 0 8 7 0 08 J 0 6 J E 0 4 a 0 2 J a NING 300 400 500 600 700 300 400 500 600 700 wavelength nm wavelength nm Fig 6 Part
14. 0 1 et et eae ome ET a a oC T1111 rror a er a E A 500 600 wavelength nm Sy Er WE Er To TO rE 0 ee Fig 4 Comparison of the Vasilkov et al 2005 A A and B A purple curves with those of Bricaud et al 1998 green curves Solid lines are A and dashed lines are B 11 Desperation is the mother of modeling you can quote us on that so to define A A and B A spectra over the 300 1000 nm range we proceeded as follows The Bricaud et al A and B curves are accepted for use from 400 to 720 nm with A B 0 between 720 and 1000 nm The Vasilkov et al curve for A was normalized to the Bricaud value at 400 nm i e 4 A A A A 400 4 400 where subscripts v and b stand for Vasilkov et al and Bricaud et al respectively The normalized 4 A was then averaged with the two normalized spectra of Morrison and Nelson seen in Fig 1 assuming that the A spectra have the same shape as a This assumption about the shapes of A and a is correct only if B 0 or if Chl 1 in which case A a in Eq 6 The resulting average A between 300 and 400 nm then merges smoothly with the A of Bricaud at 400 nm For B the Vasilkov et al curve was normalized to the Bricaud et al curve at 400 and the result was used to extend the Bricaud et al B down to 300 nm The resulting A and B spectra are shown in red in Fig 5 along with the three A spectra used to compute the average A between 300 and 400 These A and B
15. 10 mg Chl m We hope that this new IOP model for Case 1 water is an improvement over the classic model However the limitations of any IOP model must be remembered in particular see Mobley et al 2004 for limitations of the Case 1 concept IOP models may be very good on average but may or may not be very often definitely are not correct in any particular instance IOP models are best used for generic studies When modeling a particular water body especially in a closure study it is always best to use measured data to the greatest extent possible If for example measured particle absorption spectra are available they can be used in HE52 by dividing the measured a by Chl to get a which is then entered as A in a user defined AE file with E 1 as noted above The corresponding Chl value can then be entered in the HE52 UI for use in reconstituting the measured a and for modeling the scattering if not measured as just described 20 2 5 CASE 2 This model is a generic four component pure water chlorophyll bearing particles CDOM and mineral particles IOP model The UI allows users to specify individual component IOPs and concentration profiles in a variety of ways Thus the case 2 IOP model is suitable for many Case 2 waters The various ways of specifying component absorption and scattering properties are similar to what has just been described in the Case 1 IOP models and are best seen by running the UI This
16. 9 ac s bb 9 HydroScat 6 or other data HE52 usually assumes that your data are perfect and it gives you the corresponding solution of the radiative transfer equation recall Fig 1 of the Users Guide One exception to this is that if HES2 detects a negative absorption coefficient in your data file a common occurrence due to instrument noise at near infrared wavelengths where phytoplankton absorption is essentially zero it will print a warning message to the printout file and reset the negative absorption to zero and then continue running HE52 will often continue running even with obviously bad input It is not idiot proof and does not do much checking of user input data Second if you are using ac 9 data to define the IOPs and you also wish to include chlorophyll fluorescence for example the UI will inform you that chlorophyll is not a part of your IOP model and ask you to input information on the chlorophyll profile But you say The reason I bought an ac 9 was to measure the IOPs directly rather than having to measure the chlorophyll profile and then use a bio optical model to get the IOPs That s fine but you still need to know how much chlorophyll is in the water before you can compute the chlorophyll fluorescence There is an additional caveat when using WETLabs ac s data The HydroLight Standard Format for files of a and c data assumes that both a and c are measured at the same wavelengths This is true for the ac 9 instrume
17. E J PETA a 9o 0 6F q o 9 z S bf Oo Cf E 2 6 F C E Oo 4 O46 E o ig Z 6 E on E E a E 5 5 0 2 E t a E G E A 2 o0 EELEE LEENE EEEE ee ee EEEE 300 400 500 600 700 800 900 1000 wavelength A nm Fig 2 The non dimensional chlorophyll specific absorption coefficient A A used in Eq 2 Absorption by yellow matter colored dissolved organic matter or CDOM covaries with particle absorption according to a ZA 0 2 a 2 440 nm exp 0 014 A 440 3 The scattering coefficient for the particles is given by b EA b cnor 259 4 pZ F N gt 4 where the defaults are b 0 3 n 0 62 and m 1 Gordon and Morel 1983 or L amp W Eq 3 40 The yellow matter is assumed to be non scattering The user must specify both the chlorophyll concentration Chl z and the scattering phase function for the particles One way to specify the particle phase function is to give the particle backscatter fraction B b b from which HE52 generates a Fournier Forand phase function with the requested value of B Mobley et al 2002 Various formulas for B as a function of chlorophyll can be found in the literature For example Ulloa et al 1994 give an empirical formula for B at 550 nm in Case waters B 0 01 0 78 0 42 log ChI 5 and Twardowski et al 2001 present another formula B 0 0096 Chi ee Such formulas for B can be used for rough guidance in specifying t
18. Each of these represents the transfer of light from one wavelength to another longer one In the case of Raman scattering the way in which light is redistributed in wavelength is well understood For purposes of computational efficiency HE52 uses the azimuthally averaged formulation of Raman scattering described in Mobley et al 1993 Appendix A which gives the correct contribution of Raman scattered light to irradiances This formulation gives some inaccuracy in the Raman contribution to an azimuthally asymmetric radiance distribution in HydroLight The band averaged radiances computed by EcoLight are exact Although Raman scatter is well understood and depends only on the properties of pure water the interpretation of the detailed spectral shape of the Raman emission as measured in the ocean is complicated by the wavelength dependence of the absorption coefficient between the emission depth and the measurement depth See HydroLight Technical Note 10 for a discussion and examples In the case of chlorophyll and CDOM fluorescence HE52 uses the computed scalar irradiance the component absorption and various assumptions about the fluorescence efficiency and the wavelength redistribution function to calculate the amount of light fluoresced For chlorophyll fluorescence the default fluorescence efficiency of 0 02 can be changed in the CHANGE DEFAULTS form of the UI The default CDOM fluorescence quantum efficiency function is take from Hawe
19. Sky Irradiance Data There are three HydroLight Standard Formats for input of user defined sky irradiance data depending on which option is chosen in the UI as discussed in Section 5 4 The first option allows for input of the total sky irradiance data which is partitioned into direct and diffuse parts by RADTRANX The HSF data file has the format partial output of file HE52 Data Examples Sky_Irrad_Example_Editotal txt Total sun sky sea level irradiance obtained from MODTRAN output file C HE52 examples IDL MODTRAN example f1lx This file is for use as HydroLight input Record 4 unused Record 5 unused Record 6 unused Record 7 unused Record 8 unused wavelength Ed_total nm W m 2 nm 350 0 6 7579e 001 360 0 7 1849e 001 3 70 20 7 9674e 001 many wavelengths omitted 1480 0 3 8328e 002 1490 0 8 2752e 002 1500 0 1 4052e 001 1 0 1 0000e 000 58 The second option allows for input of the direct and diffuse sky irradiance components The HSF data file has the format partial output of file HE52 Data Examples Sky_Irrad_Example_Eddir_Eddif txt Direct sun and diffuse sky sea level irradiances obtained from MODTRAN output file C HE5S2 development Modtran_Ed MODTRAN example f1lx This file is for use as HydroLight input Record 4 unused Record 5 unused Record 6 unused Record 7 unused Record 8 unused wavelength Ed direct Ed diffuse nm W m 2 nm W m 2 nm 350 0 2 2731e 00
20. User Interface UL Table 1 The IOP models for absorption a and scattering b provided with HE52 The source code is in the HE52 code common directory with filename prefixed with ab e g the new Case 1 model is in routine abnewCase1 found in file HE52 code common abnewCase1_ f 2 1 PURE WATER This IOP routine returns the absorption coefficient a and scattering coefficient b for pure water as selected by the user in the UI The options available in the UI have a values based on Pope and Fry 1997 and scattering for either pure fresh water or pure sea water or have absorption from Smith and Baker 1981 and scattering by sea water The Pope and Fry and Smith and Baker absorptions are both appended by other data sets as needed for UV and IR wavelengths A user created data file of a and b values also can be read The Pope and Fry data are read from file HE52 data H20abDefaults_ SEAwater txt or HE52 data H20abDefaults_ FRESHwater txt the Smith and Baker data are read from file HE52 data H20abClearNat txt The header records in these two data files give the references for the data sets used to extend the IOPs beyond the wavelengths of the original data sets Figure 1 shows these spectra The scattering coefficients for pure sea water are computed from Morel 1974 c f L amp W page 103 b 16 06 1 80 107 500 0 A with a similar formula for fresh water The scattering phase function BAD for both pure sea water or pure f
21. a O eA EEE alae tate eens EE glen E a ata dre E 3 2 oll Pure Water et SoG Sieh Bg SG oh ata Aa GS Geile Bde Ba habe ad Ge Belg Dtag Saale aye 6 22 Co sta t ohn Oo he ee atte pe pte eee aie E eat ld ps ia nat a aro Glassie Casel sgena n Saas Vas eee Be Gr ak Saas aa eS ao Hee eS 7 24 New Casel o2vcuscudpivetedticeaad oebedes eee aa a o a eiA 10 2 4 1 Particle Absorption 224 3 a sadied on os SEES Bie BERS RE RSH aaa 10 DAD CDOM Absorpti n 400 shop ghg 09 ba dee 40 40a 4 aR te ee eae 14 DAO SSCAICNING s 3024 ote AOR Orc aie ie sale iY Ar oe Ae Ne NN tla 16 2 4 4 Scattering Phase Function 9 4 scatei vanes Be ewa eea a4 de Baad 18 DISCOS 2 oe xx ode Said Goes kee Blood cgided Gude came dal Ine ooid ears heed 21 2 6 Measured IOP Daas octe oases ce yh atic opi PING an 29 he ho PP a a 21 2 7 User written IOP Models cia Aa cect iad Svat tay Saath ease aenee na aas 25 2 8 Specific Absorption Models 025 202 enunk danetaweter eaae Pav eiamoees 25 2 9 Specie scattering Models ree oto aed sce Pals ee ea td ee alt Oe ey 28 BOTTOM REFLECTANCE MODELS 2 0 0 0 ccc ccc eee e eens 30 Sal Emite d pth Water riir eee ae wlan ace led eee aes date nee 30 3 2 Infinitely Deep Water Without Inelastic Scattering or Intemal S urees Sa s 5 ia Way a Sa a eu eee ea news 31 3 3 Infinitely Deep Water with Inelastic Scattering or Internal Sources Dynamic Bottom Boundary Depths 0 0000s 33 SEASSURFACE MODEL 0 436 sheep 0 09 ps awind bod
22. a climatology of monthly averaged O concentrations for 5 degree latitude by 10 deg longitude bins This climatology was derived from 5 years 2000 2004 of TOMS Total Ozone Mapping Spectrometer version 8 satellite data For polar regions with missing TOMS data zonal averages from the nearest latitude band with valid data are used This database is used when the user specifies the HE52 run for a particular latitude longitude and time and specifies the use of climatological ozone values by setting the ozone concentration to 99 in the UI As before a particular ozone value can be specified in Dobson units in the UI which will override the climatological value It should be noted that the ozone concentration affects the incident irradiance only below about 340 nm Although the difference in at the sea surface for very low and very high O3 concentrations can be an order of magnitude at 300 nm there is no effect on AOPs Figure 18 shows an example of the TOMS monthly average climatological ozone values as available in HE52 and the actual values for a particular day The monthly averaged values vary over the course of a year and with location from less than 150 to more than 450 Dobson units The default used in HES2 is 300 Dobson units 38 ozone Dabsons lt 150 TOMS Ozone Concentration for September OMI Total Ozone Sep 23 2007 NIVR FMI NASA ENMI MI Dobson Units Dark Gray lt 100 and gt 500 DU ash 2b aaz a
23. can be used as is These models also can be the starting point for creating models tailored to a particular user s specific situation This section briefly describes the IOP models available in the HES2 software package All IOP routines have the same fundamental structure see HE52 examples templates ab f which allows the same subroutine to be used by both HydroLight and EcoLight The source codes for these and other subroutines that are common to both HydroLight and EcoLight are in the HE52 code common directory The UI prompts the user for the various inputs needed by these routines The IOP routines allow users to model the IOPs of almost any water body For example the user can e select models for Case 1 water which are based on the user supplied chlorophyll concentration These models convert the user supplied depth dependent chlorophyll concentration to depth and wavelength dependent absorption a and scattering b coefficients e build up the a and b values for Case 2 water by defining the concentrations and specific absorption and scattering coefficients of non covarying microbial dissolved and mineral components e obtain depth and wavelength dependent a and b values from user supplied data files containing measured a and c values e g data from a WETLabs ac 9 or similar instrument c is the beam attenuation coefficient and b c a Depth and wavelength dependent scattering phase functions can be determined from associate
24. directory name where EcoLight looks for phase function files e Change the mxwork array dimension parameter in routine HE52 code EcoLight infbotm f from parameter mxwork 1000 to parameter mxwork 3200 e The first time you run EcoLight via the UI change the CHANGE LIMITS button on the RUN IDENTIFICATION form to increase the Maximum number of theta quad bands from 20 to 92 When EcoLight runs with these changes the code will recompile with the new array dimensions and EcoLight will find the 2 deg surface and phase function files when it looks under the default directory names When you are done with the high resolution runs be sure to reverse the above changes if you did not create a separate directory such as HE522deg for the 2 deg runs That is change HE52 data surfaces EcoLight which now contains 2 deg files back to HE52 data surfaces EcoLight 2deg and change HE52 data surfaces EcoLight_10deg back to HE52 data surfaces EcoLight and so on Change the maximum number of theta bands back to 20 for use with the standard band partition Note that the 2 deg grid is available only for EcoLight runs Correspondingly high resolution files for HydroLight are not yet at least available The main reasons are 1 the very small solid angles of the HydroLight quads would require extremely large numbers of Monte Carlo photon tracings to get surface files with negligible statistical noise especially at high wind speeds a
25. for a 50 per cent cloud cover with both averaged over 10 nm wavelength bands The Kasten and Czeplak adjustment for cloud fraction is very simple and cannot account for differences such as a thin cloud cover vs part of the sky covered by small cumulus clouds although both situations might correspond to a particular value of Cld in their model Their model is nevertheless better than nothing in estimating the effects of clouds on incident irradiances If very accurate irradiance calculations are to be made especially for cloudy conditions then the UI has an option for specifying a user created file of measured total irradiances rather than using values computed by RADTRAN X If this is done HE52 still uses RADTRAN X to partition the measured total into direct and diffuse parts as is needed within HES2 40 1 5 roe Torre eas err rer pores roars Por m sun at 45 O 300 total direct 7 iff 1 0 H diffuse M E he J A Ph W l 2 0 5 i 2 MERN l O a Mi 0 0 ee i piiiiiil boii sri 300 400 500 600 700 800 900 1000 wavelength nm Fig 19 Example sea surface irradiances as computed by RADTRAN X rrr TTT TTT TTT T H sun at 45 E O 150 a 04E 0 300 7 E E 450 0 36 J F i L 5 02 O F al Z i 0 1F 7 2 I O05 e cotta DESEO ETSE 310 320 330 340 350 wavelength nm 300 Fig 20 Effects of ozone concentratio
26. for generation of lroot txt files This is sometimes done when making many runs or when running HE52 on operating systems othr than MicroSoft Windows for which there is no UI The following pages therefore describe in detail the input records found on file lroot txt There are 12 records or groups of records each containing a particular type of information Each of the records is free format The names of the variables are those used in HE52 code commoniinitial f which reads Iroot txt additional documentation is given in initial f RECORD 1 Default Parameters This record contains Name of variable icompile Parmin Parmax PhiChl Raman0 RamanXS iDynZ RamanExp Example 0 400 700 0 02 488 0 00026 1 5 5 icompile is the flag that tells the RUN utility whether to use the STANDARD icompile 0 no compiling necessary or USER icompile 1 code must be re compiled executable A root for file will be generated and placed in code batch if and only if icompile 1 PARmin is the lowest wavelength included in PAR calculations 87 PARmax is the highest wavelength included in PAR calculations PhiChl is the chlorophyll fluorescence efficiency Raman0 is the Raman reference wavelength in L amp W Eq 5 89 RamanXS is the Raman scattering coefficient at the reference wavelength a in L amp W Eq 5 89 iDynZ is the flag that tells HE52 whether to use the Dynamic depth option if an infinitely deep bottom boundary is selected when ine
27. give an absorption model that roughly corresponds to the mid range of UV absorptions seen in the Morrison and Nelson data The new Case 1 IOP model uses these A and E 1 B as the default spectra for the last version of Eq 6 The tabulated A and E spectra are on file data AE_midUVabs txt 0 4 F mA pager oe TIT TT ae 0 3 F J o f 5 0 2 F E lt Ea 0 1 J hc Basie secetei E ea 300 400 500 600 700 wavelength nm Fig 5 The A red solid line and B dashed line spectra used in Eq 6 to define a for the mid range of UV absorption The purple and blue curves were averaged to produce A between 300 and 400 nm 12 Some HE52 users will likely be interested in comparisons of underwater light fields for the wide range of UV absorptions illustrated by the Morrison and Nelson spectra of Fig 3 We therefore defined A spectra for low and high UV absorptions by simply using the shapes of the Morrison and Nelson spectra absorption spectra to extend the Bricaud A from 400 down to 300 nm The B spectra were taken to be the same as for the mid range of UV absorption just discussed The user can select low file AE_lowUVabs txt mid or high UV file AE_highUVabs txt absorptions in the HE52 UI Users who have their own data for A and B can place their data in an ASCII file of the same format as AE_midUVabs txt and select the new file in the UI Jf the user has a measured a spectrum and wishes to use it in Eq 6
28. incorrectness of its output 8 In no event shall Curtis Mobley and or Sequoia Scientific Inc be liable for costs of procurement of substitute products or consequential damages however caused and on any theory of liability arising out of or related to this Agreement even if Curtis Mobley and or Sequoia Scientific Inc have been advised of the possibility of such damages 109 9 The User agrees to defend indemnify and hold Curtis Mobley and Sequoia Scientific Inc harmless from any loss cost expense claim liability demand or cause of action arising in any way from the User s use of HE52 or any products based on HE52 10 In the event that the User desires to develop a product system or service based in whole or in part on HES2 or which incorporates any portion of HES2 the User will not manufacture sell or otherwise commercially exploit such a resultant product system or service before obtaining a written agreement from Curtis Mobley granting such rights which may be granted by Curtis Mobley at his sole discretion 11 If you licensed HE52 in the United States this agreement is governed by the laws of the State of Washington 12 This license will terminate automatically if the User fails to comply with the conditions and limitations described herein On termination the User must uninstall HE52 and destroy all copies of the software and documentation 13 It is requested but not required that the use of HE52 b
29. phase function at each wavelength The only ways to have wavelength dependent phase functions are to use the power law formula for the backscatter fraction recall 2 3 or use the IOP data model to read in wavelength dependent scattering and backscatter coefficients recall 2 6 In either case a wavelength dependent Fourier Forand phase function will be generated 9 3 High theta resolution EcoLight Runs The standard HydroLight quad partition and EcoLight band partition of direction are shown in Fig 29 These directional bins are adequate for almost all purposes As shown in HydroLight Technical Note 2 outputs such as irradiances and reflectances are the same to within a percent or two if higher angular resolutions are used but the run time increases dramatically as the number of quads or bands increases However there are some simulations for which a higher angular resolution may be needed For example if the irradiance or remote sensing reflectance etc is being computed as a function of time over the course of a day a plot of irradiance vs time may show a stairstep shape rather than a smooth curve This is because the sun s irradiance is smeared out over a quad or band to computed the corresponding radiance using the solid angle of the quad or band even though the solar irradiance value is computed for the exact solar zenith angle Thus when the sun passes from one quad or band to another the smeared out sun jumps f
30. runs Or 10L actual A a aq L L a D 20 T 30L 0 0 5 1 0 IOP value Figure 16 Hypothetical actual depth profile of an IOP black line and the profiles used 10 red and 20 m green The Z a 20 m profile is offset by 0 5 to the right for clarity the z 10 m profile is offset by 1 0 when z max max 32 3 3 Infinitely Deep Water with Inelastic Scattering or Internal Sources Dynamic Bottom Boundary Depths For technical radiative transfer reasons well in all honesty because Curt can t figure out the math the water column below depth Z a does not include inelastic scattering or bioluminescence even if those effects are included between the surface and Z a In other words inelastic scattering and bioluminescence are always turned off at the lower boundary This can cause abrupt changes in diffuse attenuation functions near Z and other apparently strange behavior as the light field near z adjusts from one including inelastic max scatter above z to one without it below z These behaviors are illustrated and discussed in detail in HydroLight Technical Note 6 which is available on the HE52 documents directory or the HydroLight Users Group website To avoid problems or confusion with the application of a source free bottom BRDF in runs involving internal sources and infinitely deep water HES2 by default applies the bottom boundary condition at a w
31. shows the remote sensing reflectance R A for homogeneous infinitely deep waters with Ch 0 01 0 1 1 and 10 mg m as computed for the classic and new Case 1 IOP models The mid range UV absorption model was used in the new model The sun was placed at a zenith angle of 30 deg in a clear sky with typical marine atmospheric parameters sky irradiances were computed using the RADTRAN X sky irradiance model discussed in 5 1 The wind speed was 6 m s For the classic IOP model the particle phase function was taken to be a Fournier Forand phase function with a backscatter fraction as given by Eq 5 EcoLight was run from 300 to 800 nm with 10 nm bands We see that the R spectra are very similar for Chl 0 01 and 0 1 but that the differences can become very large at high chlorophyll values The maximum difference computed as 100 new old old is less than 20 for Chl 0 01 or 0 1 For Chl 1 the maximum difference is less than 50 at visible wavelengths 58 at 795 nm For Ch 10 the differences are as large as 243 more than a factor of three at 575 nm The larger R for high Chl is due to the greatly increased scattering as seen in Fig 8 19 0 025 0 020 0 045 Rrs 1 sr 0 010 0 005 0 000 300 400 500 600 700 800 wavelength nm Fig 10 Comparison of R as computed by the New Case 1 model with mid range UV absorption N dotted lines and the Classic model C solid lines for Ch 0 01 0 1 1 and
32. the atmospheric information entered in the user interface UI will not be used Keep in mind that a simulation of lidar input in HES2 corresponds to a horizontally infinite lidar footprint on the sea surface not a small diameter lidar beam which gives an inherenty 3D radiative transfer problem Also HTS cannot simulate time dependent effects such as beam spreading Input a file containing wavelength total irradiance and the fraction of the total that is direct irradiance FileSky_Irrad_Example_Edtot_frac txtis an example The direct irradiance will be computed from the total times the fraction and the diffuse from the total times 1 minus the fraction With this option the sun zenith angle and sky radiance models are used as for option 2 and RADTRANX will not be called 5 4 1 Running HE52 outside the 300 1000 nm range The RADTRANX sky irradiance model is the default for computing direct and diffuse sky irradiances in HE52 The databases underlying RADTRANX cover only the 300 1000 nm range which in turns limits HE52 to those wavelengths However because Option 2 does not call RADTRANX this option makes it possible to run HE52 over any wavelengths There are however important caveats to running HE outside of 300 1000 nm 45 Even though the user can input sun and sky irradiances for any wavelengths and run HES2 the IOPs and bottom reflectances will be given the values at the nearest wavelengths defined in their databases F
33. to compute the irradiances For idealized single wavelength runs Computes radiance patterns for uniform or heavy overcast skies by L amp W Eq 4 50 For idealized single wavelength runs Table 2 The sky irradiance and radiance models distributed with HE52 The user interface option column shows which routines are called when either of the sky model options is selected when running the UI 37 5 1 The RADTRAN X Sky Irradiance Model A major part of the work involved in extending the HE52 databases to the 300 1000 nm range required extending the original RADTRAN sky irradiance model of Gregg and Carder 1990 which is the default used to compute the clear sky direct sun and diffuse background sky irradiances incident onto the sea surface Dr Marcos Montes of the U S Naval Research Laboratory kindly computed and provided the needed O O and H O absorption coefficients on the format needed to extend the RADTRAN model used in H4 The RADTRAN numerical model with the extended underlying database as used in HE52 is called RADTRAN X At wavelengths near 300 nm the solar irradiance reaching the sea surface is strongly influenced by ozone absorption It is therefore important that the user provides HE52 with the correct O concentration if accurate radiometric quantities are to be computed below 340 nm AOPs will of course be insensitive to the magnitude of the incident sky irradiance To assist with this HE52 includes
34. 0 5 86099E 30 1 51607E 27 2 51447E 25 2 67395E 23 1 82323E 21 7 97094E 20 2 23438E 18 4 01593E 17 4 62802E 16 6 50 4 33072E 29 1 12023E 26 1 85795E 24 1 97580E 22 1 34719E 20 5 88976E 19 1 65100E 17 2 96739E 16 3 41967E 15 7 00 3 19997E 28 8 27739E 26 1 37284E 23 1 45992E 21 9 95443E 20 4 35196E 18 1 21992E 16 2 19261E 15 2 52680E 14 7 50 2 36439E 27 6 11597E 25 1 01436E 22 1 07870E 20 7 35510E 19 3 21556E 17 9 01374E 16 1 62007E 14 1 86699E 13 8 00 1 74655E 26 4 51782E 24 7 49300E 22 7 96826E 20 5 43315E 18 2 37531E 16 6 65837E 15 1 19673E 13 1 37913E 12 8 50 1 28778E 25 3 33110E 23 5 52479E 21 5 87521E 19 4 00600E 17 1 75137E 15 4 90939E 14 8 82380E 13 1 01687E 11 9 00 9 36746E 25 2 42309E 22 4 01861E 20 4 27371E 16 2 91402E 16 1 27397E 14 3 57116E 13 6 41856E 12 7 39684E 11 9 50 6 20826E 24 1 60589E 21 2 66345E 19 2 83238E 17 1 93126E 15 8 44322E 14 2 36677E 12 4 25388E 11 4 90223E 10 7 8 Filling in Missing Data When working with measured data the data will be available only at discrete depths and wavelengths However as HE52 solves the RTE it must know the absorption and scattering coefficients at all depths and wavelengths relevant to the run and it must know the bottom reflectance at any wavelength for which the RTE is to be solved Therefore some sort of interpolation or extrapolation must be done to define concentrations IOPs and bottom reflectances at depths and wavelengths other that those actually measured HE52 does this in two ways depe
35. 01000 300 400 500 600 700 800 9001000 Chl 1 0 Chl 10 0 OG Brrr 30 reer R E 7 O 5P E eS 9 O 3F ii 4 1SE au 3 Thee F a A 0 2 a ee Si ee coe E 0 5 J 0 0 Breeder 0 0 h E a 300 400 500 600 700 800 9001000 300 400 500 600 700 800 9001000 wavelength nm wavelength nm Fig 8 Beam attenuation c as determined by from Eq 10 and near surface values of c 0 407 and n 0 795 blue The red curves are the same a as the mid UV absorptions in Fig 6 The solid green curve is the new b c a the dashed green curve is the classic b of Eq 4 17 2 4 4 Scattering phase function When using the classic Case 1 IOP model the user must also specify the scattering phase function to be used for particle scattering This can be done in various ways e g by picking a particular phase function such as the Petzold average particle phase function defined in Mobley et al 1993 or by specifying the particle backscatter fraction B b b which is then used to generate a Fourier Forand phase function with that backscatter fraction Mobley et al 2002 Morel et al 2002 developed a phase function model for Case 1 water in which the phase function is a combination of small and large particle phase functions with the fraction of each being determined by the chlorophyll concentration BCh a Chl R Chl By 11 where a ChI 0 855 0 5 0 25 log Chi and a 1 a Fig
36. 1 4 4848e 001 360 0 2 5410e 001 4 6440e 001 370 0 2 9374e 001 5 0300e 001 many wavelengths omitted 1490 0 5 0811e 002 3 1941e 002 1500 0 8 4977e 002 5 5539e 002 1 0 1 0000e 000 1 0000e 000 The third option allows for input of the total sky irradiance and the direct fraction The HSF data file has the format partial output of file HE52 Data Examples Sky_Irrad_Example_Eddir_Eddif txt Total sun sky and fraction of direct sea level irradiance obtained from MODTRAN output file C HE52 examples IDL MODTRAN example f1lx This file is for use as HydroLight input frac direct Ed_direct Ed total Record 5 unused j Record 6 unused Record 7 unused Record 8 unused wavelength Ed total frac_direct nm W m 2 nm nondimen 350 0 6 7579e 001 0 33637 360 0 7 1849e 001 0 35365 370 0 7 9674e 001 0 36868 many wavelengths omitted 1490 0 8 2752e 002 0 61402 1500 0 1 4052e 001 0 60475 1 0 1 0000e 000 1 00000 59 The format for input of one exact wavelength as needed for lidar simulation was shown in Section 5 4 2 7 7 Bioluminescence Data There is also a HydroLight Standard Format for reading in the strength of a bioluminescing layer as a function of depth and wavelength The quantity on this file is the spectral source strength S z A in W m nm as seen in Light and Water Eq 5 107 When modeling bioluminesence in HE52 the bioluminescent source represents a horizontally homogeneous la
37. 215 224 Pope R M and E S Fry 1997 AO 36 33 8710 8723 Prieur and Sathyendranath 1981 L amp O 26 4 671 689 Quan and Fry 1995 AO 34 18 3477 3480 Smith R C and K Baker 1981 AO 20 2 177 184 Spinrad R W K L Carder and M J Perry 1994 Ocean Optics Oxford University Press New York 283 pages Twardowski et al 2001 JGR 106 14129 14142 Ulloa Sathyendranath and Platt 1994 AO 33 3 7070 7077 Vasilkov et al 2005 AO 44 14 2863 2869 Voss 1992 L amp O 37 501 509 86 APPENDIX A DETAILED DESCRIPTION OF THE RUN TIME INPUT FOR STANDARD RUNS The various subroutines and data files used by HE52 provide much of the information needed to make a standard run The remaining information is read in at run time from the file Iroot txt One such file is automatically generated by the UI for each run and its exact format is not of interest to most users The same Iroot txt file is used by both HydroLight and EcoLight Increases in options and flexibility of HE52 runs has resulted in the input files becoming increasingly complicated to the point that users are now strongly discouraged from trying to work directly with the input file However some users may wish to edit these files to make small changes in the input from one run to the next for instance changing the wavelengths or output depths rather than re running the UI This will certainly be the case for users who wish to write their own programs
38. 88 14 716 0 0947061 0 0652625 0 0642977 0 290185 0 253677 0 272243 14 894 0 0949196 0 0643945 0 0638672 0 288154 0 255801 0 274805 1 0 0 0949196 0 0643945 0 0638672 0 288154 0 255801 0 274805 This example was extracted from file HE52 data examples ac3data txt see also HE52 examples template ac9data txt for an example file with 9 wavelengths Note that the file has exactly 10 header records for identification Record 11 contains the number of wavelengths and the nominal wavelengths Records 12 and onward contain the depth a and c data with one depth per record The required units for a and c are inverse meters The last record is just the last valid record duplicated but with the depth set to 1 0 HE52 will stop reading the file when a record with a negative depth is read 55 When HE52 reads a file of ac 9 data it performs a number of calculations The a and c values are read for the discrete depths and wavelengths Minor error checking is performed For example if the depths as read are not in increasing order the depth records are re ordered values with duplicate depths are averaged However such quality control should be done before giving the data to HE32 b c a is computed for each depth and wavelength The discrete depth discrete wavelength a and b values are fit with linear splines See 7 6 for discussion of the spline fitting On all subsequent calls the splines are used to define a and b at any depth and wavelengt
39. 9s allows the total absorption to be separated into particulate and dissolved fractions which gives important information about the ecosystem However it must be understood that light is influenced only by the total absorption coefficient the total scattering coefficient and the total phase function If CDOM is assumed to be non scattering partitioning the absorption into particulate and dissolved fractions does not change the total a total b or total phase function Therefore partitioning the absorption into particulate and dissolved fractions does not change the elastic scattering solution of the radiative transfer equation Thus in the absence of fluorescence the light field is exactly the same whether or not the absorption is partitioned into particulate and dissolved components Oceanographers may care about absorption by particles vs absorption by CDOM but photons do not 21 The one optical situation for which partitioning absorption into particulate and dissolved components is necessary is in the CDOM fluorescence calculation To predict CDOM fluorescence it is first necessary to know how much light was absorbed by CDOM Therefore if the user selects the option of including CDOM fluorescence in the HE52 run the user is given the option of naming a data file containing a and c values from a filtered ac 9 These CDOM absorption values are used only in the fluorescence calculations In most instances CDOM fluorescence is a small contr
40. DUCTION The separate Users Guide gives a general overview of the capabilities of HydroLight EcoLight version 5 2 describes how to install and run the code and shows example output This Technical Documentation is designed to provide more detailed information for users who are already familiar with the basic operation of HydroLight EcoLight version 5 2 For brevity HydroLight EcoLight version 5 2 will be called HE52 when referring to common features of the codes Differences in HydroLight and EcoLight will be noted as necessary in which case the individual names will be used This documentation includes descriptions of various models available within HE52 and instructions on how to customize HE52 for a user s specific purpose Formats for data files are described as are procedures for creating new phase function and sea surface files This document assumes that the reader has basic familiarity with HE52 Ifnew to HE52 at least run the examples found in the Users Guide to get an idea of what input is required and how it can be specified It is also assumed that the reader is familiar with the basic terminology and notation of optical oceanography If this is not the case then the reader should first consult the review paper by Mobley 1995 or one of the books by Kirk 1994 Spinrad Carder and Perry 1994 or Mobley 1994 The Ocean Optics Web Book at http www oceanopticsbook info also contains much useful infomation The tex
41. E52 for comparison purposes It is called the Classic Case 1 IOP model in the HE52 user interface 2 4 NEW CASE 1 This is a new IOP model which is based on recent publications on absorption and scattering in Case 1 waters We therefore describe it in some detail for comparison with the well known classic IOP model just described which has been used by researchers for decades 2 4 1 Particle absorption It is well known that there is great variability in chlorophyll specific absorption spectra a A In particular the spectral shape of a A changes with the chlorophyll concentration owing to species composition and pigment packaging effects e g Bricaud et al 1995 1998 Thus the next step in improving the particle absorption model is to allow the chlorophyll specific absorption a A to depend on the chlorophyll concentration itself Bricaud et al 1998 therefore model particle absorption as a EA a Chl d Chl z A CAAD 8 Chi z 6 AN ChI The Bricaud et al 1998 paper gives A and E A between 400 and 700 nm Extending the Bricaud et al values from 700 to 1000 nm is easy because phytoplankton absorption is essentially zero in the IR However there are very few measurements of phytoplankton absorption below 350 nm so extending A A and E A down to 300 nm is an uncertain process Morrison and Nelson 2004 their Fig 1 show two normalized phytoplankton absorption spectra from 300 to 750 nm
42. ELS 64 8 1 Subroutines for IOP Models i co a004 4 ova ised es edad ec evesae neds 65 8 2 Subroutines for Chlorophyll Profiles 00 0 0 68 8 3 Subroutines for CDOM Absorption 0 0 0000 ce eee 69 8 4 Subroutines for Bottom BRDFs 0 0 00 ccc cee 69 8 5 Subroutines for Sky Models 0000s 70 8 6 Subroutines for Bioluminescence 0 00 cece eee 70 D gt SPECIAL RUNS sc cie enhed Sierra al oils 5 Bac Sa encutnet oo oea oe esa Se Ses 71 9 1 Defining New Quad and Band Partitions and Creating New Surface Files 72 9 2 Discretizing a Phase BuncChony 223 5 iaGe Dake gaan tees r hes 173 9 3 High theta resolution EcoLight runs 0 0000 c eee nee 78 10 WHEN SOMETHING GOES WRONG 00 00 0 eee eee eens 81 IIE REFERENCES scoot i ea EE ceed A dae a eg dass aca pcre heen EN ERSE Clg 85 APPENDIX A DETAILED DESCRIPTION OF THE RUN TIME INPUT FOR STANDARD RUNS 3 xy outbo fia 22Gb See tae aes aay 87 APPENDIX B DESCRIPTION OF ARRAYS AND I O UNIT NUMBERS 103 BY The IMISC Aay a4 3 a cleo 4 axed oo bv tee 2 two ae ea eee eee ae 103 B2 The FMISC ATAY a me eA mae aN Pn a Ie ra aah 104 B33 The DA TAPILES Array syle beee nati dadinds Seana eens adele E 105 B 4 Reserved VOU nit Numbers 0 4 542 8 sep ed Gd See ha bad Ae SOLS eed 106 APPENDIX C LICENSE AGREEMENT FOR USE OF HYDROLIGHT ECOLIGHT VERSION 5 04 108 iv 1 INTRO
43. Optmodel 1 for the classic Case 1 IOP model IOP routine abcase1 f 89 iOptmodel 2 for the Case 2 IOP model IOP routine abcase2 f iOptmodel 3 for the IOP data model IOP routine abacbb f iOptmodel 4 for the new Case 1 IOP model IOP routine abcase1new f iOptmodel 1 for a user written IOP model the user s named IOP routine in HE52 code users will be used iSkyRadmodel indicates which sky radiance model has been selected in the GUI iSkyRadmodel 0 for the analytical sky radiance model single wavelength runs only iSkyRadmodel 1 for the semi empirical sky radiance model of Harrison and Coombes 1988 routine hcnrad will be used iSkyRadmodel 2 to call a user defined sky radiance model the user s named sky radiance routine in HE5S2 code users will be used iSkyIrradmodel indicates how the sky irradiances are to be obtained If iSkyRadmodel 0 then iSkyIrradmodel 0 flags use of the analytical sky irradiance model single wavelength runs only If iSkyRadmodel 1 or 2 then iSkyIrradmodel 0 to call RADTRANX to obtain the direct and diffuse irradiances iSkyIrradmodel 1 to read a user defined data file with the total irradiances RADTRANX will be used to partition the total into direct and diffuse contributions available when iSkyRadmodel 1 or 2 iSkylrradmodel 1 to read a user defined data file with the direct and diffuse irradiances RADTRANX will be NOT be called iSkylrradmodel 2 to read a user defin
44. Ps bottom reflectance etc for your particular water body Each type of data required by HE52 e g a and c IOP data a chlorophyll concentration profile or a bottom reflectance spectrum has a standard format for that type of data You can easily replace the HE52 default or example data with your own You only need to put your data into the standard format for that type of data and then browse for the name of your data file when defining the run in the HES2 user interface Standard format data files are always uncompressed ASCII text files so that they are easy to view and edit and are independent of the computer operating system The standard file format consists of ten header records which can be used to identify the data file or make any other comments about the data as the user desires These records are read by HES2 as character strings and copied to the printout for documentation of the file read but the information in these headers is not otherwise used by HE52 Records 11 and onward contain the actual data The format of these subsequent records varies according to the type of data being read HE52 will recognize an end of file when reading a data file but it is recommended that the end of the data be explicitly flagged to prevent errors that can occur if blank records are accidentally inserted at the end of a file and are then read as data this will usually cause run termination with an error message The end of data is usually flag
45. The pure water phase function green line file pureh2o dpf has b b 0 5 23 The Fournier Forand phase functions used for dynamic determination of the phase function also can be selected by name in the UI just like any other phase function If this is done the named phase function will be used at all depths and wavelengths for that IOP component The Fournier Forand phase functions are named by their backscatter fraction For example the phase function on file FFbb0001 dpf has a backscatter fraction of 0 0001 phase function FFbb018 dpf has a backscatter fraction of 0 018 and so on However it is just as easy to specific the backscatter fraction e g 0 018 in which case the Fournier Forand phase function on file FFbb018 dpf will be used The dpf extension of the phase function files just means that the file contains a discretized phase function The files themselves are ASCII text files However it is not meaningful to open a dpf file and look at the contents in hopes of seeing the phase function as a function of scattering angle The discretized phase functions used by HydroLight are the quad averaged phase functions computed as described in Light and Water Chapter 8 This computation is carried out by the code in the HE52 SpecialRuns PhaseFunction directory which is described in 9 2 below A couple of final comments are in order regarding the use of Measured OPdata First it is not HES2 s job to quality control your ac
46. ae Peale ee pee 35 2 SRY MODELS 3 ralera say do eee tee RE ay A Ee is Bees 37 5 1 The RADTRAN X Sky Irradiance Model 00 0 c eee ee ee eee 38 5 2 The Empirical Sky Radiance Model 2 0 0 ce cece ee eee eee 42 5 3 The Analytic Sky Radiance Model 0 0 ccc eee nes 44 5 4 Input of User defined Sky Irradiances 0 0 0 0 44 5 4 1 Running HE52 outside the 300 1000 nm range 45 5 4 2 Simulation of Lidar induced inelastic scatter 0004 49 INELASTIC SCATTERING MODELS 0c cece cee ee ene eeees 51 STANDARD FORMAT DATA FILES 0 00 cece eens 52 LI Concentraton LIAL Ass 55 3 ane sen ges a Fi hee dig hon Mates eR ecarga tee sce ae eh 53 7 2 Concentration specific Absorption and Scattering Data 54 7 3 Absorption and Beam attenuation Data naunan aasan e eee ene 55 iii T4 Backscatter Data eredo Sang seven Seda sk sgh OS a wees ead Aa Goad sete ene 57 7 5 Bottom Reflectance Data gonii nn 2 tei bay BBS OP ee ea 57 TO Sky bradiance Data doeroe he nek Se ead KPa te Paw ee Sad salem as 58 Tal Biolumnescence Datas coat atten and tae aetew mane TA EAE 60 7 8 Filling in Missing DBL sas ei rc eS teas ha eek Sha kG ed Bek kt ee Te 61 7 8 1 Interpolation within the measurement range 00005 61 7 8 2 Extrapolation outside the measurement range 000 62 8 WRITING SUBROUTINES FOR IOP AND OTHER MOD
47. al of Geophysical Research L amp O Limnology and Oceanography Acharya P K et al 1998 MODTRAN User s Manual Versions 3 7 and 4 0 Draft Air Force Research Lab Hanscom AFB Ma 01731 3010 79 pages Ahn Y H 1999 Proprietes optiques des particules biologiques et minerales Ph D dissertation Univ P amp M Curie Paris France Boss Twardowski and Herring 2001 AO 40 27 4885 4893 Bricaud et al 1995 JGR 100 C7 13321 13332 Bricaud et al 1998 JGR 103 C13 31033 31044 Brunger A P and F C Hooper 1993 Solar Energy 51 1 53 64 Cox C and W Munk 1954a J Opt Soc Am 44 838 850 Cox C and W Munk 1954b J Mar Res 13 198 227 Fournier G R and J L Forand 1994 Ocean Optics XII SPIE vol 2258 194 201 Fournier G R and M Jonasz 1999 Computer based underwater imaging analysis In Airborne and In Water Underwater Imaging SPIE Vol 3761 corrected version Gathman S G 1983 Optical Engin 22 57 62 Gregg W W and K L Carder 1990 L amp O 35 8 1657 1675 Gordon H R and A Morel 1983 Remote Assessment of Ocean Color for Interpretation of Satellite Visible Imagery a Review Lecture Notes on Coastal and Estuarine Studies Volume 4 Springer Verlag New York 114 pp Gould R W R A Arnone and P M Martinolich 1999 AO 38 12 2377 2397 Harrison A W and C A Coombes 1988 Solar Energy 41 4 387 392 Hawes S K 1992 Quantum fluorescenc
48. ally made black Ed_direct Ed_diffuse 0 0 at all bands other than the Lidar wavelength which is all direct Ed the Lidar beam and nothing else For this example at 488 nm the run bands defined in the User Interface MUST include a band from 487 5 go 488 5 Also place the sun at the zenith zenith angle 0 0 in the User Interface wavelength Ed direct Ed diffuse nm W m 2 nm W m 2 nm 488 0 1 0 0 0 1 0 1 0 1 0 Fig 25 Example HSF input file for lidar simulation at 488 nm The first 8 header records are reduced in size for printing here 3 On the wavelength form it is necessary to enter a 1 nm wide band centered on the lidar wavelength regardless of the wavelength resolution of the other bands You can enter all wavelength band boundaries manually or use this trick Enter the bands 49 you want for the output using the minimim maximum bandwidth option e g 480 to 720 by 5 nm Hit continue to go to the next form This will enter the wavelength band list onto the wavelengths form Then back up to the wavelengths form and select the option to enter a list of wavelengths and manually add a 1 nm wide lidar band at the appropriate location in the list of wavelengths Save the inputs when you get to the final form so that the wavelength list will be saved for subsequent runs The final list of wavelengths for this example should then look like 480 485 487 5 488 5 490 495 715 720 The red wav
49. alues WARNING this option is not recommended as it can produce a substantial amount of output iOptDigital is a flag for the inclusion omission of the digital output file if i OptDigital 0 Droot txt will not be generated for this run if i OptDigital 1 Droot txt will be generated for this run iOptExcelS is a flag for the inclusion omission of the Excel Single wavelength output file if iOptExcelS 0 Sroot txt will not be generated for this run if i OptExcelS 2 Sroot txt will be generated for this run iOptExcelM is a flag for the inclusion omission of the Excel Multi wavelength output file if iOptExcelM 0 Mroot txt will not be generated for this run if iOptExcelM 1 Mroot txt will be generated for this run iOptRad is a flag for the inclusion omission of the full radiance printout file if iOptRad 0 Lroot txt will not be generated for this run if i OptRad 1 Lroot txt will be generated for this run nwskip sets whether to skip alternating wave bands and then interpolate to fill skipped bands nwskip 1 solves all wavelengths nwskip 2 solves every other wavelength etc Record 4b model options This record specifies the models that will be used for this run and is used by incfiles std Names of variables iJOPmodel iSkyRadModel iSkyIrradModel iChl iCDOM Example 1 1 0 2 0 iOptmodel indicates which IOP model has been selected in the GUI iOptmodel 0 for constant IOPs IOP routine abconst f will be called i
50. and then obtain b from b c a with a being determined by Eq 6 as described above This is the approach taken in the new Case 1 model which uses cE c cnor 10 where 16 lt Il 0 5 log oCh 0 3 for 0 02 lt Chl lt 2 0 for Chl gt 2 Thus the new model uses the chlorophyll dependence of c 660 from Loisel and Morel 1998 Eq 9 and assumes that c has the same chlorophyll dependent wavelength dependence as the b model of Morel et al 2002 Eq 8 The values of c and n can be chosen by the user in the HE52 UI The default values which apply to near surface waters are c 0 407 and n 0 795 from Loisel and Morel 1998 Eq 5 Figure 8 shows example a for mid range UV absorption b and c spectra for near surface waters c 0 416 and n 0 766 in Eq 10 along with the classic b of Eq 4 with b 0 3 n 0 62 and m 1 The scattering coefficients are not too different at low chlorophyll values but the new b has a different wavelength dependence and is much larger in magnitude by up to a factor of three at high Ch values Unlike in the classic scattering model of Eq 4 the wavelength dependence of b now depends on Chl and is more complicated These differences in scattering will have a significant effect on computed radiances Chl 0 01 Chi 0 1 0 04 lil 0 15 RTT a 2 6 e a 500 tii 0 00 CN aC TPT rT 300 400 500 600 700 800 90
51. andard tasks of checking for bad data applying scattering corrections and smoothing or binning the data with depth or wavelength to remove as much noise as possible while retaining the relevant physics and biology Recall the comments associated with Fig 1 of the Users Guide and the ac s caveat at the end of 2 6 Once the data have been cleaned up they can be placed in a file in the HES2 standard format for ac 9 or similar data which is illustrated below Example showing the standard format for ac 9 data as used for ac 3 data This file can be read by IOP routine abacbb Record 11 contains the number of wavelengths and the nominal wavelengths free format records 12 and onward contain the free format ac 9 data ordered as depth a wavelength 1 a wavelength n c wavelength 1 c wavelength n A negative depth flags the end of file depth a 456 a 488 a 532 c 456 c 488 c 532 3 456 488 532 0 832 0 18703 0 147457 0 128416 1 18221 1 07972 0 967512 0 965 0 199363 0 160162 0 139046 1 15831 1 05569 0 944565 1 054 0 194962 0 155575 0 135086 1 16072 1 05842 0 948916 1 188 0 183068 0 143292 0 123584 1 17185 1 07182 0 967417 L277 0 175154 0 134851 0 114446 1 18305 1 08881 0 992936 L 366 0 168063 0 128259 0 108202 1 19956 1 10532 1 01611 Many depth records omitted here 14 405 0 0961292 0 066717 0 0661964 0 298839 0 253469 0 270901 14 583 0 0954552 0 0662783 0 0653839 0 295006 0 252532 0 2700
52. are not provided in HE52 because they are seldom used However the routines provided allow these phase functions to be discretized and used in HES2 if desired 74 Defining and discretizing a new phase function require several steps which are described next Note that all of these calculations are performed within the HE52 SpecialRuns PhaseFunction directory and subdirectories e Step 1 Write a FORTRAN subroutine to define the desired phase function as a function of scattering angle y This can be done by modifying any of the distributed routines shown in Table 3 or by following the template on file HE52 examples templates phasefun f Note that any scattering phase function Bay must satisfy the normalization condition E 2x f B y siny dy 1 0 It is your responsibility to make sure that your phase function satisfies this normalization condition before doing the discretization calculations Note further that the subroutine must return a value of the phase function for any scattering angle 0 lt y lt a actually for any value of cosy where 1 lt cosy lt 1 After the new subroutine has been written give the file a name of the form mynewpf f for example and save it in both the HE52 SpecialRuns PhaseFunction Ecolight and the HE52 SpecialRuns PhaseFunction Hydrolight directory Step 2 The routines on files HE52 SpecialRuns PhaseFunction EcoLight phasef f and HE52 SpecialRuns PhaseFunction HydroLight phasef f will call your ph
53. ase function subroutine during the discretization calculations Therefore you must change the subroutine calls in both phasef f files to show the name of your phase function subroutine For example if the current call in phasef f is call pfpart cospsi pfvalue and you have named your subroutine mynewpf you must change the above call to call mynewpf cospsi pfvalue This change is made by opening file phasef f with a text editor making the change near line 36 and re saving the file 75 e Step 3 Run the batch script makediscpf_EL bat in the PhaseFunction Ecolight directory and run makediscpf_HL bat in PhaseFunction Hydrolight You can run these scripts by double clicking on the file names in Windows Explorer or by entering the file name from a command window in the appropriate directory These programs compile your new phase function subroutine and create the corresponding executable files HE52 SpecialRuns PhaseFunction discpf_EL exe and HE52 SpecialRuns PhaseFunction discpf_HL exe Check to see that these two files were created If they are not there the compilation failed probably because there was a bug in your phase function routine If debugging is required open a command window change directories to the HE52 SpecialRuns PhaseFunction Hydrolight directory and enter the command makediscpf_HL bat You can then see if there are any error messages during the compilation Repeat for the EcoLight directory e Step 4
54. asured and modeled red CDOM absorptions 15 2 4 3 Scattering Just as for absorption recent papers have presented improved models for particle scattering in Case 1 waters Morel et al 2002 Eq 14 model the particle scattering coefficient as A b EA stomerr 2 8 where v 0 5 log Chl 0 3 for 0 02 lt Chl lt 2 0 for Chl gt 2 Thus the wavelength dependence of the scattering coefficient now depends on the chlorophyll concentration unlike in the classic model of Eq 4 In particular v now lies between 1 and 0 A value of v 1 as seen in Eq 4 is known from Mie theory to be valid for nonabsorbing particles with a Junge particle size distribution of slope 4 Loisel and Morel 1998 studied the relationship between particle beam attenuation at 660 nm c 660 and the chlorophyll concentration They found the functional form c 660 c CHOT 9 The values of c and n are different for near surface down to one penetration depth as relevant to remote sensing and deeper waters Because c b at 660 Morel et al 2002 adopt the coefficients for Eq 9 for use in Eq 8 after shifting the reference wavelength to 550 nm Thus for near surface waters Morel et al 2002 use b 0 416 and n 0 766 in Eq 8 However a power law in wavelength of the form of Eq 8 generally gives a better fit for c than for b e g Voss 1992 Boss et al 2001 Thus it is probably better to model c
55. ation shall not dilute or limit Curtis Mobley s rights with respect to those portions of any derivative work in original modified or combined form incorporating any portion of HES2 Any portion of HE52 included in a derivative work shall remain subject to the terms of this Agreement 6 The User acknowledges that HE52 is valuable to Curtis Mobley and shall be held in confidence as proprietary to Curtis Mobley and that HE52 is licensed solely for the User s use subject to the terms of this Agreement The User agrees not to disclose or provide HE52 in any form to any person party or entity without the prior written consent of Curtis Mobley except that the User may provide HES2 to his employees consultants or students as reasonably necessary to exercise his rights under this Agreement 7 The User agrees that HE52 including any documentation and or instructions for its use is made available without warranty of any kind expressed or implied or statutory including but not limited to the implied warranties of merchantability fitness for a particular purpose or conformity with whatever documentation user manuals or other literature as may be issued by Curtis Mobley or Sequoia Scientific Inc from time to time The user is warned that HE52 is not fault tolerant and must not be relied upon in situations where financial loss disruption of business or research or other pecuniary loss could occur from an inability to use HE52 or from the
56. avelength dependent dynamically determined determined at run time depth below the greatest output depth requested by the user This dynamic bottom boundary depth guarantees that any effects due to the use of a source free lower boundary condition in arun including internal sources will not affect the radiances and other quantities at the greatest depth of interest Runs for a wide range of water conditions from very clear to very turbid and for wavelengths from 300 to 1000 nm show that applying the bottom boundary condition at 20 optical depths below the greatest depth of interest is sufficient to give good radiances at the greatest depth of interest HE52 therefore does the following At the start of a run it computes the wavelength dependent depth ZA Zmax requested 20 c Z a 5 A max where C Z ax5 A is the beam attenuation coefficient at the maximum output depth requested by the user It then finds z A Anax Lhe infinitely deep source free bottom boundary condition is then applied at depth the maximum of z A which occurs at some wavelength max Z4 2 Zyhinan if 2 lt 2 max max ZA Zmax requested 20 C Zinax gt A IFA gt Anax max 33 These bottom depths guarantee that the RTE is solved as deeply as needed at all wavelengths A lt Anax In order to compute the inelastic scattering source terms needed at wavelength Anax where the water is clearest and that it is solved no deeper than needed at long
57. between the sun and the downwind direction is irrelevant and is set to sunphi 0 by the front end programs If the full Cox Munk equations are used in HE52 which requires regenerating the surface files the default value of sunphi 0 must be changed manually in Iroot txt before running HE52 97 Record 8b atmospheric conditions Names of variables jday rlat rlon pres am rh wv vi wsm ro3 Example 180 60 20 29 92 5 8 2 5 15 5 99 This record contains the atmospheric parameters that will be used to specify the earth sun distance jday the solar zenith angle if iflagsky 3 jday rlat rlon ozone climatology if ro3 99 and for the Gregg amp Carder atmospheric irradiance model if used pres am rh wv vi wsm ro3 fiday is the day of the year Julian day fjday 1 0 for January 1 rlat is the latitude in degrees positive for north and negative for south rlon is the longitude in degrees positive for east and negative for west pres is the average sea level pressure inches of Hg am is the marine aerosol type 1 marine to 10 continental see Gathman 1983 for a description of aerosol types rh is the relative humidity percent wv is the precipitable water content cm vi is the average horizontal visibility km wsm is the 24 hr average wind speed m s ros isthe total ozone Dobson units set to 99 if a climatological value is to be used RECORD 9 Surface Information Thi
58. ble for specifying Acpoy Z A One option is to let deyo Z A covary with the chlorophyll absorption as it does in the classic casel IOP model Then Acpom Z4 0 2a 440 exp 0 014 A 440 It is also possible to read a user supplied subroutine or data file to obtain a poy Z A 2 9 Specific Scattering Models The UI provides several options for how to model scattering by IOP components Scattering is often modeled by a power law dependence on wavelength b z d 2 X z where X is a concentration e g Chl in mg m or mineral particles in gm m and b A mM andn are model parameters The commonly used Gordon Morel 1983 scattering model for chlorophyll bearing particles is a special case of the power law model with X Chi b 0 3 550 nm m 1 and n 0 62 550 Bey Zd 03 a Chi z This model for the scattering coefficient is used in the classic casel IOP model Gould et al 1999 presented a scattering model that is a linear function of wavelength A me XO ott Their best fit parameters for a range of water bodies are b 0 5 m 0 00113 A 550 nm i 1 62517 and n 0 62 This model can be selected in the UI when using the Case 2 IOP model Note that these two power law and linear models are models of the scattering coefficient rather than models of the concentration specific scattering coefficient 28 Scattering can also be modeled by specifying a file of specifi
59. bove for input of a file with Ed_ direct and Ed_diffuse is selected but there is only one wavelength in the file not counting the negative wavelength that flags the end of data then that input is taken to be a lidar input and all other bands in the run are given zero sky inputs i e the sky is black except at the lidar wavelength To run a lidar simulation do the following in the UI 1 Place the sun at the zenith zenith angle of 0 to simulate the lidar beam propagating straight down The other atmospheric parameters will not be used because RADTRANX will not be called nor will the Harrison and Coombes sky model be called However the UI still requires values to be entered since the UI doesn t know that the HE52 code won t use some inputs 2 Select the sky irradiance option for reading a user file containing direct and diffuse sky irradiance The file to be read must have the one wavelength format seen in Fig 25 The lidar irradiance in W m nm should be entered as the Ed_ direct value with the Ed_diffuse value set to zero Example direct and diffuse sky spectral irradiances for LIDAR SIMULATION in a black sky This example assumes the run is done for a 1 nm band from 487 5 to 488 5 containing input from a 488 0 nm Lidar plus other bands that can be anything e g 5 or 10 nm bands For input files having ONLY ONE input wavelength not counting the end of file record of negative values the sky is automatic
60. c scattering data and a file of concentration data bA bX where b is the concentration specific scattering coefficient and X is the concentration Files of mass specific scattering data for mineral particles corresponding to the mass specific absorption data seen in Fig 13 are also available in the HE52 data defaults directory Figure 14 shows the wavelength dependence of these mass specific scattering coefficients given by the various data files Again users are free to use either the provided models or their own data files to specify the specific scattering coefficient for each component If using the example mineral spectra provided with HE52 note that HE52 does not check to see if you have chosen self consistent mineral a and b spectra e g that you have chosen b for red clay if you picked a for red clay o m 11E o D 101 solid is measured Oc F dashed is extrapolated d OOF lt SD 08b 7 F m gg 07E oO t calcareous san S 52 0 6 f yellow clay a 0 5 E red clay 6 i D t 0 4 E brown earth ssis EG ggi dois sO 300 400 500 600 700 800 900 1000 wavelength nm Figure 14 Mineral mass specific scattering spectra as provided with HE52 29 3 Bottom Reflectance Models 3 1 Finite depth Water HE52 can simulate both finite depth and infinitely deep water bodies In the finite depth case a physical bottom is placed at depth z where Z is the last depth specified i
61. cent source strength in W m nm For the example of Section 6 2 of the Users Guide the file UGEx2 txt which was generated by the front end program and which contains the input for the HE52 run is shown on the next page Note that dummy names were generated by the UI as place holders for the names of various files that were not needed for this run File names for ac 9 chlorophyll and bottom reflectance data are always read by the HE52 input routines which do not know whether or not the files will be needed later by the core routines In the present example only the Chlzdata txt file will be opened and read by HE52 in the course of solving the RTE 101 O 400 700 0 02 488 0 00026 1 5 3 Users Guide Example 2 Classic Case 1 IOPs UGEX2 Ope 2 lyk yl 1 1 0 2 0 2 2 Of Ly O 1 440 0 1 0 014 2 3 440 0 1 0 014 data H20abDefaults_ S FAwater txt data defaults astarchl txt 0 999 999 1 550 0 3 1 bstarDummy txt dummybstar txt 04 05 850 0 01 0 0 0 005 550 0 01 pureh2o dpf 9 0 9 0 62 0 avgpart dpf 35 350 450 550 560 650 660 0 1 0 1 2 2r By 204 1 O 03 1 99809 he OZ 0 11 360 460 370 470 570 670 380 480 580 680 490 590 690 0 0 3 29 5925 1 3 34 1 20 80 35 0 2 4 6 8 99 z9 390 10 O72 199 9 99 400 500 600 700 410 510 610 420 520 620 430 530 630
62. coLight run This action will open a command window and perform the run this is just what the UI itself does The executable HydroLight and EcoLight codes themselves are located in the HE52 code directory named mainHL_stnd exe and mainEL_stnd exe Option 2 An equivalent way to make a run is to e Open acommand window e g via start gt all programs gt accessories gt command prompt in the Windows XP or 7 operating systems e In the command window change directories to the HE52 run directory via the cd command e g cd c HE52 run If HES2 is installed in the C directory the command prompt will then be c HE52 run gt e Type runHE exe or runEL exe at the command prompt and hit enter An execution window will open and the code will run just as in Option 1 However for both Option 1 and Option 2 the execution window will close automatically at run termination which is undesirable if error messages appear there The next two options allow you to make a single run and force the command window to stay open at the end of the run Option 3 To force the command window to stay open after run termination e Open a command window and cd to the HE52 run directory e Type the command code mainHL stnd exe lt batch Iroot txt at the command prompt and hit return I root txt is ofcourse replaced by the actual name of your Troot txt file This pipes the input file to the HydroLight executable and makes the run The comma
63. common incfiles_default for common incfiles_user for 1f95 dbl nco nlst nap ndal nchk ntrace inln npca nsav stchk o1 nw nwo c common incfiles_user for 1f95 dbl chk nco nlst pca nsav stchk ntrace ml winapi win nvsw nw c common w f90 1f95 dbl chk nco nlst pca nsav stchk ntrace ml winapi win nvsw nw c common f90 1f95 dbl nco nlst nap ndal nchk ntrace inln npca nsav stchk o1 nw nwo c f ml msvb 1f95 dbl co nlst nchk pca nsav stchk nw nwo c for ml msvb 1f95 obj stack 750000 ml msvb lib common HE52info lib nomap winconsole out mainHL_stnd exe move mainHL stnd exe del obj del mod Refer to the Lahey Fortran 95 documentation for description of these compilation options There is a corresponding set of commands in file HE52 code Ecolight makeExeStnd bat that compiles the EcoLight code The compilation listing on the command window goes by very quickly and can be too long for the number of lines retained for viewing in the command window When debugging you can save the entire listing to a text file via makeExeStnd HL bat gt complist txt You can then examine complist txt with a text editor to see any error messages Compilation error messages issued by the Lahey compiler are usually quite specific and tell you the routine line number and type of problem encountered 84 11 REFERENCES Journal abbreviations AO Applied Optics JGR Journ
64. d e anew surface wind speed or index of refraction is needed e anew scattering phase function is to be prepared and added to the collection of available phase functions HydroLight Technical Note 2 in HE52 documents discusses the angular resolution and shows that little is gained by going to a finer quad resolution Likewise water surfaces describing wind speeds between U 0 and 15 m s and water indices of refraction between n 1 32 and 1 38 are available with the distributed code These U and values cover the full range of oceanic conditions for which the Cox Munk sea surface model is appropriate and for which HE52 should be run Thus new surface runs are needed only if new quad band layouts are also being generated We discourage all but the most advanced users with very special needs from altering the default quad band partitioning However many users want to add their own scattering phase functions to the default set that comes with HE52 so special runs for discretizing phase functions are sometimes needed Because of the infrequent need for making special runs control of these runs is not incorporated into the UI Special runs can be made only by creating the needed small input files with a text editor and perhaps also modifying a subroutine or writing a new one e g to define a new phase function and then submitting the run from a command window 71 9 1 Defining New Quad and Band Partitions and Creating New Surface Files
65. d measurements of the backscatter coefficient if available e provide input to user written IOP models with any number of components The total IOPs of a water body are built up as a sum of IOPs attributable to the various components of the water body Thus the total absorption coefficient is computed from Light and Water Eq 3 10 ncomp doa 3 a z Here a z A is the absorption coefficient of the i component of the water body e g pure water chlorophyll bearing phytoplankton CDOM or mineral particles which in general is a function of the depth z and wavelength A The number of components in the IOP model is ncomp A similar equation is used to compute the total scattering coefficient b To complete the specification of the water IOPs a scattering phase function must be specified for each component In some IOP models the user explicitly selects the phase function to be used for each component in others the phase functions are pre determined The Measured IOP data model also has the option of allowing the phase function to be determined dynamically i e as HydroLight or EcoLight runs to solve the radiative transfer equation from depth and wavelength dependent scattering and backscattering coefficients read from user supplied data files Table 1 lists the various IOP models available in HE52 These models are then described name components comments pure 1 pure water returns a and b values for pure wate
66. ded to list all of the band boundaries The a and b values as returned by the IOP model AT THE BAND CENTERS will be used The band averaged sky radiance will be used RECORD 7 Inelastic Scattering and Internal Sources Gives the flags specifying whether or not internal sources and inelastic scatter are to be included in the run Names of variables ibiolum ichlfl icdomfl iraman icompchl Example 0 1 0 1 2 ibiolum is a flag for the inclusion omission of bioluminescence if ibiolum 0 there is no bioluminescence present if ibiolum 1 the run includes bioluminescence user supplied routine sObiolum or its replacement see 8 6 is required ichifl is a flag for the inclusion omission of chlorophyll fluorescence if ichlfl 0 the is no chlorophyll fluorescence present if ich fl 1 chlorophyll fluorescence is present routine chlzfunc or chizdata is called icdomfl is a flag for the inclusion omission of CDOM fluorescence if icdomfl 0 there is no CDOM fluorescence present if icdomfl 1 CDOM fluorescence is present routine acdom is required iraman is a flag for the inclusion omission of Raman scattering if iraman 0 there is no Raman scattering present if iraman 1 Raman scattering is present icompchl is the integer index for the chlorophyll component and will equal zero if chlorophyll is not specified in the run icompchl is used only when chlorophyll fluorescence is included 95 RECORD GROUP 8 Sky Model Th
67. detect the new subroutine compile it and incorporate your new BRDF into the HydroLight bottom calculations The only fundamental constraint on the bottom BRDF used in HydroLight is that the BRDF depend only on the difference of the incident and reflected azimuthal angles i e the bottom must be azimuthally isotropic Note also that the irradiance reflectance of a non Lambertian bottom depends on the incident lighting therefore the bottom reflectances available in the UI are not used if the bottom is non Lambertian A non Lambertian BRDF 69 routine must define both the directional pattern of the bottom reflected radiance and its magnitude Doing this properly requires a thorough knowledge of BRDFs There is a lengthy set of notes on BRDFs in the HE52 documents directory 8 5 Subroutines for Sky Models Although the sky models provided and discussed in 5 are adequate for most purposes the UI has an option for specifying a user defined data file containing sky irradiance obtained from direct measurements or from a separate model such as MODTRAN You can also specify a different routine to produce the radiance distribution in the UI or write your own sky irradiance or radiance pattern models by following the corresponding templates found on files HE52 examples templates skyirrad txt and skyrad txt This is the approach to be taken if you wish to use measured or externally modeled sky radiances e g from a separate run of an atmosph
68. e Record 9 depth m Chl mg m 3 0 0 s5 5 0 0 67 10 0 T2 1 50 2 1 20 0 1D 25 0 0 4 lis 0 4 Note that this file has exactly 10 header records If you don t need 10 records to identify your file enter place holder records 1 e not blank lines as was done here for records 8 and 9 In this example there are then six data records The last record flags the end of the file with a negative depth Note that the last record is just the previous record repeated but with the depth changed to a negative number This same format is used for mineral particle concentration data see for example file HE52 data examples minzdata txt and for the absorption by CDOM ata given reference wavelength Note that the concentration data must have units that are consistent with the concentration specific absorption and scattering spectra 53 7 2 Concentration specific Absorption and Scattering Data Concentration specific absorption and scattering spectra are needed by some IOP models to convert chlorophyll or mineral concentrations to absorption and scattering coefficients Such data files have the same format as just seen for concentration profiles except that the depth is replaced by wavelength For example the mineral particle mass specific scattering spectrum for red clay from file HE52 data defaults bstarmin_redclay txt looks like the following RED CLAY mass specific mineral scattering coefficient bmin wavelength in m 2
69. e to turn your IOP model into a subroutine suitable for calling by HE52 The documentation in HE52 templates ab f explains how this must be done Both HydroLight and EcoLight use the same format for IOP subroutines which are found in the HE52 code common directory Thus you only have to write one IOP subroutine which is then useable by both HydroLight and EcoLight As explained in the ab f template you must give your IOP subroutine a unique name e g mylIOPs or abmodel5 and save it as a file with the corresponding name e g mylOPs f in the HE52 code common directory Your model for a and b is then ready for use by HE52 To run HE52 with your IOP model you select the USER IOP MODEL option when running the UI Then enter the name of your routine e g myIOPs the number of components and a descriptive name to be used to identify each component e g pure water phytoplankton or bubbles Note that these names are used only by the UI for prompting you 66 for the input for each component The UI will then lead you through the component by component specification of the absorption and scattering properties the concentration and the phase function If you have named your IOP model myIOPs in your file named mylOPs f for example the UI will generate a call to your subroutine with the following form call myiops z wavenm ncomp acomp bcomp atotal btotal This call is placed in an automatically creat
70. e efficiencies of marine fulvic and humic acids Master s thesis Dept of Marine Sci Univ of South Florida St Petersburg FL 92 pp Kasten F and G Czeplak 1980 Solar Energy 24 177 189 Kirk J T O 1994 Light and Photosynthesis in Aquatic Ecosystems Second Edition Cambridge University Press Cambridge 509 pages 85 L amp W see Mobley 1994 Light and Water see Mobley 1994 Loisel and Morel 1998 L amp O 43 5 847 858 Mobley C D 1994 Light and Water Radiative Transfer in Natural Waters Academic Press San Diego 592 pp Revised version available on CD from the author Mobley C D 1995 The Optical Properties of Water Chapter 43 in Handbook of Optics Second Edition Volume I McGraw Hill and Optical Society of America 56 pages Mobley C D B Gentili H R Gordon Z Jin G W Kattawar A Morel P Reinersman K Stamnes and R H Stavn 1993 AO 32 7484 7504 Mobley Sundman and Boss 2002 AO 41 6 1035 1050 Mobley C D H Zhang and K J Voss 2003 L amp O 41 1 part 2 337 345 Mobley C D D Stramski W P Bissett and E Boss 2004 Oceanography 17 2 60 67 MODTRAN see Acharya et al 1998 Morel A 1974 Optical Aspects of Oceanography Academic Press 1 24 Morel A 1988 JGR 93 C9 10749 10768 Morel A and S Maritorena 2001 JGR 106 C4 7163 7180 Morel Antoine and Gentili 2002 AO 41 30 6289 6406 Morrison and Nelson 2004 L amp O 49 1
71. e file HE52 data defaults apstarchl txt contains a default chlorophyll specific absorption coefficient which is shown in Fig 12 This particular spectrum is taken from Morel 1988 his Fig 10c which is Prieur and Sathyendranath 1981 Table 2 Column 3 scaled so that a 440nm 0 05 m mg Chl which is a typical value for a 440nm This spectrum was selected as the default because its normalized version with a 440nm 1 shown in Fig 2 is used in the classic casel IOP model Note that there is at least a factor of five difference among measured values of specific absorption e g Prieur and Sathyendranath reported values between 0 02 and 0 1 m mg Chl at 440 nm so the values in the default file may or may not describe your actual water body chlorophyll specific absorption coef a m mg 0 00 E fice ek E EE E A N N E a A ee ae 300 400 500 600 700 800 900 1000 wavelength A nm Figure 12 The default chlorophyll specific absorption spectrum on file apstarchl txt This spectrum is based on Prieur and Sathyendranath 1981 26 Several mass specific absorption spectra for mineral particles are included in the HE52 data defaults directory Files containing values taken from Fig 4 3A of Ahn 1999 are available for brown earth calcareous sand yellow clay red clay and the average of these four These files are named astarmin_X txt where X labels the type of mineral particle in the file Ahn measu
72. e suitably referenced or acknowledged in publications papers reports presentations or other communications for which HE52 was used as a part of the study being reported upon It is requested but not required that the user of HE52 provides to Curtis Mobley a copy of all publications reports or other documents in which the use of HE52 is acknowledged THE USE OF HYDROLIGHT ECOLIGHT VERSION 5 2 IMPLIES THE USER S ACCEPTANCE OF THE ABOVE AGREEMENT 110
73. e that the water column between depth 0 the mean sea surface and depth z generally has depth dependent IOP s For infinitely deep water HE52 automatically computes the BRDF of the infinitely deep source free homogeneous layer of water below depth Zna and applies this BRDF as the bottom boundary condition at depth Zma Section 9 5 of Light and Water describes these calculations as performed by HydroLight The infinitely deep layer of water below depth z is not a Lambertian reflector 31 Because the water column is assumed to be homogeneous below depth Zax different choices for z can give greatly different light fields between depths 0 and Za if the water IOPs vary strongly with depth This is illustrated in Fig 16 which shows a true IOP profile e g the absorption coefficient at a particular wavelength for the upper 30 m of a water column and the IOP profiles that will be used in HE52 if z a is set to 10 and 20 m Clearly these IOP profiles represent different radiative transfer problems and the light fields will be max different in the regions above z max for these three situations When modeling infinitely deep water with depth dependent IOPs the user must always take z to be far enough max below the maximum depth of interest that the light field in the region of interest is not significantly affected by light from below z This depth often must be determined by max a few trial and error
74. ed data file with the total irradiances and the fraction of direct to total RADTRANX will NOT be called Parameters iChl and iCDOM indicate which IOP components are chlorophyll and CDOM respectively according to which IOP model is used For example when the new Case I IOP model is used component 2 is chlorophyll small particles for scattering calculations but containing the total Chl absorption and component 4 is CDOM This is used when fluorescence is included in the run and absorption by chlorophyll and CDOM must be separated 90 RECORD GROUP 5 IOP Specification This group of seven records provides HE52 with information about the IOPs for the various components included in the run Record 5a number of components The first record gives the number of components expected in the IOP model the number of components built into the ab routine followed by the number of concentrations that will need to be read Names of variables ncomp nconc Example 2 3 nconc gt ncomp Normally the number of components will equal the number of concentrations needed The exception is when CDOM or chlorophyll fluorescence is included when CDOM or Chlorophyll are not included as model components For example if you include CDOM fluorescence with the Case 1 water model you will have to specify CDOM absorption in the UI and ncomp 2 but nconc 3 Record 5b component concentrations This record specifies the concentration of
75. ed file HE52 code batch root for There are several important points to note about your or any IOP subroutine e An IOP subroutine gives HE52 information about the absorption and scattering coefficients but not about the corresponding phase functions The phase function to be used with each component is not built into the subroutine for a and b because users often want to change the phase function being used to model a particular component the new Case 1 IOP model is an exception For example you may wish to try several phase functions each with a different amount of backscatter when modeling a particle component Such flexibility is a powerful feature of HES2 The UI will still require you to provide concentration specific a and b spectra and concentration information for each component even if they are defined differently in your IOP model If fluorescence is included you will be asked by the UI to specify which model components correspond to the fluorescing material s included For this reason users are strongly encouraged to try to write their IOP models to work with the UI options whenever possible e You can do whatever you wish within the IOP subroutine to compute the component a and b values HES2 requires only that the format of the call to the subroutine be fixed i e the list of variables must be exactly as shown in the ab f template file and that the subroutine return the component and total a and b values T
76. elengths were added manually to the list to define the 1 nm lidar band centered at 488 0 nm Note that the input irradiance in W m nm times the 1 nm bandwidth defines the lidar irradiance in W m as is conventional when describing lidar irradiance in a very narrow band This process was used in a series of EcoLight runs to simulate Raman scatter for an infinitely deep homogeneous water body with a chlorophyll concentration of 0 05 mg m using the new Case 1 IOP model Bandwidths of 1 5 and 10 nm were used for the nominal run bandwidths and the RTE was solved to a depth of 50 m the wind speed was 0 Figure 26 shows the water leaving radiance in the region of the Raman output band centered near 582 nm HydroLight Technical Note 10 gives a detailed discussion of this example and of interpretation of Raman scatter output excitation at 488 nm 9 0E 06 8 0E 06 Fig 26 Example EcoLight 7 0E 06 A on computed Raman water 5 sor 06 leaving radiances induced by 4 0E 06 a lidar input of 1 W m at 488 F 3 0E 06 nm for three different band TOEG widths in the emission region See HTN10 for a full discussion 1 0E 06 p 0 0E 00 560 570 580 590 600 610 emission wavelength nm 50 6 Inelastic Scattering Models HE52 has the option of including three types of inelastic scattering in multi wavelength runs Raman scattering by water chlorophyll fluorescence and CDOM fluorescence
77. epresents the input for the i component for i to ncomp Names of variables itype i ibbopt i bbfrac i BfrefPL i BfOPL i BfmPL i Example 1 1 0 03 550 0 01 0 itype specifies what type of concentration input will be used itype 0 component concentration is constant with depth 1 component concentration will be given by a user supplied subroutine 2 component concentration will be read from a data file ibbopt specifies how the scattering phase function will be provided ibbopt 0 phase function file explicitly specified use pfname 1 construct a phase function to match a specified b b ratio 2 dynamically construct a phase function using b b as specified by backscatter data for use only with the MEASURED DATA IOP model 2 means that the file of backscatter data includes water value 2 means that water values have been removed from the backscatter data 3 b b specified by a power law see 2 3 using next three input values 93 BfrefPL is the power law reference wavelength used only if ibbopt 3 BfOPL is the power law b b value at the reference wavelength used only if ibbopt 3 BfmPL is the power law m value used only if ibbopt 3 Record 5h phase function file names The next ncomp number of records give the names of the files containing the discretized phase functions to be used with each component of the IOP model The order MUST match the order of the components in the subroutine for a and b see
78. er wavelengths especially beyond 700 nm where water absorption is high Use of the dynamic bottom boundary depth is the default in HE52 whenever bioluminescence or inelastic scatter are included and the water column is infinitely deep If no source terms are included or the bottom is placed at a finite depth then the bottom boundary condition is applied at z requested just as in previous versions of HydroLight If desired the dynamic bottom boundary option can be turned off in the Change Defaults form of the HE52 User Interface In order to avoid a new type of confusion output at depths below the user requested maximum depth z requested is not included in the various output files Ignorance is bliss Experience shows that the dynamic bottom boundary condition typically adds about 20 to the run time compared to the use of a fixed bottom depth That is a reasonable computational price to pay in order to avoid spurious results in simulations where the effects of sources are of interest at great depths The penalty for not going deep enough can be an apparently wrong answer for your problem 34 4 The Sea Surface Model HE52 models the wind blown sea surface using Cox Munk 1954a 1954b wind speed wave slope statistics temperature and salinity dependent Fresnel reflectances and Monte Carlo ray tracing as described in Light and Water Chapter 4 It should be noted that the Cox Munk slope statistics are based o
79. eric radiative transfer model like MODTRAN 8 6 Subroutines for Bioluminescence If you check the BIOLUMINESCENCE option when running the UI HE52 will read the specified data file or call the subroutine named in the appropriate UI text box to compute the bioluminescence source function S as a function of depth and wavelength The function S Z A computed by this routine is discussed in Light and Water Section 5 16 see in particular Eq 5 107 Subroutine HE52 code common sObiolum f is an example of such a subroutine As always the particular formulas in file SObiolum f are just examples of how to compute S each user must replace these formulas with ones that describe the internal source distribution of interest Note also that when you include bioluminescence in HE52 you are simulating a horizontally homogeneous bioluminescing layer within the water body whose details are specified in SObiolum f or in your subroutine not a point source of bioluminescence such as a single disturbed organism 70 9 SPECIAL RUNS Standard HE52 runs are those that use the various defaults for quad and band partitioning phase functions and surface reflectance and transmittance files to solve the radiative transfer equation The default files and standard runs are all that is needed by the vast majority of HE52 users A special HE52 run is required in only three circumstances e anew HydroLight quad or EcoLight band layout is neede
80. ew quad partitions and instructions on how to create new surface files That code is not distributed with HE52 for the simple reason that it is now a rather complex process to create new surface files for all of the wind speeds and indices of refraction and for both the HydroLight quad and the EcoLight band partitions In the event that you have a problem that really does require a new quad band partition please contact us for user support If we are unable to talk you out of changing the quad band partition then it is quite frankly easier for us to make the needed code modifications do the runs and send you the new surface files than to explain all of the subtle code changes and how to perform the runs that loop over wind speeds and indices of refraction Suffice it to say that a change in the quad band partition requires re creation of all of the surface files and rediscretization of all of the phase function files New surface and phase function files are needed because they depend on the quad and band partitions Those calculations can require a day of time to set up the runs and several days of computer time to do the calculations We therefore again strongly discourage users from modifying the standard HE52 quad and band partitions 9 2 Discretizing a Phase Function As noted above all phase functions must be re discretized if the quad or band partition is changed which is rarely if ever necessary However users often want to add new p
81. for HE52 43 5 3 The Analytical Sky Radiance Model The two routines that are selected when the IDEALIZED SKY MODEL option is selected in the UI are intended for specialized radiative transfer studies that need only a simple sky radiance distribution e g a sun in a uniform background sky or a heavily overcast sky The sky radiance is given by Lay L 1 Ccos6 where 0 is the viewing direction measured from 0 in the zenith direction C 0 gives a uniform background sky C 2 gives a cardioidal sky and C 1 25 gives a good approximation to a heavy overcast Brunger and Hooper 1993 Integration of Lxy 8 cos 0 gives the diffuse plane irradiance E diffuse 2nL 1 2 C 3 which sets the value of L given E diffuse Rar total from the user input Here Rais the ratio of diffuse to total sky irradiance R 0 for a sun in a black sky Rap 1 for a completely diffuse sky with no sun visible This simple sky model is intended only for use in idealized radiative transfer studies at a single wavelength It is not available in runs at more than one wavelength 5 4 Input of User defined Sun and Sky Irradiances There are three options that allow users to input their own measured or modeled sun and sky irradiances using data on Hydrolight Standard Format HSF data files 1 Input a file containing wavelength and total sun sky irradiance HE52 will then call the built in RADTRANX sky irradiance model to partiti
82. g sz g ag Ly ae ar ae Fig 18 Top The TOMS monthly average ozone concentration for September as used in HE52 Bottom The measured ozone concentrations for a single day 23 Sept 2007 from http toms gsfc nasa gov ozone ozone html 39 Figure 19 shows the total direct sun and diffuse sky downwelling plane irradiances Eat the sea surface for clear sky typical open ocean atmospheric conditions the sun at a 45 deg zenith angle and a typical ozone concentration of 300 Dobson units RADTRAN X always outputs its irradiances at 1 nm resolution HE52 automatically averages those irradiances over the wavelength bands requested in the run Figure 20 shows the effect of varying ozone concentrations near 300 nm for the same general conditions as Fig 19 Although at 300 nm is an order of magnitude higher for low ozone 150 than for high 450 ozone significantly affects the irradiances only below about 330 nm For runs at 350 nm or greater the value of the ozone concentration used in the HE52 run is irrelevant The RADTRAN X clear sky direct and diffuse sky irradiances are modified as described in Kasten and Czeplak 1980 if the cloud fraction Cld is greater than 0 25 0 lt Cld lt 1 E total Cid E total clear 1 0 75 Cld and E diffuse Cld E total Cld 0 3 0 7Cld Here total E direct E diffuse Figure 21 shows the clear sky irradiances of Fig 19 along with the irradiances
83. ged by a negative depth or wavelength as seen below Since data are read one row at a time you need to flag the entire last row for each column An easy way to do this is just to repeat copy and paste the last row of valid data but then change the depth or wavelength to a negative number 52 The names of the files containing various data are given to HE52 by entering the file name at the appropriate location in the UI By default HE52 looks in the HE52 data directory for data files this occurs if you type in only the file name itself You may wish to place your data files in other directories in which case the UI lets you browse for the data file the full path name is then automatically passed on to HE52 The following sections show the standard formats for various types of data 7 1 Concentration Data Chlorophyll or mineral particle concentrations as used by the IOP models have data records that consist of pairs of depth and concentration values For example a file of measured chlorophyll data might look like the following Example standard format Chlorophyll data file To be read by HE52 subroutine chlzdata on file chlzdata f Ten header records for identification of the user s data are expected Record 11 begins with free format pairs of depth in m and Chl in mg Chl m 3 values The last value is flagged by either an end of file or by negative depth and Chl values Record 8 unused in this exampl
84. gm for use by IOP ab routines These values times the concentration of mineral particles in gm m 3 give the scattering coefficient of the mineral particles in 1 m These values are computed from Fig 4 3B and Table 3 II of Y H Ahn Proprietes optiques des particules biologiques et minerales Univ P amp M Curie Paris France 1990 The Ahn bb values Fig 4 3B are divided by bb b 0 0067 Table 3 II to get b Data were 400 700 nm WARNING Extrapolated by eye and splines to 300 and 1000 nm for use in H5 0 Spectrum may be unrealistic in the 300 400 and 700 1000 nm regions wavelen bmin nm m 2 gm 300 0 0 80000 305 0 0 80096 310 0 0 80194 315 0 0 80294 320 0 0 80395 and so on to 990 0 0 44298 995 0 0 44148 1000 0 0 44000 1 0 1 0000 Note that these mineral b values have units of m gm which is consistent with measuring the mineral concentration in units of gm m The product then gives a scattering coefficient in units of m as required by HE52 The end of data is flagged by a record with a negative wavelength 54 7 3 Absorption and Beam attenuation Data A common use of HE52 is to compute light fields using absorption a and beam attenuation c coefficients as measured by a WETLabs ac 9 or ac s or similar instrument as input via the MEASUREDIOP DATA option Note that HE52 cannot process your raw ac9 or ac s data Before giving the ac 9 or similar data to HE52 you must perform all of the st
85. h Note Routine abacbb reads in a and c and then computes a and b If you would rather convert a and c to a and b during your own data processing rewrite the abacbb routine slightly to read in a and b and remove the b c a conversion calculations Note It is customary when processing ac9 data to first subtract out pure water values from the measured raw a and c values HE52 therefore assumes that any file of a and c data has had pure water subtracted out of the tabulated values HE52 then automatically adds in the pure water values it computes the total absorption and scattering coefficients WARNING If HE52 needs values for a and b at depths or wavelengths outside the range of the original data then the nearest measured values will be used rather than terminating the run with an error message or using the splines to extrapolate which can cause large errors For example if the first wavelength of your ac 9 is 412 nm then the IOPs measured at 412 will be used for all wavelengths less than 412 nm Thus you can start a HE52 run at 350 nm even if your ac 9 data starts at 412 nm but you cannot reasonably expect to get accurate results at much less than 412 nm See 7 8 for more discussion of how HE52 fills in missing data 56 7 4 Backscatter Data The format for backscatter data measured for example by a HOBILabs HydroScat 6 or WETLabs bb 9 instrument is similar to that for a and c data except that the data records 12 and onwa
86. hase functions BAD to the available selection using the standard quad or band partitions in which case the calculations described in this section are needed Scattering phase functions analytically defined or tabulated as pairs of scattering angle and phase function cannot be used directly by HE52 This is because the HE52 code computes radiances averaged over quads or bands The code thus needs to compute how strongly radiance is scattered from one quad or band in EcoLight to another Exact scattering angles are thus integrated discretized over finite solid angles of incident and scattered directions within the quads or bands The phase function discretization calculations average the phase function over all pairs of quads pairs of bands for EcoLight as described in Light and Water 8 2 The main calculation for HydroLight is the evaluation of Eq 8 13 and a corresponding azimuthally averaged equation for EcoLight phase functions The HE52 code comes with many phase functions already discretized and ready for use These phase functions are defined by subroutines in the HE52 SpecialRuns PhaseFunction Hydrolight and 73 HE52 SpecialRuns PhaseFunction Ecolight directories The corresponding files of discretized phase functions are found in the HE52 data phasefun HydroLight and HE52 data phasefun EcoLight directories The discretized phase function files have names of the form dpf these files are all ASCII text files the dpf extension jus
87. he IOP subroutine must return values for any possible depth and wavelength that might be used in an HE52 run If you have data at discrete depths or wavelengths you must use an interpolation or extrapolation scheme in the IOP routine to define the a and b values at all other possible depths and wavelengths The HE52 core mathematical 67 routines call the IOP routine thousands of times and at very closely spaced depths not just at the depths were output is to be saved when solving the radiative transfer equation so doing this in a computationally efficient manner can have a significant impact on the runtime e HE52uses the absorption and scattering coefficients at the wavelength band centers not band averages Using a and b at the band centers is not generally a large source of error because a and b vary slowly and smoothly on a scale of 1 to 20 nm If your a and b do vary rapidly with wavelength then use smaller band widths for the run 8 2 Subroutines for Chlorophyll Profiles It is often convenient in the IOP models just discussed to be able to call a subroutine that returns the chlorophyll concentration Chl in mg Chl m given the depth in meters The chlorophyll concentration then can be used as input to bio optical models for a and b as is done in the subroutine on file abcase1 f for example A chlorophyll subroutine will always be called to obtain Ch if chlorophyll fluorescence is included in the HES2 run Two ways of selecting
88. he new routines to be utilized within HE52 in a straightforward manner The following sections give a few additional words of advice on writing your own routines User written subroutines for IOP models sky radiance models etc MUST be placed in the HE52 code user directory This helps keep user code separate from the distributed code The HE52 UI will detect the presence of new routines in the user directory or any changes to existing routines in the HE52 code Hydrolight and HE52 code Ecolight directories recompile the routines as necessary and create new executables However at least make sure your routines compile correctly before trying to run them in HE52 This avoids confusing HE52 bugs with bugs in your own code This is easily done by opening a command window going to the directory containing the new or 64 modified routine and typing the command LF95 c filename f where filename f is the name of your new routine or of the one you modified If there are error messages continue debugging your code until it compiles correctly Ifthe compilation is successful then you can run HE52 and begin the process of figuring out if your new code is actually doing what you want it to do Section 10 gives additional advice on debugging new code and tracking down errors that cause run termination before error messages can be written 8 1 Subroutines for IOP Models The most common and important task in tailoring HE52 to a specific water bod
89. he phase function in the absence of other information about the phase function However it must be remembered that B correlates poorly with Chl and there can be order of magnitude variability in the measured value of B for a given Chl value in a particular data set The particle backscatter fraction can chosen to be the same at every wavelength or be a function of wavelength according to the power law n Ao BO B J Values for B A and n are entered in the User Interface If it is desired to have the particle backscatter fraction be a function of both depth and wavelength then the IOP data model described in 2 6 must be used to read in separate files of b z A and b z A such that B Z A b z A b z A has the desired values Note that the absorption model of Eq 1 reduces to that of pure water if Chl 0 which is slightly different than some formulations e g Morel and Maritorena 2001 Eq 16 18 L amp W Eq 3 27 which include a small amount of background CDOM absorption even in the absence of phytoplankton The inclusion of a small amount of background CDOM is reasonable but in all honesty we just got tired of explaining to HydroLight users why they didn t get exactly the same results as for pure water when they plugged in Chl 0 in the Case 1 IOP model In any case the difference is negligible except at extremely low Chl values in which case the model is suspect anyway This IOP model is retained in H
90. hen the optical properties of the surface will be accurately modeled in most situations The exception may occur if the sun or viewing direction is near the horizon when wave shadowing by large gravity waves may be significant It should be noted that the Monte Carlo ray tracing that computes the surface reflectances and transmittances exactly conserves energy as light passes through the sea surface However conservation of energy across a wind blown sea surface has some radiometric subtleties when expressed in terms of radiance or irradiance These matters are discussed in HydroLight Technical Note 7 found in the HE52 documents directory The sea surface slope depends on the wind speed The Fresnel reflectance of an air water surface is determined by the real index of refraction n of the water which in turn depends on the water temperature and salinity The index of refraction varies over visible wavelengths 400 700 nm by only a few percent about a value of n 1 34 This value was therefore 35 used to generate the sea surface reflectances and transmittances in previous versions of HydroLight Those sea surface reflectances and transmittances were then functions only of wind speed However the dependence of n on wavelength temperature and salinity becomes greater at UV and IR wavelengths as seen in Fig 17 HE52 therefore models the sea surface reflectance and transmittance properties as functions of wind speed temperature salini
91. his script runs both of the executables discpf_EL exe and discpf_HL exe which perform the calculations to create the new discretized phase function file for both EcoLight and HydroLight The new dpf files are placed in the HE52 data phasefun Ecolight and HE52 data phasefun Hydrolight directories The printout from these two runs will be placed in the Printout _EL txt and Printout_HL txt files in the PhaseFunction directory Inspect the printouts with a text editor to verify that the discretization run terminated normally Also check the HE52 data phasefun Ecolight and HE52 data phasefun Hydrolight directories to see if the new dpf files are there Finally use a text editor to check file HE52 data phasefun filelist txt to make sure that the name of your new file was added to the bottom of the list of available dpf files If everything is in order the new phase functions are now available for standard HE52 runs Note 1 The HydroLight discretization program adds the file name e g Sta4_550 dpf to the end of the file named filelist txt in the HE52 data phasefun directory This filelist txt file is read by the UI to generate the pulldown menu of available discretized phase functions The files are shown in the UI in the order listed in filelist txt You can edit this file to change the order of the files in the menu or to remove files from the available list Note 2 The discretization programs will not overwrite an existing dpf file The
92. how to write your own subroutines for various models The HE52 source code is written almost entirely in ANSI standard FORTRAN 77 in order to make it easily portable to almost any computer with a FORTRAN 95 compiler some new routines in HE52 use FORTRAN 95 All input and output files are written as ASCII text files to assure easy transfer of files between computers with different operating systems Writing new subroutines for an IOP model for example does require knowledge of FORTRAN However HE52 comes with templates that show you how to fill in the blanks to create IOP and other models so this process is relatively easy even for inexperienced programmers The built in IOP models for example are just specific implementations of the template for ab models which is file HE52 examples templates ab f However writing an IOP model can be complicated if there are many components and different models or data files for specifying the absorption and scattering by those components The best way to learn how to write a IOP or other subroutine is to compare the appropriate template file all of which are on the HE52 examples templates directory with a corresponding existing subroutine such as HE52 code common abnewcase 1 f Making the effort to write new routines in the same style as the templates will help you to develop routines that best mesh with existing formats and methods which will significantly simplify any code changes and allow t
93. ibution to the total light field and has an almost negligible effect on quantities such as the water leaving radiance An important additional feature of the Measured IOP data model is the option of reading a file containing backscatter coefficients b z A as measured by a WETLabs bb 9 HOBILabs HydroScat 6 or similar instrument If such data are available HE52 can use the backscatter fraction B z A b z A b z A to generate a phase function having the measured backscatter fraction at each depth and wavelength Here the b z A values are those obtained from the ac 9 data If this option is used the user does not select a phase function for component 2 from within the UI When processing a and c data from an ac 9 or ac s the convention is to remove the contribution by pure water to the measured total HE52 therefore assumes that standard format files of a and c data have water removed The chosen pure water values are then automatically added back in to create the total IOPs needed to solve the RTE However there is no convention on removing pure water backscatter values when processing bb 9 or HydroScat 6 data When reading a standard format file of backscatter data the user must indicate on the UI whether the b data do or do not include backscatter by pure water When b z and b z data are used to dynamically determine the phase function according to the backscatter fraction a Fournier Forand Fournier and Jonasz 1999 c f Fourn
94. icle absorption coefficients computed by the new Case 1 model of Eq 6 red and the classic Case 1 model of Eq 2 blue The red solid line is for the high UV absorption dotted is the mid range UV absorption and dashed is the low range UV absorption The purple line is absorption by pure water 2 4 2 CDOM absorption Figure 7 shows measured CDOM absorption spectra down to 200 nm at several locations in Florida waters on both log and linear ordinates data of Lore Ayoub personal communication For wavelengths greater than 300 nm these spectra are acceptably well modeled by a function of the form of Eq 3 a Z A aZ Ap expl S A 7 The red line in Fig 7 shows the spectrum predicted by Eq 7 with A 440 nm S 0 0162 nm and the average for the spectra shown of a 440 This value of S was determined from the average values of a at 300 and 440 nm The functional form 7 is used to model CDOM absorption down to 300 nm as needed for HE52 This model underestimates CDOM absorption at shorter wavelengths but that is irrelevant for HE52 When incorporated into the new Case 1 IOP model a z A is set to f a z A with default values 14 of f 0 2 A 440 nm and S 0 014 nm just as in Eq 3 for the classic Case 1 model However the user can change the values of f and S in the HE52 UI if other values are desired log a_CDOM 1 m 1 m a_CDOM Wavelength nm Fig 7 Me
95. ier and Forand 1994 phase function is used This closed form analytical phase function is based on Mie theory and is parameterized by the real index of refraction of the particles and the slope of the Junge size distribution By varying the particle index of refraction and size distribution phase functions can be generated with very small to very large backscatter fractions A collection of discretized Fournier Forand phase functions with backscatter fractions from 0 0001 to 0 5 is built into HE52 Figure 11 shows selected phase functions from this collection When the option to dynamically determine the phase function from the backscatter fraction is chosen HE52 interpolates within this collection to generate a phase function at each depth and wavelength with the needed amount of backscatter This interpolation requires very little increase in run time compared to using the same phase function for all depths and wavelengths 22 b b 0 4 0 1 0 0183 0 01 phase function sr th oT Tt t 0 001 t 0 0001 1 f poi il 0 1 1 0 10 0 100 0 scattering angle deg pure water phase function sr jai v T 1 1 1 1 1 1 1 1 1 1 1 T T T T 0 30 60 90 120 150 180 scattering angle deg Figure 11 Selected Fournier Forand phase functions red lines The black line labeled b b 0 0183 is the Petzold average particle phase function on file HE52 data phasefun avgpart dpf
96. include salinity and temperature effects so this feature was added to version 5 36 5 Sky Models HE52 must know the sky radiance incident onto the sea surface from all directions and for all wavelengths included in the HES2 run to be made To increase flexibility in using modeled or measured sky radiances this information is provided by two independent subroutines The first subroutine returns the direct and diffuse components of the sky downwelling plane irradiance These irradiances are used to set the magnitude of the sky radiance The second subroutine returns the angular pattern of the sky radiance distribution This radiance pattern is integrated over direction to compute a sky irradiance which is then forced to equal the irradiance given by the irradiance subroutine Together these two routines yield a sky radiance distribution with the desired magnitude in all directions HE52 comes with two versions of each routine as shown in Table 2 which we now discuss filename type user interface option comments radtranx f irradiances SEMI EMPIRICAL SKY MODEL henrad f radiance SEMI EMPIRICAL SKY MODEL cosirrad f irradiances IDEALIZED SKY MODEL cosrad f radiances IDEALIZED SKY MODEL Computes irradiances using RADTRAN X Recommended for general use Computes the normalized radiance pattern using Harrison and Coombes 1988 Recommended for general use Uses user supplied input from the UI
97. is record gives information needed by whichever sky radiance model is being used in the run The general form of the record is Record 8a sky model parameters Names of variables iflagsky nsky skydata 1 skydata 2 skydata nsky iflagsky 1 if the idealized sky models are being used 2 if the semi analytic sky model is being used with solar zenith angle being specified 3 if the semi analytic sky model is being used with time and location being specified nsky is the number of values to be read in the remainder of the record The format of the record depends on which sky model is being used If iflagsky 1 then nsky 5 and skydata 1 to skydata 5 contain the following Names of variables suntheta sunphi C rsky Edtotal Example 30 0 0 0 1 25 0 3 1 2 where suntheta is the solar zenith angle in degrees suntheta 0 0 for the sun at the zenith and suntheta 90 0 for the sun at the horizon sunphi is the solar azimuthal angle in degrees relative to the wind direction sunphi 0 0 is downwind and sunphi 90 0 places the sun at a right angle to the wind The default is sunphi 0 0 see Note 2 below C is the cardioidal parameter C in Light and Water Eq 4 50 C 0 0 gives a uniform background sky and C 2 0 gives a cardioidal sky C 1 25 gives a heavily overcast sky rsky is the ratio of background sky to total scalar irradiance 0 0 lt rsky lt 1 0 rsky 0 0 for a black sky sun only and r
98. lastic sources are present If iDynZ 1 inelastic sources are present and an infintely deep bottom is selected HE52 will compute the light field to a greater depth roughly 20 optical depths at the clearest wavelength This helps guarantee a good solution at all depths as described in 3 3 RamanExp is the wavelength dependence of the Raman scattering coefficient on the excitation wavelength as described in HydroLight Technical Note 10 RECORD 2 Run Title This record gives the descriptive title for the run The title can be up to 120 characters long Name of variable ititle Example Users Guide Example 1 RECORD 3 Rootname This record gives the root name to be used for file generation The name can be up to 32 characters and must not contain blanks or other special characters Name of variable rootname Example ugex1 RECORD GROUP 4 Options Record 4a output options This record specifies the output options Name of variables iOptPrnt iOptDigital iOptExcelS iOptExcelM iOptRad nwskip 88 Example 0 1 2 1 1 1 iOptPrnt is a flag for the specification of the amount of printout if OptPrnt 1 Proot txt will contain only minimal runtime output if OptPrnt 0 Proot txt will contain the standard output including component IOPs and selected radiances and irradiances if OptPrnt 1 Proot txt will contain extensive output including all radiance arrays and some intermediate v
99. line tells how the specific absorption line 1 and specific scattering line 2 will be given for each component Names of variables ibstropt i bstarRef i bstar0 i coefl coef2 coef3 Example 1 550 0 3 1 0 62 999 The b specifications are ibstropt 0 user supplied data file read to get b values 1 Power Law used Chlorophyll only 2 Linear GAM model Chlorophyll only 3 Gordon Morel model used Chlorophyll only If ibstropt 1 or 2 up to five more parameters are used if needed otherwise 999 bstarRef reference wavelength for model bstar0 b at reference wavelength Ifibstropt 1 coefl m for power law coef2 n of the power law 92 If ibstropt 2 coefl m of GAM model the slope of the linear model and coef2 i of GAM the offset value in the linear model coef3 n of GAM the exponent for the concentration Record 5f Specific scattering data file names The next ncomp lines of input give the names of the files containing the specific scattering for each concentration option The order MUST match the order of the components in the subroutine for a and b see 8 1 File names that are not needed for the run components for which ibbopt 0 may be stored with the name bstarDummy txt as a place holder Names of variable bstarfile i Example bstarchl txt bstarDummy txt Record 5g type of concentrations and phase functions This record consists of ncomp lines of input where each line r
100. m a command window in which case the window remains open so that any error messages can be read Finally some users perform massive simulations under control of their own code for automatic generation of the needed input files in which case the UI is not used to open the command window and run HE52 Thus it is sometimes necessary to manually perform a run by explicitly issuing commands in an already open command window or by double clicking on the appropriate file when viewed by Windows Explorer To describe how this is done it is assumed that e The run input file lroot txt has been created and placed in the HE52 run batch directory e The name of the lroot txt file has been placed in the HES2 run runlist txt file The runlist txt file contains the names of the one or more Iroot txt files that define the run or runs to be made Multiple runs defined by different input files can be made by listing the input files in the runlist txt file This is what the UI does when multiple passes are made through it via the ADD ANOTHER BATCH RUN option on the final UI form The Iroot txt file may have been created by the UI by a text editor or by the user s own program that mimics the output of the HE52 UI There are then several ways to make a run as follows 81 Option 1 The easiest way to make a run is then to open Windows Explorer to the HE52 run directory and click on either runHL exe to make a HydroLight run or runEL exe to make an E
101. n observations of the ocean surface and therefore include the effects of whatever gravity and capillary waves where on the sea surface at the time of their observations The Cox Munk statistics are sometimes thought of as describing the slopes of capillary waves In some situations this viewpoint is reasonable because much of the variance in the slope of the sea surface is indeed due to the smallest capillary waves The larger gravity waves are responsible for most of the variance of the sea surface elevation but they contribute less to the variance of the surface slope These matters are discussed in more detail in HydroLight Technical Note 1 which can be found in the HE52 documents directory Hydrolight uses Monte Carlo ray tracing of millions of rays or photon packets through tens of thousands of randomly generated sea surface realizations to compute the radiance reflectance and transmittance functions that describe the optical properties of the sea surface see L amp W 4 7 When doing Monte Carlo simulations of rays interacting with the sea surface it is the slope of the surface at the point where a ray intersects it that determines the directions of the reflected and transmitted rays The magnitudes of the transmitted and reflected rays are determined by the associated Fresnel reflectance and transmittance which are determined by the real index of refraction of the water Thus if the slope statistics and water index of refraction are correct t
102. n on sea surface total irradiances 41 1 20E 00 1 00E 00 diffuse clear sky E 8 00E 01 direct clearsky z total clear sky diffuse cid 50 400E 01 direct cid 50 a total cid 50 2 00E 01 0 00E 00 300 400 500 600 700 800 900 1000 wavelength nm Fig 21 Example sky irradiances for a clear sky and a 50 cloud cover 5 2 The Empirical Sky Radiance Model The sky radiance distribution not just the sky irradiance must be known in order to solve the RTE Thus it is necessary to define the directional pattern of the sky radiance distribution Several semianalytic models for the angular pattern of the sky radiance distribution can be found in the literature One is probably as good as the other for most HE52 applications because errors in the IOPs or other inputs used in a run are usually far more important than small errors in the sky radiance distribution and because the AOPs are generally very insensitive to the pattern of the sky radiance distribution However some quantities such as the surface reflected radiance sea surface glitter pattern can be strongly dependent on the details of the incident radiance HE52 uses the Harrison and Coombes 1988 semianalytic clear sky model to define the relative pattern of the incident sky radiance Figure 22 shows the angular pattern of the clear sky radiance distribution as generated by their model for a 30 deg solar zenith angle
103. n the list of depths where output is requested in the UI The default is to assume that the physical bottom is an opaque Lambertian reflecting surface whose irradiance reflectance R E E is specified by the user This irradiance reflectance is automatically combined with a Lambertian bi directional reflectance distribution function BRDF to generate the needed radiance reflectance properties of the bottom Equation 4 81 of Light and Water gives the needed bi directional radiance reflectance BRRF of the physical bottom the BRRF equals the BRDF times the cosine of the incident polar angle Data files containing irradiance reflectances for various sediments and benthic biota are provided with HE52 These files include are based on measurements from 400 to 750 or 800 nm which were extrapolated by eye to 300 and 1000 nm for use in HE52 Thus the reflectances may be unrealistic near 300 and 1000 nm These reflectance spectra are shown in Fig 15 As with all other such data sets these spectra are provided with HE52 as examples of bottom reflectances they may or may not be adequate for modeling your water body Section 7 5 describes how to add your own data files to the list of available bottom reflectance spectra The source code for the default Lambertian bottom BRDF is contained on file HE52 code Hydrolight BRDFLamb f The assumption ofa Lambertian bottom is justified for most simulations However HydroLight can simulate non Lambertian bottom
104. nd 2 no one has needed high resolution radiances and the irradiances are the same for both HydroLight and EcoLight Finally the 2 deg files described here have not been distributed with code versions through 5 1 4 because very few users want them and even fewer actually need them However they are available on request 80 10 WHEN SOMETHING GOES WRONG HE52 standard runs are usually made by using the user interface to specify the run conditions and to automatically execute the run including if needed compilation of new user written routines Normally this works well The HE52 UI opens a command window and starts the run the code runs to completion and window then closes after execution and the UI displays a message that the run has completed However when developing new code it is almost always the case that things don t work correctly on the first try If the new code doesn t compile or execute correctly the run can terminate before any output and error messages are written to the printout the Proot txt file The command window then usually closes automatically a particularly annoying feature of the Windows operating system before any error messages can be read This can also happen if an input user data file is incorrectly formatted and causes an error in reading or processing the data which causes the run to terminate abnormally In these situations it is necessary to bypass the UI and run HE52 manually fro
105. nd window will remain open at the end of the run so that any error messages can be read 82 Option 4 A set of commands equivalent to Option 3 is e Open a command window and cd to the HE52 code directory e Type the command mainHL stnd exe lt run batch Iroot txt at the command prompt and hit return This again pipes the input file to the HydroLight executable and makes the run Option 3 or 4 should be used when debugging new code such as user written IOP subroutines or trying to determine why a run terminated abnormally perhaps without an error message in the printout file Finally it is sometimes wise to force a recompilation of all code just to make sure everything is up to date and compatible This can happen if old and new codes from different versions of HE52 are mixed in which case the need for recompilation may not be recognized by the UI To manually recompile all HydroLight code e Open a command window and cd to the HE52 code Hydrolight directory e Type the command makeExeStnd HL bat at the command prompt and hit enter The same process is used to force recompilation of all EcoLight code via the command makeExeStnd EL bat in the HE52 code Hydrolight directory The actual compilation commands issued to the Lahey Fortran 95 compiler in order to compile the HydroLight code and generate mainHL_stnd exe are found in HE52 code Hydrolight makeExeStnd bat These commands are as follows 83 del obj copy
106. nding on whether the value needed falls within or outside of the range of 60 measurements For example suppose you have ac9 measurements at 9 wavelengths from 412 440 715 nm If HES2 needs an IOP at 420 nm then it will interpolate between the measured values at 412 and 440 However if it needs a value at 390 outside the range of measurements then it will extrapolate below 412 by simply using the value at 412 7 8 1 Interpolation within the measurement range In the original release of HydroLight version 4 0 cubic splines were used to interpolate in single variable data e g chlorophyll as a function of depth or reflectance as a function of wavelength and bi cubic splines were used to interpolate in depth and wavelength e g with IOP data However some users came to grief with this interpolation scheme because their data were not sufficiently smooth for the successful use of cubic splines To use cubic spline fitting the data must be continuous in at least the first derivative The essence of the problem is illustrated in Fig 27 which shows linear and cubic spline fits to hypothetical depth vs concentration data In the left panel of Fig 23 the data vary relatively smoothly with depth and there is not much difference in the linear and cubic spline fits However in the right panel Fig 27 the data are noisy and there is a sharp change in concentration near a depth of 15 Now the spline fit shows large overshoots
107. ne the extrapolation in a way that is consistent with the know characteristics of the data That is not the case in this example About the only time such a naive extrapolation can be justified in HES2 is if a run needs to be started at 360 nm say in order to approximately include the effect of Raman scatter or CDOM fluorescence on the light field from 400 nm onward and then only the solution from 400 onward will be considered as valid Similarly if your IOP data start at a depth of 5 m and stop at 22 m for example then the values at 5 m will be used from the surface down to 5 m and the values at 22 m will be used at all depths greater than 22 m Again that may or may not give good results depending on how homogeneous the water is It must be kept in mind that if you want to run HES2 using measured data then the measurements should cover the range of depths and wavelengths of interest for the simulation 63 8 WRITING SUBROUTINES FOR IOP AND OTHER MODELS As seen previously HE52 comes with many built in subroutines for IOP models sky models etc You can use these models as they are and most users will find these built in models to be adequate However some users will want to write IOP models of their own either to read in data on a different format or to use different bio geo optical models to convert measured chlorophyll or mineral profiles into the absorption and scattering coefficients required by HE52 This section describes
108. nt However the ac s instrument measures 24 the same number of wavelengths for both a and c but the exact wavelengths for each can differ by a few nanometers Therefore when working with ac s data the user should spline the a and c spectra to a common set of wavelengths say at 5 nm resolution and use those data to create a standard format data file for input to HES2 2 7 USER IOP MODELS This option on the UI gives users convenient access to their own IOP models from within the HE52 UI If this IOP model is chosen on the UI the user is asked to enter the name of the subroutine containing the user s IOP model and to enter a name for each component of the user s IOP model These component names are then used by the UI to identify the components whose IOPs can be specified in a manner similar to that for the case2 model Note that unlike the other IOP models there is no corresponding ab file in the HE52 code common directory because you the user must write the desired subroutine defining your IOP model Section 8 1 describes how to write IOP subroutines which can then be used with the other UI option Note that user written IOP and any other user written subroutines must be placed in the HE52 code user directory The HE52 code as distributed contains an example of a user written IOP model on file TenComplOPs f in the HE52 code user directory The model builds up the total IOPs from 10 components which in the original application we
109. of the bottom will be read from a HE52 standard format file of bottom reflectance data see 7 5 A value of rflbot is always read but is used only if ibotm 1 RECORD 11 Output Depths This record gives the depths at which output is to be saved for post run analysis The depths as read can be either dimensionless optical depths or geometric depths z in meters The last depth specified is taken to be the maximum depth of interest Or Zmax Where the bottom boundary condition will be applied The water below the last depth is assumed to be homogeneous if ibotm 0 If ibotm gt 1 the Lambertian bottom is placed at the last output depth The record has the form Names of variables iop nznom zetanom 1 zetanom 2 zetanom nznom Example 0 5 0 0 1 0 2 0 5 0 10 0 iop is a flag for optical or geometric depth as follows if iop 0 then the zetanom values are GEOMETRIC depths in meters 99 if iop 1 then the zetanom values are OPTICAL depths nznom is the number of depths where output is desired zetanom 1 0 0 zetanom nznom Z are the depths where output is desired Note that if the run contains inelastic scattering then the run must use geometric depth Note also that zetanom 1 must always be 0 0 the depth in the water just below the mean air water surface and that zetanom nznom is by definition the maximum depth of interest Recall from the discussion of Section 7 1 Note 1
110. on this total into direct sun and diffuse sky irradiances as needed by the RTE solution code File Sky_Irrad_Example_Edtotal txt in the HE52 data examples directory shows the required HSF format for such files When using this option the sun angle and atmospheric conditions used by RADTRANX should match as closely as possible to the conditions for the user s irradiances so that the partition into direct and diffuse parts will be as accurate as possible 44 2 2a Input a file containing wavelength direct sun and diffuse sky irradiances File Sky_Irrad_Example_Eddir_Eddif txt is an example With this option the sun zenith angle entered in the user interface UT is used to determine where the sun whose direct beam generates the user s direct irradiance values should be placed and the default Harrison and Coombs sky radiance model is used to determine the angular distribution of the radiance corresponding to the user s diffuse sky irradiances However RADTRANX will NOT be called so the atmospheric information entered in the user interface UI will not be used A special case of option 2 enables the user to Input a lidar irradiance at a single wavelength and have the sky irradiance at all other wavelengths be set to zero This simulates a lidar input in an otherwise black sky This option is discussed in more detail below File Sky_Irrad_Example_Lidar488 txt is an example With this option RADTRANX will NOT be called so
111. onent 1 be pure water component 2 be CDOM component 65 3 be pigmented particles component 4 be mineral particles and component 5 be microbubbles The order of the components is not important in building up the absorption and scattering coefficients but it is very important when you later specify when running the UI the component input e g which data file contains the concentration profile for each component or which scattering phase function is to be used for each component If component 1 is chosen to be pure water then you must select the phase function for pure water for component 1 if component 4 is mineral particles then you must select a phase function for component 4 that is appropriate for mineral particles and so on Decide how you are going to compute the absorption and scattering coefficients for each of your components For example you can call the HE52 routine pureH2o to obtain the pure water a and b values You might want to use a simple formula to model the CDOM absorption coefficient and assume that CDOM is non scattering You might want to call HE52 subroutine chlzdata to read in a file of your chlorophyll data and then use a bio optical model to convert the chlorophyll value at depth z into the component a and b values atz and You might want to read your own data file of mineral particle a and b values Such decisions are yours to make Modify one of the existing IOP subroutines or fill in the ab f templat
112. or example the concentration specific absorption and scattering spectra found in HE52 data defaults are defined only for 300 1000 nm and even then the values near 300 and 1000 are often just educated guesses because very few measurements of such things have been made outside 350 to 750 nm Thus the user must first update any needed IOP or bottom reflectance files to include values for the wavelengths of the desired run The default Harrison and Coombes model for the angular pattern of the sky radiance does not depend on wavelength and thus becomes more and more inaccurate outside the visible wavelengths for which is was developed Note that the sky radiance pattern is effected by Rayleigh scattering which depends on the fourth power of the wavelength Thus sky radiance patterns can be significantly different at near UV wavelengths compared to near IR wavelengths These differences should have little effect on computed AOPs but may become significant when computing radiances at wavelengths far from the visible The index of refraction varies with wavelength so the option to use a wavelength dependent index should be chosen in the UI This should have almost no effect on computed AOPs In any case if doing a run outside of 300 1000 nm you must manually change the default wavelength limits in file HE52 frontend HFEdflts txt to allow the UI to run outside its standard 300 1000 nm range This change is made to line 3 of that file which in
113. oup website and supplement the information in this Technical Documentation The HTNs issued to date are o o UUO e a BRDFs are used in HydroLight and Monte Carlo codes a ama HTN3 Sept2002 Advice on choosing output depths HTN4 Dec 2002 The RADTRAN X atmospheric model Jan 2007 How H spectral outputs are converted to visual color as CIE chromaticity coordinates and RGB values HTN6 May 2008 Advice on selecting bottom depth in runs with inelastic scattering Note this advice has now been automated in the code via the dynamic lower boundary depth option so this note is now obsolete Aug 2008 Conservation of energy across the sea surface in H Oct 2010 How H computes Forel Ule color indices HTN9 Dec 2010 New options for input of user defined sky irradiances begun in version 5 1 2 Note these options are now discussed in Section 5 4 of this Tech Doc so this HTN is now obsolete HTN10 Jan 2012 Interpretation of Raman scatter output Table 1 HydroLight Technical Notes issued through October 2013 2 IOP MODELS HE52 comes with several built in models for describing water inherent optical properties IOPs namely the water absorption coefficient a and scattering coefficient b as functions of depth and wavelength and the scattering phase function which depends on scattering angle and possibly also on depth and wavelength These models are accessible from within the graphical user interface UI and
114. r index index of refraction of water 1 34 by default rad48 0 842439 asin 1 refr critical angle for internal reflection by a level surface relerrs relative error for ODE integration abserrs absolute error for ODE integration areset absorption coefficient or flag for routine abconst breset scattering coefficient or flag for routine abconst deltazK the depth increment for use in generating closely spaced pairs of depths for K function computations acoef nz the absorption coefficient at the deepest depth bcoef nz the scattering coefficient at the deepest depth wavel current nominal wavelength in nm 104 w FETTE the bottom reflectance 14 rflbot 15 wndspeed the wind speed in m s 17 epsilon parameter in L amp W Eq 4 39 19 atotal total absorption coef the the current depth and wavelength 21 aOR Raman scattering coefficient Raman scattering reference wavelength 23 PhiChl chlorophyll fluorescence quantum efficiency 24 delbbTol tolerance for changine phase function according to b b value 25 22 wave used only by EcoLight for dynamic depth optimization min wavelength for PAR calculation 27 PARmax max wavelength for PAR calculation salinity in PSU for use in computing index of refraction temp temperature in deg C for use in computing index of refraction refr index of refraction B 3 The DATAFILES Array 26 PARmin salinity jii
115. r called by other IOP models constant 1 the total a and b values returns depth independent a and b for a single wavelength Classic Case 1 1 pure water Gordon Morel Case 1 water model 2 particles 3 co varying CDOM requires the chlorophyll concentration and particle phase function as input This is the Case 1 model found in HydroLight version 4 1 pure water a NEW Case 1 water model based on 2 small chlorophyll bearing recent bio optical models for absorption particles and scattering requires only the 3 large chlorophyll bearing chlorophyll concentration as input particles Recommended as a replacement for the gt co varying CDOM classic Case 1 model of version 4 a generic 4 component IOP model with pure water 1 2 chlorophyll bearing particles 3 CDOM 4 mineral particles great flexibility in how the user can define component optical properties Recommended for general use Measured IOP data 1 pure water Reads standard format see 7 3 data files to obtain a and b Optionally an 2 everything else e g as measured by an unfiltered ac 9 additional data file can be read to obtain the backscatter coefficient b for use in determining the phase function User written any number of components Allows the user to access user written default 4 can be increased inthe IOP subroutines from within the HE52
116. r absorption by chlorophyll IfCDOM is included in the run CDOM absorption will be specified by the options selected in the UI Note that the CDOM subroutine must return the CDOM absorption value for a depths and wavelengths relevant to the HES2 run 8 4 Subroutines for Bottom BRDFs For finite depth water the both HydroLight and EcoLight assume that the bottom is an opaque Lambertian reflecting surface which is fully described by specifying the irradiance reflectance of the bottom However HydroLight has the option of using a non Lambertian bottom The default Lambertian bottom for HydroLight is found on file HE52 code Hydrolight BRDFLamb f HydroLight by default calls this subroutine If you wish to use a non Lambertian BRDF for the bottom boundary the procedure is as simple e Write a function subroutine defining the desired BRDF An example of such a subroutine which defines the Minnaert BRDF is given on file HE52 examples templates BRDFMinn f The function should be named BRDFbotm just as is seen in files BRDFLamb f and BRDFMinn f e Move the file BRDFLamb f from the HE52 code Hydrolight directory to some other directory for safekeeping Note to return later to using the Lambertian form you will need to reinstall the original BRDFLamb f file from the safekeeping directory or from the HE52 installation CD and add your file e g add file BRDFMinn f to the HE52 code Hydrolight directory The next time the HE52 UI runs it will
117. rd now contain the depth and backscatter coefficient values rather than a and c values See file HE52 data examples HydroScat6 txt for an example data file The HE52 data processing e g interpolating or extrapolating in depth and wavelength is similar to that just described for ac 9 data Note There does not seem to be any convention on removing pure water values when processing HydroScat 6 bb 9 or similar backscatter data because the scattering contribution by pure water is usually a small contribution to the total except in very clean water The UI therefore requires you to indicate whether or not your file of backscatter data has had pure water backscatter removed HE52 will then know whether or not to subtract pure water backscatter from your data in order to determine the particle backscatter fraction for use in defining a Fournier Forand phase function 7 5 Bottom Reflectance Data As explained in 3 1 for finite depth water a physical bottom is placed at depth z Light and Water Eq 4 81 then gives the needed bi directional radiance reflectance of the physical bottom The irradiance reflectance for a physical bottom is most easily communicated to HE52 by placing measured wavelength dependent reflectances ona file in the HE52 standard format for bottom reflectances and then giving the file name to the UI The format for reflectance data is the same as for concentration specific data except that the data records are now pair
118. re pure water CDOM minerals and 7 phytoplankton species This model can be used as is or can serve as a template for modifications to define other IOPs as needed by a particular user Note that when running any IOP model with more than 4 components the number needed for the Case 2 IOP model the maximum number of IOP components must be increased on the UI Change Limits form 2 8 Specific Absorption Models Some of the IOP models require a chlorophyll or mineral particle concentration to be converted to absorption and scattering coefficients Such conversions are made by equations of the form aA aN X 2 The absorption coefficient a is always in units of m a is the chlorophyll or mass specific absorption coefficient and X is the component concentration For chlorophyll a is in m mg Chl and X is the chlorophyll concentration in mg Chl m For mineral 25 particles a is in m g and X is the mineral particle concentration in g m HE52 uses these default units because they are what is standard in the literature when expressing concentrations of chlorophyll and mineral particles HE52 comes with several examples of component specific absorption and scattering spectra The user s a and X values can however be in any units so long as the product of mass specific absorption times concentration has units of inverse meters e g a in m ug Chl and X ug Chl m The user can easily add additional spectra Th
119. red these spectra in the laboratory for 400 750 nm These curves were extrapolated by eye to 300 and 1000 nm for use in HE52 The spectra are therefore very uncertain and perhaps simply wrong near 300 and 1000 nm These specific absorption curves are shown in Fig 13 As noted above the mineral particle mass specific absorption spectra have units of m g thus the mineral concentrations entered via the UI must be specified in g m These data files are all provided with HE52 as examples of what is available in the literature no statement is made about their accuracy or applicability to your particular water body Users wishing to use their own chlorophyll or mineral specific absorption spectra can create files on the format described in 7 2 and then select those files from within the UI I 2 mera oT 0 25 ae D 3 solid is measured 2 0 20 fs dashed is extrapolated Q pA calcareous sand to ES Ilow cl ellow cla Ex O15 y y ae red clay oO Ls N oO A brown earth 56 0 10 Raa average Qf k T Q Re 6 0 05 F c QO ro 0 USSSA 2 Oban aaa 300 400 500 600 700 800 900 1000 wavelength nm Figure 13 Mineral mass specific absorption spectra as provided with HE52 CDOM absorption is usually modeled as an exponentially decaying function of wavelength Acpom Z acnom E exp y A Aol 27 The model parameters and y can be specified in the UI Various options are availa
120. refore your new phase function must have a unique name If you want to replace an existing dpf file delete it before running the discretization program Note 3 The discretization programs set several parameters defining the numerical accuracy of the phase function discretization The default values are adequate for most phase functions and will not need to be changed by most users Note 4 The numbers in the HydroLight dpf files are discretized phase functions denoted by B r 1 u v in Light and Water Chapter 8 see Eq 8 13 stored as a 2D array These 77 numbers bear no obvious resemblance to the phase function values themselves so don t waste your time trying to figure out what the individual numbers are Likewise the numbers in the files for EcoLight are band averaged values which are not easily interpreted Note 5 As previously noted in 9 1 the discretization calculations depend on the quad and band layout If you must use a quad or band layout other than the defaults shown in Fig 24 you must re create all of the files of discretized phase functions and all of the surface files The phase functions must be rediscretized after creating the new surface files because the phase function discretization routines read the quad and band layout information from the surface wind files Note 6 It is not currently possible to discretize wavelength dependent phase functions and then have HE52 automatically select a different
121. resh water is L amp W Eq 3 30 BAD 0 06225 1 cos y 10 10 f Smith amp Baker abs Pope amp Fry abs 10 10 10 sea water scat fresh water scat abs or scat coef m 10 10 300 400 500 600 700 800 900 1000 wavelength nm Fig 1 The pure water IOPs available for selection in the HE52 UI The pure water IOP model is not an explicit option on the UI because it is of no interest to most users However simulations for pure water can be obtained simply by using the classic Case 1 IOP model and using a chlorophyll concentration of 0 The same pure water IOPs are accessed within each IOP model by calls to subroutine HE52 code common abpure f which reads the chosen pure water data file and can be used for the same purpose by user written IOP subroutines 2 2 CONSTANT This IOP model simply gives the total including water a and b values at a single wavelength for a homogeneous water body It is designed for idealized radiative transfer studies and pedagogical purposes rather than for modeling actual water bodies This is the only IOP model that allows the depth to be measured as non dimensional optical depth C rather than as geometric depth z in meters The UI requires the user to pick the total scattering phase function Because constant is for runs at a single user selected wavelength various options such as the inclusion of inelastic scatter are automatically disabled if this IOP model is used 2 3
122. rom one theta band to another which causes a small jump in the solar radiance used to initialize the RTE solution These jumps are typically only a percent or less 78 in magnitude and will not affect AOPs but may give an unsatisfactory appearance to some plots Users encountering such output have occasionally requested the ability to do higher resolution runs EcoLight but not HydroLight surface files and phase functions have therefore been created as follows e The theta resolution is 2 deg from the equator to the pole The polar cap has a 1 deg half angle the band near the equator is also 1 deg e The same phase functions are available as in the regular version of EcoLight e Surface files are available for the same wind speeds as the regular EcoLight 0 to 15 m s by steps of 1 m s e Surface files are available only for a water index of refraction of 1 34 Therefore the surface reflectance and transmittance properties are the same for all wavelengths water temperatures and salinities The ability to do runs with the high resolution 2 deg grid is purposely not built into the HE52 UI because such runs are rarely needed Indeed the use of the high resolution grid is discouraged unless you absolutely do need it for some special simulation requiring highly accurate irradiances or radiances as a function oftheta Using angular resolutions higher than the standard grid is in most cases just a waste of computer time like doing run
123. routine also gives the user additional options for modeling Case 1 waters e g by specifying the microbial and CDOM components as desired and setting the mineral concentration to zero The various mass specific absorption and scattering spectra available in HE52 are discussed in 2 8 and 2 9 below 2 6 MEASURED IOP DATA A common use of HE52 is to compute light fields using absorption a and beam attenuation c coefficients as measured by a WETLabs ac 9 ac s or equivalent instrument as input The IOP model on file HE52 code common abacbb f is designed for precisely this task This IOP model divides the total absorption and scattering into two components component is pure water and component 2 is everything else namely the particles and dissolved substances detected by the ac 9 after pure water values are subtracted from the measured totals according to the recommended ac 9 data processing protocol For many but not all waters the measured signal can be attributed primarily to particles and the scattering for component 2 can be modeled with a particle type phase function The preparation process for including ac 9 data in an HE52 run is described below in 7 3 This IOP model also has the option of reading in data from both unfiltered and filtered ac 9 instruments The unfiltered ac 9 file gives the total particles plus CDOM absorption and the filtered ac 9 file gives only the CDOM absorption Having filtered and unfiltered ac
124. s 1992 this function is shown in L amp W Fig 5 11 The revised versions on CD of Light and Water Sections 5 14 and 5 15 give the details about the calculations used for Raman scattering and fluorescence respectively See 3 3 for comments regarding the depth where the bottom boundary condition is applied when the water is infinitely deep and inelastic scattering is included in the run Note that the option for skipping every other wavelength and filling in the missing wavelength bands by interpolation is not available with inelastic scattering is included in a run 51 7 STANDARD FORMAT DATA FILES As seen above HE52 comes with built in spectra for water absorption and scattering phytoplankton absorption and related quantities needed for its standard IOP models Data files with the reflectances of typical bottom materials such as sea grass sediment and corals are included These default data sets are suitable for generic studies There are also various example data files of showing how to input measured chlorophyll or mineral particle concentration profiles IOP data for absorption scattering and backscatter and the like However if you wish to use HE52 to accurately model the optical environment for a particular time and location to show agreement closure between measured and predicted optical quantities then the built in data files should be replaced by data files appropriate for the chlorophyll concentration water IO
125. s e g Mobley et al 2003 Section 8 4 describes how to write HydroLight subroutines for non Lambertian bottoms EcoLight always assumes that a finite depth bottom is Lambertian 30 0 8 E solid ie measured gunn avg ooid sand t dashed is extrapolated J clean coral sand E aana avg hardpan 1 avg dark sediment 4 green algae J avg coral E J avg turf algae 0 45 red alqae f i J avg seagrass J brown algae E J avg macrophyte 0 2 ee AY 1 avg clean seagrass 4 avg kelp o D A reflectance R m EATON aa 300 400 500 600 700 800 9001000 wavelength A nm Figure 15 Bottom irradiance reflectances as provided with HE52 Solid lines are measured dashed lines are subjective extrapolations 3 2 Infinitely Deep Water Without Inelastic Scattering or Internal Sources The complicated situation for infinitely deep water requires explanation Consider first the case of no inelastic scattering no Raman scattering or chlorophyll or CDOM fluorescence and no internal sources no bioluminescence This is called the source free case Now Z gives the maximum output depth for the run i e the maximum depth at which we wish to obtain output from HE52 For a study in the open ocean we might take Zmax 50 m even though the water is optically infinitely deep In this case HE52 always assumes that the water is homogeneous below depth z and has the same IOPs as are given by the IOP routine at Z Zaa Not
126. s at 1 nm wavelength resolution unless you really do have measurements at 1 nm resolution However if high resolution is required runs can be as follows Directory names and code must be changed from their default values The recommended way to keep things straight is to make a copy of all of the HE52 files in a new directory called for example HE522deg or reinstall HE52 from the CD but install to a directory named HE522deg The original standard grid code and data will then be left unchanged and you can run either the standard or 2 deg versions depending on which UI is run the one in HE52 frontenda HE52WinUI exe or the one in HE522deg frontend HE52WinUI exe Then make the following code changes in the new HE522deg directory e Rename the existing EcoLight surface directory HE52 data surfaces EcoLight to something like HE52 data surfaces EcoLight_ 10deg e The Ecolight high resolution surface files are distributed in directory HE52 data surfaces EcoLight 2deg Rename this directory to 79 HE52 data surfaces EcoLight This is the default directory name where EcoLight looks for surface files e Rename the existing EcoLight phase function directory HE52 data phasefun EcoLight to something like HE52 data phasefun EcoLight_ 10deg e The EcoLight high resolution phase function files are distributed in directory HE52 data phasefun EcoLight_ 2deg Rename this directory to HE52 data phasefun EcoLight This is the default
127. s of wavelength and reflectance values rather than wavelength and concentration specific a or b values See any of the files in directory HE52 data botmrefl for an example of such a data file For bottom reflectances you can add your data file to the list of files in the UI pull down menu for bottom reflectances This is done as follows e Place the wavelength vs reflectance data on a file with the HE52 standard format seen in the file HE52 examples template Rbottom txt e Give the file a meaningful name e g Rmud txt and place the file in the HE52 data botmrefl directory e Use a text editor such as Notepad or Wordpad to add the name of your file to the 57 filelist txt file in the HE52 data botmrefl directory The UI reads filelist txt to get the names of the available files of bottom reflectances The files are shown in the UI in the order listed in filelist txt You can edit this file to change the order of the files in the menu or to add or remove files from the available list If you wish to use an analytical formula to compute the bottom reflectance given the wavelength you can either use your formula to generate a data file to be read by rbottom f or rewrite routine HE52 code common rbottom f to use your formula rather than read a data file HydroLight can also use a non Lambertian bottom in the finite depth case To do this follow the directions in 8 4 and the example routine HE52 examples templates BRDFminn f 7 6
128. s record gives the wind speed in meters per second at an anemometer height of 12 m the index of refraction of the water the water temperature in deg C and the salinity in PSU Names of variable windspd refr temp salinty Example 5 0 1 34 20 0 35 0 These values are used to select the appropriate surface reflectance and transmittance files in the data surfaces directory Interpolation between files is used if the windspeed is not an exact match to one of the available files If refr lt 0 0 then temp and salinity are used to compute the index of refraction using the formula in Quan and Fry 1995 98 RECORD 10 Bottom Reflectance This record defines the type of bottom boundary Names of variables ibotm rflbot Example 1 0 2 ibotm is a flag for the type of bottom boundary as follows if ibotm 0 the water column is infinitely deep The water below depth Zmax to be specified in record 10 below is taken to be homogeneous with IOPs equal to the values at depth z The non Lambertian BRDF of the infinite layer of water below depth Zmax 18 Computed automatically from the IOPs if ibotm 1 the bottom is an opaque Lambertian reflecting surface located at depth Zma The irradiance reflectance of the bottom is taken to be rflbot independent of wavelength Note that 0 lt rflbot lt 1 if ibotm 2 the bottom is an opaque Lambertian reflecting surface located at depth Zma The wavelength dependent irradiance reflectance
129. sion 5 or as HE52 are being licensed to the User on a non exclusive non transferable license for use in scientific research The following requirements which with the preceding paragraphs constitute an agreement for use of HY DROLIGHT by the User hereafter called the Agreement are to be upheld 1 HYDROLIGHT ECOLIGHT version 5 may be installed on the same two computers on which the Lahey Fortran 95 Express compiler that comes with HES2 has been installed You may not network HE52 or otherwise use it on more than two computers or computer terminals at a time 2 This entire notice must be retained within each main program of the source code 3 The following notice must be legibly displayed on the monitor or other output when HE52 is performed HYDROLIGHT ECOLIGHT version 5 2 is Copyright c 2013 by Curtis D Mobley HYDROLIGHT ECOLIGHT IS EXPERIMENTAL AND IS LICENSED AS IS WITHOUT REPRESENTATION OF WARRANTY OF ANY KIND EITHER 108 EXPRESS OR IMPLIED THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF HYDROLIGHT ECOLIGHT IS WITH THE USER 4 Title to HES2 and all portions thereof shall at all times remain with Curtis Mobley HE52 is licensed not sold to the user 5 Any alterations variations modifications additions or improvements to HE52 or merger into other program material to form a derivative work made by the User shall be made or done at the User s own risk and expense Any modification or combin
130. sky 1 0 for a fully overcast sky no sun visible Edtotal is the downwelling spectral scalar irradiance due to sun and background sky incident onto the sea surface the units are W m nm 96 If iflagsky 2 then nsky 3 and skydata 1 to skydata 3 contain the following Names of variables suntheta sunphi cloud Example 45 0 0 0 0 0 where suntheta is the solar zenith angle in degrees defined as above sunphi is the solar azimuthal angle defined as above cloud is the cloud cover 0 0 lt cloud lt 1 0 cloud 0 0 for a clear sky and cloud 1 0 for a solid overcast If iflagsky 3 then nsky 3 and skydata 1 to skydata 3 contain the following Names of variables hour cloud sunphi Example 20 5 0 5 0 0 hour is Greenwich Mean Time in hours Minutes are expressed as a fraction e g 21 40 is 9 25 PM GMT cloud is the cloud cover defined as above sunphi is the azimuthal angle defined as above For this option the solar zenith angle will be calculated from the specified day of year latitude and longitude specified in record 8b Note 1 If you wish to write your own sky model you can modify the front end programs to read in the data needed by your sky model using the skydata array Then pass the needed data on to your model in the manner seen in routine initial f Note 2 HE52 uses an azimuthally averaged form of the Cox Munk wave slope equations In this case the azimuthal angle sunphi
131. stepping past very sharp discontinuities This problem can be avoided by replacing sharp steps discontinuities in the data with slightly rounded steps Figure 28 shows an example of the absorption coefficient a A measured by an ac9 water absorption has been removed The green line shows the values of a A as obtained by linear interpolation between the measured values shown by diamonds The green curve is the values used by HE52 to solve the RTE at any wavelength between 412 and 715 nm 7 8 2 Extrapolation outside the measurement range However suppose the run is done from 300 to 800 nm There are no measured IOPs below 412 or beyond 715 nm Therefore HE52 will use the measurements at the nearest wavelength for the values below 412 and above 715 These values are shown in red in Fig 28 62 absorption coef 1 m 300 400 500 600 700 800 wavelength nm Fig 28 Measured ac9 absorption coefficient symbols as fit by linear interpolation green lines within the range of measurement and as extrapolated outside this range red lines The extrapolation below 412 in Fig 28 will allow HE52 to run but it is very unlikely that the results will be meaningful because the real absorption curve almost certainly does not look like the flat red curve below 412 Extrapolation much outside the range of measured data is almost guaranteed to be incorrect unless done with a physically based model to defi
132. t Light and Water Radiative Transfer in Natural Waters Mobley 1994 describes in considerable detail the mathematical methods employed in HydroLight That book is the primary reference for HydroLight The source code itself is documented by references to the equations of Light and Water Throughout this report the names of mathematical variables are written in italics e g U z or zeta The names of computer programs directories and files are written in a sans serif font e g abcase1 f or Pupcast2 txt Path names are written using the Microsoft Windows DOS format with a backstroke e g data phasefun hydrolight avgpart dpf which is appropriate for PC operating systems User input to and output from programs is shown in Courier Options on the Graphical User Interface are shown in SMALL CAPS Most of the figures in this Technical Documentation were generated in color but printed in black and white for the hard copy The color figures can be seen on the pdf version of the Technical Documentation which is found in the HE52 documents directory on the distribution CD 1 1 HydroLight Technical Notes HydroLight and EcoLight users often pose excellent questions about how various computations are done how to interpret outputs and the like Questions of general interest requiring detailed explanations sometimes result in a HydroLight Technical Note HTN These HTNs are found in the HE52 Documents directory and on the HydroLight Users Gr
133. t file 12 Mroot txt output file 13 Sroot text output file 14 PAR model input 15 first surfwind input file read opened read and closed in initial 16 second surfwind input file read opened read and closed in initial f 17 Lroot txt output file 18 19 reserved for future system output User Available unit numbers 20 39 available for use by user written IOP models etc Core HE52 datafiles input 40 Specific absorption data e g astarchl txt andastarmin txt 41 chlz datafile 42 ac9 datafile 43 bottom reflectance data file 106 44 not used in HE52 45 Edtotal data file total incident irradiance as a function of wavelength 46 49 reserved for future system data files Discretized phase function files dpf files 50 99 There are actually ncomp files opened units 50 to 49 ncomp Note these files remain open throughout run 107 APPENDIX C LICENSE AGREEMENT FOR USE OF HYDROLIGHT ECOLIGHT 5 2 This license agreement is also contained in the source code and in the UI HydroLight Ecolight version 5 2 SINGLE USER LICENSE AGREEMENT HYDROLIGHT ECOLIGHT version 5 2 is Copyright 2013 by Curtis D Mobley HYDROLIGHT and ECOLIGHT are owned by Curtis D Mobley and are protected by United States copyright laws and international treaties These computer programs named HYDROLIGHT and ECOLIGHT and consisting of various main programs and subroutines hereafter referred to collectively as HYDROLIGHT ECOLIGHT ver
134. t file Example file Sky Irrad Example Eddir Eddif txt was created in this way MODTRAN and RADTRANX are much different in their inputs and calculations so it is difficult to do corresponding simulations Nevertheless the figure shows that the total irradiances Ed_total computed by each are within 10 of each other except in the strong atmospheric absorption bands However MODTRAN and RADTRANX partition the total into direct and diffuse components in considerably different ways Fortunately the differences in MODTRAN and RADTRANX irradiances make almost no difference in apparent optical properties AOPs such as the remote sensing reflectance R That is after all why AOPs are of value Figure 24 shows the R spectra for these two sky inputs and for Case 1 water with a chlorophyll concentration of 5 mg m using the new Case 1 IOP model in HE52 The R spectra are the same to within 2 at all wavelengths and they are usually much closer 47 0 005 0 004 Modtran Rrs Radtran Rrs 4 0 003 2 0 002 4 0 001 0 000 T T 350 450 550 650 750 850 950 wavelength nm Fig 24 Remote sensing reflectances R computed with the MODTRAN and RADTRANX input irradiances of Fig 1 and Case 1 water with a chlorophyll concentration of 5 mg m The code in subroutine HE52 code common irradat f is set to read in the user s irradiances and assume that the data are values e
135. t reminds us that they are files of discretized phase functions ready for use in HE52 Note that there are separate files with the same names for HydroLight and EcoLight in the respective directories The HydroLight and EcoLight discretized phase function files are not interchangable because the quads and bands are different Table 3 shows the discretized phase functions distributed with HE52 phase function defining subroutine in file containing the discretized reference HE52 SpecialRuns phase function in directory PhaseFunction HE52 data phasefun isotropic pfiso f isotrop dpf L amp W Eq 5 105 pure water pfpure f pureh2o dpf L amp W Eq 3 30 Petzold average particle pfpart f avgpart dpf L amp W Table 3 10 column 6 Fournier Forand pfff f FFbb0001 dpf Fournier and Forand 1994 n and nu are user for 0 01 backscatter defined parameters fraction to FFbb500 dpf for 50 backscatter fraction Morel large and small particle pfCase1Large f pfCase1Large dpf Morel et al 2002 pfCase1Small f pfCase1Small dpf Table 3 Phase functions distributed with HE52 Subroutines defining several other phase functions are available in the HE52 SpecialRuns PhaseFunction directories These include the one term Henyey Greenstein phase function on subroutines name pfothg f and Kolelevich large and small particle phase functions on pfLarge f and pfSmall f Discretized versions of those phase functions
136. taken at the Bermuda Atlantic Time Series BATS site The BATS Chl values ranged between 0 002 and 0 606 over the course of a year with a mean of 0 152 mg Chl m Although their spectra are similar above 365 nm they are highly variable with season and depth between 300 and 365 nm This variability is likely due to mycosporine like amino acids MAAs which strongly absorb near 320 nm Figure 3 compares the Morrison and Nelson spectra blue curves with the Bricaud et al a of Eq 6 evaluated for Ch1 0 05 mg m red curve the Morrison and Nelson spectra are normalized to the Bricaud value of a 400 The shapes of the Morrison and Nelson spectra are consistent with the Bricaud values for low Chl values Vasilkov et al 2005 present spectra for A A and B A 1 E A between 300 and 400 nm as derived from absorption measurements in coastal California waters Figure 4 shows their A A and B A spectra compared with those of Bricaud et al 1998 The differences at 400 nm reflect the different databases i e different waters used to derive the coefficients 10 ar S S E S E 0 014 0 012 0 010 abs coef m7 tripe a a aa a T x biti Chl Pie ea ep rT 0 05 mg a 500 wavelength nm 3 Fig 3 Comparison of Bricaud et al a for Chl 0 05 red with the Morrison and Nelson normalized absorption spectra blue dotted is summer solid is winter 0 4 0 9 0 2 A or B
137. the subroutine are provided with HE52 The first option is to use the chlorophyll subroutine on file HE52 code common chizdata f to read a file of chlorophyll data This subroutine is automatically called if you select the USER SUPPLIED DATA FILE option for either of the casel IOP models The name of the data file is specified in the UI read in from the Troot txt file and passed to the chlzdata routine This routine reads this file of measured depth vs chlorophyll data fits a linear spline to the data and uses the spline to compute the chlorophyll concentration at any depth The HE52 standard format for the data files read by chlzdata f is seen in 7 1 or on HE52 examples templates chizdata txt The other option is to write your own subroutine This is done by selecting the USER SUPPLIED SUBROUTINE option when running the UI and providing the name of your subroutine If you do wish to write your own chlorophyll subroutine follow the template on file HE52 examples templates chizfunc f or modify either of the two distributed routines These two options should be adequate for most purposes In either case the core code will call the appropriate chlorophyll subroutine chlzdata or your specified subroutine for your problem Note that chlorophyll profile must be specified for all depths relevant to your run 68 8 3 Subroutines for CDOM Absorption CDOM absorption can be specified as a function of depth and wavelength in the same manner as fo
138. ty and wavelength 1 386 T T moor T T ee 127k S26 7 6 ad ee S 6 1 40 oN S240 Te oo SOO siti S 40 T 404 S36 Te 1 55 1 354 index of refraction 1353 Tr rrr ttt or I2 Laaa a ere ine 300 400 500 600 700 800 900 1000 wavelength nm Fig 17 Real index of refraction of water as a function of salinity S psu temperature T deg C and wavelength Values are computed by Eq 3 of Quan and Fry 1995 The Monte Carlo surface calculations are performed as a special run using code described in 9 1 These calculations are time consuming but they need be done only once for each combination of index of refraction and wind speed The results are stored the files in directory HE52 data surfaces as a function of index of refraction from n 1 32 to 1 38 and wind speeds of U 0 to 15 m s When HE52 runs it then uses the formula of Quan and Fry 1995 their Eq 3 to convert the salinity and temperature as specified by the user in the UD to n at each wavelength The available surface files are then interpolated to the needed value of n and U at each wavelength Temperature and salinity effects on the Fresnel reflectance cause less than a 5 variability in remote sensing reflectance near 300 nm and less than 3 at visible wavelengths compared to the use of a value of n 1 34 at all wavelengths as in previous versions of HydroLight However some users have requested the ability to
139. ual Information Solutions Inc All other brand names are trademarks of their respective holders Disclaimer Curtis D Mobley and Sequoia Scientific Inc reserve the right to revise the HydroLight and EcoLight software Users Guide Technical Documentation or other publications and products with no obligation to notify any person or any organization of such revision In no case shall Curtis D Mobley or Sequoia Scientific Inc be liable for any loss of profit or any other commercial damage including but not limited to special consequential or other damages Technical Support Technical support for HydroLight and EcoLight can be obtained in accordance with the user s license agreement from Curtis D Mobley Lydia K Sundman 425 641 0944 ext 109 lsundman sequoiasci com fax 425 643 0595 curtis mobley sequoiasci com If you encounter a problem during a HydroLight or EcoLight version 5 run please e mail us the following e The Licensed to name and Serial Number of the copy you are running These are found on the distribution CD on opening form of the User Interface and at the top of the printout e A description of the problem error messages or other pertinent information e The input file the Iroot txt file for the run which is found in the HE52 run batch directory e The printout file the Proot txt file from either the HE52 output Hydrolight printout or HE52 output Ecolight printout directories e Any user s
140. up to four components and is used only when a concentration is set to be constant with depth Names of variables compconc j j 1 nconc Example 0 1 1 0 Record 5c Specific absorption parameters This record consists of nconc lines of input where each line tells how the specific absorption will be given for each component Names of variables iastropt i astarRef i astar0 i asgamma i Example 4 440 0 0 06 0 0114 The a specifications are iastropt 0 user supplied data file read to get a values 1 Pope amp Fry absorption model used pure water only 91 2 Smith and Baker absorption model used pure water only 3 Prieur Sathyenranath Morel model used Chlorophyll only 4 Exponential model used CDOM only The other a parameters are only used when iastropt 4 astarRef reference wavelength for exponential model astar0 a at the reference wavelength asgamma exponential decay constant Record 5d Specific absorption data file names The next nconc lines of input give the names of the files containing the specific absorption for each concentration option The order MUST match the order of the components in the subroutine for a and b see 8 1 Filenames that are not needed for the run may be stored with the name astarDummy txt as a place holder Names of variable astarfile i Example astarchl txt astarcdom txt Record 5e Specific scattering parameters This record consists of ncomp lines of input where each
141. upplied input data files used by the run e If your run included a user defined IOP model or function or if you included bioluminescence please also send us root for file from the HE52 code batch directory and any user defined subroutines we would need to repeat the run This information will greatly increase the speed at which we can troubleshoot the problem HydroLight Users Group HydroLight licensees are strongly encouraged to join the HydroLight Users Group at http tech groups yahoo com group HydroLightUsers This web site is used to make announcements about HydroLight distribute detailed technical notes about HydroLight algorithms post bug fixes share new code written by users and for communication of general interest between HydroLight users A file of FAQs is maintained there Acknowledgment Dr Marcos Montes of the U S Naval Research Laboratory provided invaluable assistance in extending the RADTRAN sky irradiance model to 300 and 1000 nm He has also assisted with beta testing and developed several internal changes to the core code which have improved its numerical efficiency and accuracy Dr Eric Rehm also performed very useful beta testing of the EcoLight code and comments on Appendix A ii TABLE OF CONTENTS MINTER CVU CATION Scum ote a Siesta te ice Gist ig a raat id a a tes le eee as i ha aR 1 1 1 HydroLight Technical Notes i0 s40 db Kast ae eee sae eae andan ee pa 04 2 TOP MODELS r ila
142. ure 9 shows phase functions determined by Eq 11 along with the frequently used Petzold average particle phase function 100 000 10 000 ia Ss Z 1 000 i S 3 0 400 a gt 0 010 O a 0 0010 0 00016 O 30 60 90 120 150 180 scattering angle deg Fig 9 Phase functions for small orange and large red particles as given by Morel et al 2002 Phase functions as given by Eq 11 for Ch 0 01 purple 0 1 blue 1 0 teal and 10 0 green and the Petzold average particle phase function black are also shown 18 It should be noted that the Morel et al 2002 phase functions have smaller backscatter fractions B 0 014 for the small particles to 0 0019 for the large particles than the Petzold phase function B 0 018 This is consistent with what is known about the phase functions for algal particles e g Ulloa et al 1994 or Twardowski et al 2001 recall Eq 5 The Morel phase functions of Eq 11 and Fig 9 are adopted for use in the new Case 1 IOP model AILIOPs for Case 1 water are now determined by the user specified chlorophyll profile after selecting the particular of A and B spectra corresponding to low medium or high UV absorption To illustrate the quantitative differences including the combined effects of absorption and scattering coefficients and the particle scattering phase function between the classic and new Case 1 IOP models Fig 10
143. xactly at the listed wavelengths The code then interpolates between the listed wavelengths to get sky irradiances at 1 nm resolution and those values are then averaged over the run wavelength bands to get band averaged irradiances just as is done with RADTRANX values which are at 1 nm resolution This allows the run wavelengths to be independent of the data wavelengths However the values in the run printout for a given nominal wavelength will be slightly different than the values in the input data file for the same wavelength because the printout shows the band averaged irradiances which depend on the size of the run wavelength bands as well as on the input data Because AOPs are almost unaffected by the incident sky irradiances the main use of the new sky irradiance input options comes if radiometric variables themselves radiances and irradiances are to be compared with measurements or are required with great accuracy Then it is critical to have accurate values of the input sun and sky irradiances which set the magnitude of the entire computed light field The other use of the new options is to allow HE52 to be run independently of RADTRANX e g when it is necessary to make runs outside the standard 300 1000 nm range 48 5 4 2 Simulation of Lidar Induced Inelastic Scatter It is also possible to use HE52 to simulate lidar induced Raman scatter and CDOM and chlorophyll fluorescence The code has been modified so that if the option 2 a
144. y is writing a subroutine to provide HE52 with the IOPs namely the absorption a and scattering b coefficients and the scattering phase function of the water body The HE52 software package comes with six IOP subroutines for computing a and b as described in 2 Five of these routines abconst abcase1 abnewCase1 abcase2 and abacbb are suitable for use in a wide variety of studies Users can specify individual component IOPs and concentration profiles The abcase2 routine for instance is suitable for many Case 1 and Case 2 waters If you still need to develop your own IOP model the main thing to remember is that the total IOPs of a water body are built up as a sum of IOPs attributable to the various components of the water body Thus the total absorption coefficient is computed from ncomp aoa 2 az d iz Here a z A is the absorption coefficient of the i component of the water body which in general is a function of the depth z and wavelength A A similar equation is used to compute the total scattering coefficient b The number of components in the IOP model is ncomp The default array dimensioning in HE52 allows ncomp to as large as 4 if you need more than 4 components just increase the component limit on the CHANGE LIMITS form of the UI The process of writing your own JOP subroutine is straightforward e Decide what components will be included in the model and in what order For example you might let comp
145. yer of bioluminescence Such bioluminescence does occur in nature e g in the milky seas of the Arabian Gulf HE52 cannot model a localized source of bioluminescence which gives an inherently 3D radiative transfer problem HE52 always assumes that the bioluminescence is isotropically emitted The first few rows and columns of the example data file HE52 data examples SobiolumData_10nm txt are shown below EXAMPLE BIOLUMINESCENCE DATA showing the Hydrolight Standard Format The values are the bioluminescent source strength 50 in W m 3 nm Vales were selected to match the output from the example bioluminescence FUNCTION contained in s0biolun It approximates a bioluminescing layer with a gaussian wavelength distribution Simulating a typical concentration of dinoflagellates The bioluminescence source function is modeled as a gaussian in wavelength with parameters SOtotal 4 00E 05 W m 3 waved 480 0 nm Sigma 15 0 nm FWHM 33 8 nm The bioluminescence source function is modeled as a combination of hyperbolic tangents in depth with parameters zupper 10 0 m Zzlower 13 0 m zscale 2 0m Record 11 gives the number of wavelengths and the wavelengths 34 350 0 360 0 370 0 380 0 390 0 400 0 410 0 420 0 430 0 5 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 5 50 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 0 00000E 00 6 0

HydroLight Technical Documentation

Contents

Download Pdf Manuals

Related Search

Related Contents