User`s guide PETROCK_GUIDE
Contents
1. xst yst xwin ywin xmar ymar sub area cut method data points per grid point minimum and scale length lithological weight rock type weights coordinate rounding store used petrophysical data store used lithological polygons O O mean parameters file std deviations file for bloxer program number of rejection rules inclusion pairs row text 3 2 input parameters file for Pl l 1 Axydnkr3 dat 2 Alitolb lit 3 4 200 200 25 25 5 2 6 1 sub area max dev 7 100 sub area points 8 5 9 dis 1 radius 10 2 inverse power 11 2s 12 2700 missing values T3 0 14 2 duplicate points 15 0 16 1 17 0 18 Petrock3 dat output 19 Stdrock3 dat output 20 me l output 21 9 22 1 4 23 2 4 24 3 4 25 4 4 26 6 4 2 5 4 28 Du Ll 29 8 1 30 9 1 lito rock exclusion 34 35 1 felsic 2 dyke 3 volcanic 4 sediment 5 metam 6 other The three lines that start with a character on the top are used as comments lines Petrophysics file The 1 st line defines Polygons file The 2 nd line defines the the name and path to the petrophysical data file name and path to the lithological polygons file 10 Xmi xma ymi yma xst yst The 3 rd line defines x and y coordinates of the start West and South and the end East or North of the rectangular study area over which the data will be gridded i e the research ar2. This parameter can have values 0 use all data points 1 use the mean 2 use the median of duplicate points or 1 discard all duplicate data points Duplicate data points arise for example from petrophysical samples obtained from deep drillings Because most samples are taken from the weathered rocks from the surface their petrophysical properties can be quite different Therefore it may be advantageous to remove duplicate points totally Coordinate rounding The 15 th line defines the rounding of coordinates in terms of units Value 0 0 means that the coordinates are not rounded at all Since the coordinates are defined in kilometers value 0 1 would round them to the accuracy of 100 meters If duplicate data points are replaced by the mean or the median the rounding could be used to decrease the original amount of data points This could be advantageous in areas where there are lots of samples available Store used petrophysical data The parameter on 16 th line defines whether the petrophysical data used in the gridding are stored in a separate file 1 or not 0 The name of the file is PETROSIN DAT When processing local areas it may be useful to plot the location of data points that were actually used to create the grid and to discard the rest of the values 13 Store used lithological polygons Likewise the parameter on the 17 th line defines if the lithological units used to create the grid would be stored 1 or not 0 The
3. for the tables but NOPA must be defined exactly See chapter 5 for additional information about these parameters 16 The second line is the header of the first polygon It defines the number of vertices NOV 139 the index number of the polygon IP 1 the area of the polygon A 173 8139 km read from the MID file the code number of the lithological unit IC 80 and the host status of the polygon IH 1 The host status defines if the polygon is an actual host polygon IH 1 or if it is a hole inside a host polygon IH 0 The following 139 lines define the x and y coordinates of the first polygon Note that in LIT files the x and y coordinates refer to West East and South North directions respectively The last vertex point of the polygon must be equal to the first one 1 e the polygons must be closed The line immediately after the last vertex point of the first polygon contains the header of the second polygon and so on 6 3 Output files The columns of the output file e g PETROCK DAT that contains the gridded values are X the x coordinate of the center of the grid point Y the y coordinate of the center of the grid point Pl the mean of the data points inside the search radius P2 inverse distance weighted IDW mean of the data points inside the search radius P3 areal lithologically weighted mean P4 areal lithologically weighted mean IDW P5 areal background lithological weighting P6 areal back
4. s q L 3370 3375 3380 3385 3390 X Easting km Figure A 1 Example of the tuning of polygon vertices a detail of an actual polygon The small black dots are the original points and the large red dots are the optimized ones For this polygon n o 3317 the number of vertex points was reduced from 2633 to 570 23
5. std where std is the standard deviation of the data inside the sub area IWEI 0 or inside each lithological unit IWEI 1 2 3 IWEI and a are used to cut out or to reduce the effect of values that for some reason have been given incorrect lithological code Remember that the geological boundaries do not represent actual formations of the nature For example if a low density sedimentary unit contains lots of dense sample points that actually belong to the surrounding area setting IWEI 3 can remove these outliers provided that their amount is small compared to the number of samples of the main sedimentary unit Sub area points The 7 th line defines the minimum number of sample points inside the sub area MSP If the sub area investigation area in Fig la contains too few samples the 11 size of the sub area will be increased by the size of the margins M until the number of points 1s large enough Data points per grid point The 8 th line defines the minimum number of sample points per grid point MGP If the number of points is too small the search radius r in Fig 1b is doubled in size until the number of points is large enough Radius minimum and scale length The 9 th line defines the minimum length of the search radius r and the scale length s used to normalize the distances The initial value of the search range should be large enough that it would contain the MGP points by default However large search radius can
6. the closest points e mean basic lithological weighting f mean alternative lithological weighting Because at least ten points were used to evaluate each gridded value maps d f show much smoother and flatter variations than map c The flattening is also caused by the additional limiting condition which removes some outlying high density samples On the other hand maps d f correspond to the underlying geology much better than map c which was obtained using conventional interpolation Even without any geological weighting map d shows good correlation with the underlying geology Because of the additional limiting the new maps d f reveal the base level of the more correctly particularly above the granite areas of low density The basic lithological weighting used in map e delineates the lithological units The alternative lithological weighting further emphasizes the boundaries of 20 different geological units Note that the gridding was made using the more detailed 1 1 million digital map instead of the 1 5 million map shown in Fig 1 8 Miscellaneous Note that PETROCK only works with rectangular coordinates e g Finnish National KKJ coordinate system Geographical coordinates defined as latitude and longitude values must be converted into rectangular coordinates beforehand The various input parameters have different kind of an effect on the results In particular the number of points per a grid po
7. 15 EBB Pyroxene granite and monzonite 1 885 1 87 Ga 16 Granodiorite 1 89 Ga 17 MN Gabbro diorite 1 89 1 87 Ga 18 Tonalite 1 92 1 91 Ga 19 Mica schist and migmatite 20 Mica schist 21 EEE Metavolcanic rocks 1 92 1 88 Ga 22 MJ Serpentinite and other rocks of ophiolite c 1 96 Ga 23 Garnet gneiss and diorite Lapland granulite belt 24 7 Anorthosite 25 7 Foliated gabbro and granodiorita 1 95 1 93 Ga 26 Gneissic granite and hornblende gneiss 27 Quartzite and conglomerate 28 EE Metavolcanic rock and mica schist the Kittil allochthon 29 UU Calc silicate rock black schist basic volcanic intercalates 30 Quartzite with intercalates c 2 3 2 0 Ga 31 UH Layered intrusions 2 44 Ga 32 J Matic intermediate and felsic metavolcanic rocks 2 5 2 0 Ga Archean 33 Latest Archean granitoids 34 7 Metavolcanic rocks of the greenstone association 35 Metasediments of the greenstone association 36 Biotite hornblende gneisses and migmatites 37 Tonalite trondhjemite gneisses and migmatites Faults and major shear or thrust zone X Kimberlite province xx Impact site Fig 1 Geological map of Finland Korsman et al 1997 The yellow square shows the location of the detailed study area discussed in the text 7800 7 77004 76004 7500 7400 7300 7200 North km 7100 7000 6900 6800 6700 6600 it eee fh
8. 97 and the petrophysical database of rock density and magnetic susceptibility values measured in laboratory conditions Korhonen et al 1997 The digital map used in PETROCK consists of 92 lithological units corresponding to geological classes of different rock type stratigraphy and age Because of the high level of details of the geological map of Finland in the scale of 1 1 million a simplified version in the scale of 1 5 million is shown in Figure 1 The map consists of 7922 polygons of which 1959 are duplicates resulting from holes inside the actual polygons The polygon data consists of about 450000 polygon vertices Because the determination of the locations of the sample points inside the polygons is the most time consuming part of the PETROCK method the polygon data has been optimized using the LITOTUNE program see Appendix in advance After the tuning the reduced number of polygon vertices is about 290000 The petrophysical database contains more than 130 000 samples Before the actual gridding each petrophysical data point has been given a lithological code 1 92 based on its location on the geological map Petrophysical data points outside the borders of all lithological polygons are given class number O zero The density data are limited between 2200 and 3200 kg m which approximately refers to 3 0 x the standard deviation 170 kg m around the mean value 2700 kg m The main purpose is to remove outliers erroneous data va
9. PETROCK Lithologically weighted interpolation of petrophysical data User s guide Version 1 1 2011 Markku Pirttij rvi University of Oulu Department of Physics 1 Introduction The PETROCK program is used to interpolate irregularly sampled petrophysical data e g density or magnetic susceptibility or geochemical data K Zn Au on an even grid The two main purposes of the PETROCK program are a to create improved maps of the geographical distribution of petrophysical parameters and b to create initial 3 D models for potential field modeling and inversion The program uses moving window strategy in which the gridded values are estimated from the nearest sample points using both the mean and the inverse distance weighted mean The program also uses the digital geological lithological maps to provide areal weights and background weights for points inside the same lithological unit A limiting condition can be defined to remove outliers from the petrophysical data inside the investigation area or inside each lithological unit If the limiting condition is based on the mode largest class or the median the gridding can yield the base level of the lithological units better than if the mean of the samples was used Thus in areas where sampled data are not available the PETROCK gridding should be more reliable than normal interpolation would be In addition rock rejection rules can be defined that either reject some rock types fr
10. andard deviation related to ME1 SD2 standard deviation related to ME2 AD average deviation related to ME1 ND number of data points inside the sub area The LITOVALS DAT and LITOERRS DAT files contain the petrophysical background values and standard deviations per each investigation area The files are formatted so that the columns are the lithological units The order of the columns depends on the study area and therefore it is defined in the header of the file In the PETROSIN DAT file the columns are the same as in the original petrophysical data file The LITHOSIN BNA file is stored in Atlas BNA file format which is defined for example in the user manual of the Surfer program by Golden Software The following example illustrates the BNA format Pname Sname 4 35359463 7779 234 3537 069 7775 823 3537 310 7771 241 3538 019 7769 422 Pname Sname 2 3538 646 7776 931 3538 229 7777 292 18 The example defines two line segments which consist of four and two points respectively Each line segment has a header line that contains two character variables that identify the objects and a numeric parameter NPL The absolute value of NPL defines the number of points of each line segment A negative value means that the object will be a line whereas a positive value means that the object will be interpreted as a closed area polygon sphere or ellipse The polygon must be closed so that the first and the l
11. ast point are the same For a single point NPL 1 and for an ellipse and a circle NPL 2 To learn more about BNA file format see Surfer documentation or the web document at lt http www softwright com faq support boundary_file_bna_format html gt 7 Example Figure 4 shows various mapping results from an area around the cities of Hyvink and H meenlinna in the southern Finland The size of the study area is 100 km by 100 km Geologically the area is characterized by high density gabbros in the center and the low density granites in the surroundings The rest are metavolcanic rocks mica schists granodiorites and granitoids cf Fig 1 Map a shows the geological units 1 5 million scale and map b shows the locations of the 2351 petrophysical samples Map c shows the interpolation results from conventional inverse distance power of two interpolation method Golden Software s Surfer 7 When comparing map c with the background geology the inverse distance interpolation reveals two main weaknesses It brings up too many details from the individual samples and incorrectly interpolates the sparsely sampled area in the south Nonetheless 1t should be noted that in this specific case the inverse distance interpolation was much better than other conventional interpolation methods kriging minimum curvature etc Map d shows the mean bulk density obtained using the new mapping method without any geological or inverse distance wei
12. d the polygon file can be very large several megabytes it is recommended that they are stored in the program folder together with the executable file and each gridding experiment is run in a separate subfolder This requires that each subfolder contain a copy of the PETROCK BAT batch file The batch file should contain only one line that defines the path to the actual executable file C PETRO PETROCK PETROCK EXE or PETROCK EXE where denotes the previous directory level The PETROCK program can then be run double clicking the icon of the batch file inside the Explorer program of Windows Note that separate versions of the PETROCK INP script file should be located inside each subfolder The program stores its results in column formatted text files To visualize the gridded data as a contour or an image map a separate plotting program must be used e g Golden Software Surfer or Geosoft Oasis Montaj If the results are not satisfactory the PETROCK program can be run again using different initial processing parameters 5 Input parameters The following example describes the syntax of the PETROCK INP script file Note that all file names text strings must be put inside hyphens e g here Note also that if the program is run interactively the parameters are provided in the same order as in the script ETROCK4 program petrophysics file polygons file 3575 3790 69255 7000 Lae Le xmi xma ymi yma
13. e a background value of each lithological unit To improve the estimate of background value outliers are removed using a limiting condition based on the standard deviation around the mean median or mode of the investigation area In addition rejection rules based on the hierarchical rock classification of the samples can be used to reject allow certain rocks from in certain lithological unit This allows for example removing all but sedimentary rocks from sedimentary units and thus greatly improves the estimate of petrophysical background value The actual background weighting is made by using the background estimates of the lithological units below the five special points Fig 3c as additional data values 4 Program usage Before using the program ensure that the petrophysical data file and the file containing the lithological polygons are available and in the correct format see chapter 6 Immediately after the program has been started it asks whether the input parameters are read from a file recommended or given interactively from the keyboard The name of the input file must be PETROCK INP If the input file does not exist the parameters should be given from the keyboard In this case the PETROCK INP file will be created automatically and it can be used in further examples thereafter The format of PETROCK INP and the meaning of various input parameters are discussed in the chapter 6 Because both the petrophysical data file an
14. e is computed using the points inside current sub area the level can change between sub areas which creates a kind of mosaic pattern if the size of the sub area is small and the lithological weighting is strong 12 Missing values The 12 th line defines the parameter value used for grid points that do not fit inside any lithological polygon Thus grid points that locate outside the national borders or over the sea area are given a constant value Typically this value should represent the mean value of the petrophysical data Note that points outside the lithological units are given code 0 Although these values will not be used in lithological weighting they will be added to the sub area mean Rock type weights The 13 th line defines the rock type weight RWEI This parameter can have values 0 no weighting 1 weighting per sub area or 2 weighting per lithological unit Rock type weighting gives additional weight for grid points and works like the lithological weighting Instead of the proportional area of the unit the abundance of the rock type the proportion of specific rocks to the total number of samples in the sub area is used as a weight Thus the main rock type in the sub area RWEJ 1 or in the separate lithological units RWEI 2 is given more weight than others Note that this parameter is still purely experimental and its use is not recommended Duplicate points The 14 th line defines how to handle duplicate data points
15. ea R in Fig la The last two parameters define the grid spacing in x and y directions Note that in PETROCK x and y coordinates refer to East West and South North directions respectively Xwin ywin xmar ymar The 4 th line first defines the x and y size of sub area that is the moving data window that passes through the main mapping area C in Fig la The third and the fourth parameter on this line define the x and y width of the margin area M in Fig 1 around the computation area Note that the width of the investigation area is actually C 2xM For example the x width is xwin 2xxmar Sub area cut method The 5 th line defines the reference value s used to limit the petrophysical data within each sub area This parameter can have a value IWEL 0 1 2 or 3 If IWEI 0 the mean of all data values inside the sub area are used as a reference point The other IWEI values use 1 the mean 2 the median or 3 the mode largest class of each lithological unit as separately for each lithological unit If the parameter deviations within the sub area or lithological unit are large this method can be used to cut out outliers and samples from incorrectly classified petrophysical samples Sub area max dev The 6 th line defines the multiplier a for the standard deviation that sets the limiting range around the reference point discussed above Only the values within range p dp lt p lt p dp where p is the reference value IWED and dp a
16. each unit by the area of J If C does not contain any lithological units all the grid points inside C are given a user defined mean value or the value reserved for missing data a b c I M C Figure 3 a The components of the moving window strategy total research area R computational area C and margin area M b the search radius r around a grid area g and c the five special points of a grid point The grid points must also locate inside the area defined by the polygons within national borders and islands Thus for each grid point the center and the four corner points are first tested Fig 3c If all special points are outside the lithological polygons the grid point is given the mean value or missing data value Although it would be more accurate to check if the small rectangle around the grid point g in Fig 3b overlaps with any of the polygons testing only the special points is faster If at least one special point is above some polygon the petrophysical samples within the search radius r around the grid point Fig 3b are sought for and given weights based on the inverse of their distance from the grid point If not enough sample points are found the search radius is doubled in size as many times as needed Once the petrophysical data values their inverse distance weights and the lithological weights are known the geologically weighted mean of the pe
17. etamorphic rocks are coded into 118 classes as 5 x y z other rocks and ore bodies are coded into 36 classes as 6 x y z For the PETROCK program the rock type is re defined as a real number floating value using only the first two digits of the rock type code Note that the original character codes of the rock types must have been transformed into real numbers before the data can be used in PETROCK To be able to handle duplicate data values correctly the petrophysical data file must be sorted manually according to increasing x and y coordinates before it can be used in PETROCK 6 2 Input polygon data The LIT file containing the vertex points of the lithological polygons of the digital geological map The LIT file has been prepared using the MIF2BNA program See the documentation of the MIF2BNA program Pirttij rvi 2005 for information how to transform the original polygon data from Mapinfo MIF MID files into LIT format The following example illustrates the format of the LIT files 7922 458000 139 1 173 8139 80 1 3536 588 7780 130 3537 366 7775 912 30372039 7773 585 39372701 7771 419 3538 661 7769 882 3542303 7764 564 35364587 7780 138 3536 588 7780 130 257 2 412 8682 81 1 3529 204 77774384 3530 652 7776 194 etc The first line defines two important parameters a the number polygons NOPA 7922 and b the total number of polygon vertices NOVA 458000 Note that NOVA is used merely to allocate enough memory
18. flatten the gridding and enhance the effect of anomalous sample points The scale length depends on the size of the computation area and the power of the inverse distance weighting Normally the minimum radius is equal to the grid spacing and the scale length is equal to one If the distances are large it may be advantageous to increase the scale length to distribute the inverse power to greater distances Inverse power The 10 th line defines the power of inverse distance weighting The higher the power the more details the grid can show This however may not always be desirable since the effect of anomalous points will create spike like dots in the map Normally the inverse power should be 1 or 2 The inverse distance weights are computed as w sls t where s is the scale length and rf is the distance of the sample point from the grid point Thus if the distance is equal to zero the weighting factor is equal to one If the distance is equal to the scale length the weighting factor is equal to 0 5 Lithological weight The 11 th line defines the standard lithological weight LWEI Lithological weighting means that the mean value of the lithological unit units the grid point belongs to is used as an additional data value s as if it were a point inside the search radius LWEI gives additional importance to this mean value and thus creates a grid map where each lithological unit stands out having a common level However since the mean valu
19. ghting Map e shows the density map obtained using the basic weighting scheme where the areal proportions of the lithological units are used as weights Map f was obtained using the alternative lithological weighting which emphasizes the lithological boundaries In these examples the grid size was 1x1 km the size of the computational area was 5x5 km and the margins were 25 km wide The minimum number of sample points per investigation area and per grid point and were 100 and 10 respectively 19 Duplicate points were replaced with their median value Additional limiting condition was used where the reference point was the median and the limits were 1 0 x the standard deviation of each lithological unit in each investigation area 6780 hos 6760 E E 6740 E S S fon SEL G S 5 6720 Ly E 5 Z 2 57207 Z 6700425 f po 3 u 6680 sie es La a 3300 3320 3340 3360 3380 3400 3300 3320 3340 3360 East km 6780 6780 6760 6760 6740 E 6740 E ph a 5 6720 S 6720 5 Z j Z l Z 6700 6700 7 gt 6680 l 6680 6680 3300 3320 3340 3360 3380 3400 3300 3320 3340 3360 3380 3400 3300 3320 3340 3360 3380 3400 East km East km East km Bulk density i kg m 2500 2600 2700 2800 2900 3000 Figure 4 Mapping example a geological units cf Fig 1 b sample locations c conventional inverse distance interpolation d mean of
20. ground lithological weighting IDW P7 background lithological weighting P8 background lithological weighting IDW P9 the median of the data values inside search radius P10 background value of the lithological unit below the grid point Ll lithological code of the unit below the grid point RD search radius used for the grid point Note that grid coordinates start from the SW corner of the study area The end points of the sub area are shared with the two adjacent sub areas In practice this means that the eastern and northern sides of the sub area are processed twice In the STDROCK DAT file the columns of the file of the standard deviations are X Y S1 S2 S3 S4 S5 S6 S7 S8 RD NU and NI where S1 S8 are the standard deviations of the 17 gridded values corresponding to the weighting methods P1 P8 RD is the search radius actually used NU is the number of petrophysical sample points used to evaluate the grid point and NI is the number of sample points ignored due to the additional limiting condition based on the standard deviation The SUBAREAS DAT file contains statistical information about each investigation area moving window The columns of the file are X the x coordinate of the center of the sub area Y the y coordinate of the center of the sub area MEl simple mean ME 2 lithologically weighted mean areal weighting MED median MOD mode MIN minimum value MAX maximum value SD1 st
21. int and the multiplier of the additional limiting condition affect the results Moreover different weighting methods give rise to different results and artifacts may appear at some locations Because the results of the gridding are not unique it may be necessary to run the program multiple times with different input parameters to get acceptable results I wrote the PETROCK program in Spring 2004 at the Geological Survey of Finland in Espoo as a part of the 3DCM 3 D crustal model project funded by the Academy of Finland I added background weighting scheme and rock rejection rules in 2010 when writing a scientific paper on the subject The program was written for the 32 bit Windows operating system using standard Fortran90 language The program can be compiled and run on other computer platforms provided that the associated petrophysical data files and geological map files exist 9 References Korhonen J V S vuori H and Kivek s L 1997 Petrophysics in the crustal model program of Finland In Autio S Ed Geological Survey of Finland Current research 1995 1996 Geological Survey of Finland Special Paper 23 157 173 Korja A 1989 The new rock classification of the petrophysical database and planning of concurrent classifications Report Q11 27 0 89 1 Geological Survey of Finland 47pp Korsman K Koistinen T Kohonen J Wennerstr m M Ekdahl E Honkamo M Idman H amp Pekkala Y Eds 1997 Suomen kallioperakar
22. lues and other exceptional data values obtained from ore bodies for example The resulting density data consists of 129 252 points The mean value and standard deviation are 2725 0 and 134 8 kg m respectively Likewise the magnetic susceptibility data were pre processed by removing values smaller than 10 SI units including negative values and taking the 10 base logarithm of them The resulting logio normalized susceptibility data consists of about 126 000 points the mean and standard deviation being 2 163 0 00687 and 0 8551 respectively The locations of the sample points used in this study are shown in Figure 2 Figure 2 and the results hereafter are mapped in the national rectangular KKJ coordinate system of Finland Bedrock of Finland 1 5 000 000 Caledonian tectonic units 1 E Schists gneisses or intrusions of variable origin Paleozoic 2 MI Alkali rock pipe livaara and carbonatite Sokli 3 E Sandstone and shale Cambrian Neoproterozoic 4 E Sandstone and shale Vendian Mesoproterozoic 5 Dolerite dykes northern Finland 6 MN Dolerite sills Jotnian 7 Sandstone and shale Jotnian 8 _ Rapakivi granite 9 EE Gabbro anorthosite 10 Dolerite dyke swarms Subjotnian Paleoproterozoic 11 E Quartzite and conglomerate molasse of Lapland 12 BN Postorogenic granitoids c 1 8 Ga 13 Late orogenic granites 1 85 1 8 Ga 14 Granite and granodiorite 1 88 1 86 Ga
23. name of the file is LITHOSIN BNA and it uses the Atlas BNA file format These files can be read in as overlay maps in Golden Software s Surfer and BLOXER Output mean parameters file The 18 th line defines the name of the output file where the parameter values of the gridding are stored This file contains the results from different kind of weighting methods See the next chapter to find out the file format Output std deviations file The 19 th line defines the name of an output file where the standard errors of the gridded data are stored See the next chapter to find out the file format Output bloxer file The 20 th line defines the name of an output file for BLOXER program Pirttij rvi 2005b This file is used to import petrophysical data into the top elements of 3 D density models used in gravity modeling software such as GRABLOX Pirttij rvi 2009 The exported value corresponds to background weighting with inverse distance weighting 1 e without basic areal weighting A dummy column with zeros is added for elevation height values An additional column derived from to the search radius can be used to assign weights fix free status for the data Number of rejection rules The 21 st line defines the number of rejection rules NRR to be applied during background weighting The maximum value is 92 rules one for each unit Lito rock exclusion inclusion pairs The following NRR lines lines 22 30 define the actual rejec
24. nput file In the SUBAREAS DAT file PETROCK stores information such as the basic mean lithologically weighted mean standard deviation median mode and the number of points inside sub area about each sub area If the user chooses to the petrophysical data and the lithological polygons actually used Lines 16 and 17 in PETROCK INP are stored in the PETROSIN DAT and LITHOSIN BNA files 6 1 Input petrophysical data The columns of the petrophysical data file are X Y P IC IR where X the x coordinates easting of the sample points Y the y coordinates northing of the sample points P the value of the petrophysical sample IC the lithological code of the sample 0 92 IR the rock type of the sample point 1 6 Note that in PETROCK the x and y coordinates are given in a rectangular coordinate system where x and y axes are positive towards East and North respectively Geographical coordinates defined as latitude and longitude values must be converted into rectangular coordinates beforehand Note also that normally distances are defined in kilometers km The rock type is based on the hierarchical classification system of the rock samples used by the Geological Survey of Finland In this system plutonic rocks are coded into 85 classes as 1 x y z dyke like rocks are coded into 36 classes as 2 x y z volcanic rocks are coded into 89 15 classes as 3 x y z sedimentary rocks are coded into 33 classes as 4 x y z m
25. o surrounding points Output IP int 4 array dim MP with 1 important amp O unimportant point int 4 the total number of important points just in case if needed ote the polygon must be closed so that the last point is same as the first point i e m XP NP XP 1 and YP NP YP 1 this means also that the actual unique number of vertices is NP 1 ote small polygons that have less than MNP points are not processed Note if angle ALF 0 then the opening angle condition is not used Pirttij rvi 2003 2004 subroutine polytune xp yp np rma alf ip n implicit none integer parameter mnp 5 integer i il i2 np n integer dimension np ip real rma alf x0 y0 rr angl ang2 real dimension np xp yp ip 1 np 1 if np gt mnp then x0 xp 1 22 y0 yp 1 do i 2 np 1 rr rr xp 1 x0 2 yp 1 y0 2 ip i 0 if rr gt rma then ip i 1 x0 xp i yO yp i rr 0 else if alf 0 then il i 1 12 itl angl atan2 yp 1 yp 11 xp 1 xp 11 ang2 atan2 yp i2 yp 1 xp 12 xp 1 if abs angl ang2 gt alf then ip 1 1 x0 xp 1 yO yp 1 rr 0 end if end if end do end if n sum ip np return end subroutine polytune 70104 POZA L K y 4 7005 nee i q E 7000 e E go z lt 2 po E z i Z k d F 6995 4 Pes p L gt P e E A f dl e Y e990 eos i L F E i i 4 0 6985
26. om particular lithological unit or allow only certain rocks in certain lithological unit The petrophysical data are read from column formatted text files The polygon information of the geological maps is read from text files that have been converted from Mapinfo MID MIF files into a special LIT file format using the MIF2BNA program To reduce the number of unnecessary polygon vertices the polygon data can be optimized using POLYTUNE see Appendix Input processing parameters are given via keyboard or using a separate script file PETROCK INP Output data are saved into column formatted text files 2 Table of contents 1 MroducHon csm dd 2 Ze Table OF CoMo las 3 Dy OA A e EEEE 4 P aora a a E KA E E E niga went oaag ah Gents sealiuenduayn E A A gestae a eaten es 9 Dec MULAN ANIC LETS sennae id 10 6 PUL formats A A eee De leche A 15 6 1 Input petrophysical data lt s iassanestecsaedesdebgoansussbgedadasgantescadtesdetauastesosgassasnatadtatenusnatees 15 6 2 Input polygon datan oE n EA REE EEA idad 16 OS OULU file Sinore e a E a a E E T 17 Te Example aida dio 19 Oe VS Ce CU A A E a E O ce eee A A eke 21 A A a ea a ae chs Se EA EA A a he ae 21 PROD CTR sais Resta A lots a e a a a Aa E 22 Keywords Petrophysics Interpolation Density Magnetic susceptibility 3 Interpolation method The lithological weighting of petrophysical data is made using the digital geological maps of Finland in the scale of 1 1 million Korsman et al 19
27. pide A oo a a E O E a E 3050 3150 3250 3350 3450 3550 3650 3750 East km Fig 2 Locations of the petrophysical samples of Finland Because of the vast size of the research area covering the whole Finland 700 km by 1200 km and the large number of polygon vertices and petrophysical samples the gridding is implemented using a moving window strategy The definitions of the gridding concept are illustrated in Figure 3 A rectangular xy coordinate system is considered The research area R is systematically processed by a moving window which is the computational area C which contains the evenly spaced grid points where the mappings are to be made The computational area is surrounded by margin area M Together M and C define the investigation area For each C the petrophysical data and the lithological polygons are limited by the investigation area J This makes computations faster because it allows using only a fraction of the data and the polygons when mapping the grid points inside C However to ensure that a sufficient amount of data points locate inside the margins M are doubled in size as long as needed until the number of petrophysical data points inside is greater than the predefined minimum value Before processing the individual grid points the surface areas of the lithological units below the current rectangular investigation area are computed The lithological weights are obtained by normalizing the surface areas of
28. tion rules as pairs of lithological code LIT 1 92 and rock type ROK 1 6 The last line in the PETROCK INP file is used as a comment to remind the user how the rock types are defined in the hierarchal classification system used by Geological Survey of Finland Currently only the first hierarchy level of the rock types is used 1 felsic 2 dyke 3 volcanic 4 sediment 5 metamorphic and 6 other rocks although in the petrophysical data file the rock types are saved as floating point values two level hierarchy More information about the classification is given in a GTK report Korja 1989 14 The rejection can be done per rock type or per lithological unit If rock type is defined as a positive value ROK gt 0 all samples of that rock type are rejected from the background estimate of the lithological unit If the rock type is defined as a negative value ROK lt O0 that particular lithological unit is allowed to contain samples from that rock type only absolute value of the given value of course In the example script all but sedimentary rocks were allowed in sedimentary lithological units 1 2 3 4 and 6 and only felsic rocks were allowed in granitic units 8 and 9 More over sedimentary and felsic rocks were not allowed in unit 5 6 File formats The PETROCK program saves the results the gridded parameter values and their standard deviations into two files the names of which are given as the end of the PETROCK INP i
29. trophysical parameter can be computed at the grid point Before processing the data it is possible to use coordinate rounding and consequent removal of multiple data values for example from a drill hole Multiples can be replaced with the mean or the median value or they can all be accepted or neglected Important parameters affecting the gridding results are also the minimum number of data values per investigation area Nmin the minimum number of points per grid point Ain and the minimum search radius Yin used to search for the nearest petrophysical data points Note that the size of the computation area should be adjusted so that it contains enough data values and provides sufficiently accurate lithological weights for the computation area PETROCK uses several weighting methods and stores all the results into the same output file It stores the standard deviations and the number of petrophysical data points used to compute the grid values into another output file Additional information mean median maximum minimum standard deviation average deviation number of points about each investigation area is stored into third output file The weighting provided by the above mentioned method is weak when considering areas where samples are not available To provide a better guess in such locations an alternative background weighting is used In this method the mean median or mode largest class of the lithological unit is used to defin
30. tta Berggrundskarta Over Finland Bedrock map of Finland 1 1 000 000 Geological Survey of Finland Special Maps 37 21 Pirttij rvi M 2005a MIF2BNA Conversion of polygon data of Mapinfo MIF MID files into LIT file format User manual Report Q16 2 2005 1 Geological Survey of Finland Pirttij rvi M 2005b BLOXER visualization and maintenance of 3 D block models version 1 5a user s guide University of Oulu Department of Physics 38 pp lt https wiki oulu fi display mpi oulu fi Block model maintenance gt Pirttij rvi M 2009 GRABLOX2 3D modelling and interpretation of gravity and gravity gradient data version 2 0 user s guide University of Oulu Department of Physics 62 pp lt https wiki oulu fi display mpi oulu fi Gravity inversion using block model 2 gt Appendix POLYTUNE Fortran90 algorithm for tuning the vertex points of polygon data POLYTUNE a subroutine that divides and marks the points of a polygon into important and unimportant unnecessary ones based on the maximum distance and maximum opening angle between the points Language Fortran 90 Digital Visual Fortran 6 6 Input XP real array dim MP x coordinates of the polygon vertices YP real array dim MP y coordinates of the polygon vertices NP int 4 actual number of polgon vertices RMA real 4 the maximum distance between two adjacent points ALF real 4 the maximum angle the point it can have with respect to the tw
Download Pdf Manuals
Related Search