Home

EM4Soil - uplink

1. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Distance m Figure 11 1 Using coordinates without enough resolution produces strange graphics If your operating system is the W7 then verify the following 1 is the Pack Service 1 installed ii is the option run programs made for previous versions of Windows active iii are the permission for read write and modify files in the folder active If the program is not working properly do the following before contact EMTOMO 1 open a MSDOS command prompt window go to the start button all programs accessories ii move to the folder using the cd command for example cd users emtomo iii run the program just typing the name and pressing the return iv after the problem occurs make a print screen of the MSDOS windows Send this together your comments to EMTOMO 76 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion 12 Examples DUALEM 642 survey Figure 12 1 shows a site location map of a survey carried out with DUALEM 642 The survey has more than 7000 sites which correspond to a more than 42000 measured values Ext Dzy Molene Soye Print cla Axut Distance Y mr aso o APTUS Liistance X im Figure 12 1 Site location of a DUALEM 642 survey Raw Data x D f E m et 276 2 reS 582 1 lHistanece X ra Figure 12 2 Apparent conductivity map for sensor HCP 2 m This data set is used with the kind perm
2. Figure 8 10 Defining SA parameters Save and Print entrances in the menu bar can be used to save the results Note Use the Exit entrance in the menu bar to back to the EM4Soil program 63 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 9 Input File Format 9 1 Input Data Files Standard EM4soil format All the input files required by EMA4soil and output files generated by EM4soil are ASCII files Files with data acquired with LIN instruments along a transect should have the following format For DUALEM s1 Name of line N h X Y Z HCPIlm PRP1Im Lat Lon Time Sequence of values row by row For DUALEM s Name of line N h X Y Z HCP2m PRP2m Lat Lon Time Sequence of values row by row For DUALEM s21 Name of line N h X Y Z HCPIm PRPIlm HCP2m PRP2m Lat Lon Time Sequence of values row by row For DUALEM s42 Name of line N h X Y Z HCP2m PRP2m HCP4m PRP4m Lat Lon Time Sequence of values row by row For DUALEM s421 Name of line N h X Y Z HCPIlm PRPIm HCP2m PRP2m HCP4m PRP4m Lat Lon Time Sequence of values row by row 64 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion For DUALEM s642 Name of line N h X Y Z HCP2m PRP2m HCP4m PRP4m HCP6m PRP6m Lat Lon Time Sequence of values row by row For EM31 EM38s Name of line N h X Y Z VDM HDM Lat Lon Time Sequence of values ro
3. x Cait Sarvey nat DstsPocsssing Irversixe Save ant lelp Abat Actiuir Plot Raw Data ie Plat Filtered Data oom Tata mode ACP _ Plot Decimated Data 983 101 Noisc Analysis Plut Ney Curreded Piot Data Shifted Pint Initial Mardel Plot Inverted Model Ilot Dota Lesponse PA Plul DOI PEP 99 2 ZORA AST Ano 96 5 BSA N Distance m Quit Cata mode PRF 193 0 298 3 3e77 1972 5065 685 8 795 9 394 8 964 2 Distance m Survey METI9 Ares Dats Sensor Dual121 Figure 1 1 Spatial distribution of apparent soil electrical conductivity o mS m along a transect measured using a DUALEM 421 in horizontal coplanar HCP and perpendicular coplanar PRP modes of operation and spacing of Im 2m and 4m EMASoil is capable of inverting multifrequency EM data along a survey line i e transect The software is capable of processing and inverting EM data collected by either Geonics EM38 EM31 and the EM34 as well as DUALEM instruments In addition the software is capable of carrying out inversion of data collected with PROFILER and GEM instruments EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E TH Soil v0 CM Latcrally Constrained In cision l aig x Ext Suw IInp DazaProcessng I wersion Savc Mirt Hap Abot Artinn Plot Raw Data lot Filtered Data Plul Decimate d Dala Plot Noise Analysis Plat Neg Corrected Plot Data Shifted l ot Initial Mode Lo fluUliv
4. And E Selecting data for inversion ID x Exit Figure 3 3 3 Data to invert continuation YES to be inverted with N n or NO the data will not be inverted The available data will be highlighted in green in red if there is no data ES Method of inversion Ioj x Exit Choose the method Cumulatiwe Functions Full Solution Figure 3 3 4 Choosing the approach linear or full solution 4 Formats 4 1 Format of Input data file The format of the input data file is explained in Section 9 of this Manual 100 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 4 2 Format of Initial Model The format for the file containing a 1D initial model which will be used in the Q2D inversion is as follows NL Di Do ee D nt 1 Ci C gt Cn mS mS gt MS Where NL is the number of layers D Ci amd mS are the depth of the bottom interface m the conductivity mS m and the magnetic susceptibility mSI or ppt of the ith layer Example of a 1D model of 5 layers 5 0 4 0 8 1 3 2 0 3 0 3 0 3 0 3 0 3 0 6 0 6 0 6 0 6 0 6 0 The format for 1D initial models for 1D inversion is similar but should contain the reference to the constraints of each parameter using the symbol F free parameter or C constrained 101 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion EMTOMO 1s a registered mark 2011 102
5. Data to be filtered E All Iw HCP PRP I 1m 2m 4m OK Figure 7 2 2 The Filter parameters menu If the number of negative values in the data set is less than 30 of all data the user can try to correct or delete them 50 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion It must be noted that correcting data can originate false anomalies The user must pay attention to this relevant issue Grid Data HEF 1 HtP 2 HtF 4 a E 450 0 a F E ra ra ra a u 4800 a 480 0 a a a m m m m a E b 470 0 470 0 4610 460 0 200 5 S100 B20 0 800 0 8100 Been 0 H D BO peera E Distance X m Distance X m Distanee X ira PHF 1 FRF PRF 4 T H 490 0 H H p g 4600 al a a F E 47 0 480 PAGG Blo Fa0 0 TLA 8100 be rt AO F100 20 0 Distance X m Distance X m Distance X fm 10 0 pli A o0 0 40 0 marma Figure 7 2 3 Example of a gridded data set 5I vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 7 3 Profile mode Clicking in Add Profile the arrow of your mouse will change to a cross Click the left mouse button at the beginning of the profile and move till the end of the profile keeping the button pressed A menu appears showing the coordinates of the end of the profile Figure 7 3 2 when you leave the button In this menu it is possible to identify the profile name it The sites included in the profile will be those inside of a reg
6. Co e CNI F F gt e FNL 1 Fi F gt secera END Where NL is the number of layers D and C are the depth of the bottom interface m and the conductivity mS m of the ith layer F or Ci are used to say that ith is Free or Constrained NOTE when the data are input through the OTHER sensor the initial models should contain the value of the magnetic susceptibility mSI after the conductivities see the section about the use of OTHER format at the end of this manual For example NL D Do DNL 1 Ci C sae CNL Msl Ms2 MSNi 71 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 9 2 3 Initial model for EM sounding inversion The format of the file containing a 1D initial model for inversion of EM soundings is as follow N Cond1 Cond Min Cond Max Thi Thl Min ThlMax Cond2 Cond2Min Cond3Max Th2 Th2Min Th2Max CondN CondNMin CondNMax Where N is the number of layers Condi CondiMin and CondiMax are the average minimum and maximum values of the conductivity n mS m of i layer Thi ThiMin and ThiMax are the average minimum and maximum values of the thickness in m of the i layer If Min and Max values of a parameter are equal and equals to the average values the parameter will not change during the inversion 9 2 4 Default Initial models for Q2D inversion In normal conditions a initial model default is built when a profile is input in EMASoil The number
7. Defining the model slices elevation Settings Torx Exit Data Grid geometry Raw Data min sites xmax Plot sites C NO sites amin Conductivity Slices Resistivity Elev m Contour type Depth m C Linear Logarithmic Rescale axis YES OK Minimum elev 21 1 Maximum elev 28 9 Figure 7 1 6 Defining the model slices depth Settings Maps displayed in the MAP module have axis with different scale The Rescale axis option allows to have axis with the same scale 48 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 7 2 Data Processing Some basic data processing can be done in the map module These options are mainly devoted to prepare data for Q3D inversion However they can be used in any imported data E EM45oil Map module Exit Display Data Processing Tools Profile mode Inversion Zoo Action pres Undo Rotation 7 30 AAO k Plot amp DUALEM 421 Editi NotePad Plot Re cridding Q Filter Data Correction negative values Filter Elevation Delete negative values Figure 7 2 1 The Data Processing menu Grid files are necessary for the Q3 D inversion Data files contain in general randomly spaced measuring sites and should be converted into a regular grid before inversion To create a grid file the user must specify the algorithm and its parameters as well as the output file The output file has the mask EMGr it is assumed by default oxi Ex
8. Ibl lt 1 0 indicates a non fractal signal similar to Gaussian noise and c Ibl 0 indicates a pure white noise signal Therefore the slope can be an indication of the noise level of the data However as b is a measure of heterogeneity in a fractal signal it can be used to have general information about the geoelectrical structure of the survey area The reading of Everett and Weiss paper is recommended to the interested users Peun Sensor Plot Raw Data i Plot Filtered Data gi Plot Decimated Data Ca increments mode HCP PSD mode HCP Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response 4 Plot DOI 0 198 6 397 2 595 8 794 4 993 0 2 9 2 4 19 i43 0 8 0 3 Distance m Log wavenumber bis 1 10 b2s 0 79 b3sa 0 80 me M Cond m8 m NAN onan Nd ke WN wy Quit Ca increments mode PRP PSD mode PRP bls 0 89 b2s 0 06 b3s 0 96 Cond mS m 0 0 198 6 397 2 595 8 794 4 993 0 2 9 2 4 19 13 0 8 03 Distance m Log wavenumber Figure B1 Left increments of apparent conductivity for each sensor Right PSD of the apparent conductivity with values of b for each sensor 91 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion APPENDIX C Correction of the sites location Positioning survey sites with the use of an external GPS is a common procedure In general the GPS doe
9. Save Initial Model 1D Save Initial Model 2D Save Processed Data Data Corrected Filtered Data Save Inverse Model t 272 Resampled Data Save Inversion Misfit Decimated Data Save Model Response Meg Corrected Save EC profiles Base line Corrected Location Corrected Save All Results PRI UTM coord Save DOT Result Save Medium Prop d Figure 5 8 1 The Save menu 4 yo EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Inverted models should be save using the option Save Inverse Model These files can be imported in the Map module to construct maps showing the conductivity distributions with depth All results obtained during a session with the program can be saved in a project file PRJ The user can also save results obtained from DOI analysis as well as those obtained from Medium properties 42 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion In the Print entrance the user can print the results in a file according to a specific format PS PDF PNG TIFF WMF that can be chosen using the Metafile format option PNG 1s the default format Jave Print Help About Metafile Formak Prink Raw Data Print Processed Data Filtered Data Print Inverted Model Resampled Data Print Data Response Decimated Data Print DOI result Data Statistics Print Fitting graph Noise Analysis Print Medium Prop Megat Corrected B L Corrected Figure 5 8 2 The Print m
10. 05 a program for 1D Laterally Constrained Inversion 5 3 4 Merge Data Files This entrance allows the merge of different data files into one file The user will be asked to input the files one by one finishing input files clicking in the Cancel button The user will be asked to save the merged file Only files from the same survey 1 e data acquired with the same instrument at same height etc should be merged 5 3 5 Convert Data Format The OTHER sensor which introduced in this new release of EMA4Soil allows processing and inverting any combination of EMI sensors In this new option a specific format is used to input the data see section 9 for formats The entrance Convert Data Format allows the user to convert files written in the standard format into OTHER or vice versa aT the Conversion Sele select type Conversion Standard to QTHER OTHER to Standard Figure 5 3 2 Converting input files into another format 5 4 Input 5 4 1 Input Sensors Selecting the Sensors option enables the user to choose the appropriate EM sensor used to collect co data along a transect and to be input into EM4Soil Selecting the appropriate EM sensor allows EM4Soil to select the appropriate algorithm to invert the o data EM Laterally Constrained Inversion m Input Data Processing Inversion Input Sensors Input Data Select sites EMSounding a0 to Map Figure 5 4 1 Screen shot of the EM4Soil menu operation b
11. 581 6630 2831 000 15330 00 0000000E 00 588 4750 2894 000 15509 00 0000002E 03 0 00 0 00 Q0 0000002E 03 0 00 0 00 Q0 0000004E 03 0 00 0 00 OQ oon co wJ w 40 0 2000000 1 220000 15000 00 1 5 1 X Y Z Ca Inp Quad Sus Lat Long Time 1 320000000000000E 002 0 500000000000000 1 250000000000000E 002 1 00000000000000 1 180000000000000E 002 1 50000000000000 rm 1 E m 1 0000000E 00 552 0920 3719 000 24251 00 0000000E 00 548 5830 3698 000 24097 00 0000000E 00 556 9490 3800 000 24464 00 9999998E 03 0 00 0 00 9999998E 03 0 00 0 00 9999998E 03 0 00 0 00 ooo www j 69 yo EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion The codes for the sensors are Race Ra a EM31 MK2 E31 9800 es The codes for the mode of acquisition are 70 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 9 2 Initial model 9 2 1 Initial model for quasi 2D inversion The format for the file containing a 1D initial model is as follows NL Di Do D NL 1 C C2 ssas ONL Where NL is the number of layers D and C are the depth of the bottom interface m and the conductivity mS m of the ith layer Example of a 1D model of 3 layers with conductivities of 20 100 and 20 mS m 3 10 30 20 100 20 9 2 2 Initial model for 1D inversion of profiles The format for the file containing a 1D initial model is as follows NL Dj Do D NL 1 Ci
12. Michael Gehrig GEHRIG Inc for their comments and suggestions 94 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion USING OTHER sensors Manual This Software is produced by EM TOMO Email emtomog gmail com EMTOMO February 2013 95 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 1 Introduction The OTHER sensor is the option that allows putting together data collected with different EMI instruments and jointly inverting the data set The OTHER sensor is also used to input data to be analysed with the DMO Tool Detecting Metal Objects Tool The majority of the processing and graphical tools available for standard formats are also applied to data input as OTHER There are particularities that are explained in the following sections 2 Input Data Input the data using the entrance in the menu bar and one of the options line or area After that the program will display the main characteristics of the data in a table Figure ig 101 xl Exit Sensor Space Freq Mode PROF 1 22 3000 vom Pror 1 22 6o00 vom er e ee j Figure 2 1 Displaying the main data characteristics The red highlighted cells means that there are more than 30 of negative apparent conductivity values in the first sensor Finishing the input the program will display the data according its type data from a line or data from a survey area Exit Survey O
13. Raw Data Plot Filtered Data Data Filtered mode HCP Plot Decimated Data i Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response 62 Plot DOI 3 298 3 397 7 4971 5965 695 9 7983 894 8 994 2 Distance m Figure 5 5 6 Filtered DUALEM 421 generated using the Weighted Average The Sheppard filter is a five coefficient filter 0 0857 0 34285 0 4857 0 34285 0 0857 Figure 5 5 7 shows filtering of DUALEM 421 o using Sheppard filtering onstrained Inversion 5 Exit Survey Input DataProcessing Inversion Display Settings Save Print Help About Action Data Filtered mode HCP Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Response i aaa 298 3 397 7 4971 5965 6959 795 3 894 8 994 2 i Distance m Figure 5 5 7 Filtered DUALEM 421 generated using the Sheppard filter 17 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 5 3 Resampling The data in a line can be resampled at uniform spacing There are two options for the resampling In the first GPS option the data elevation and coordinates are interpolated The new spacing between measurements should be greater than 1 5 average spacing in the original data This is the value suggested by the program to the user This option should
14. a data set of three PROFILER lines 75 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion 11 Common Errors that must be avoided troubleshooting Select the correct sensor there is no default The format of the input files must be followed strictly The most part of the troubles are related with mistakes in the data file You can not input data only for one mode of a sensor If you only have data for one mode fill the other modes with 9 0 Take care with the number of measuring sites you have and with the right sequence of the measured values The program can verify that you have made a mistake with the number of points but it can not verify the correctness of the instrument height or if the values have been acquired in a VDM mode or with a PRP sensor When filling the boxes in the screen follow the displayed format Use only data of good quality in the inversion The final model greatly depends on the quality of your data With low quality data one cannot expect a high quality final model If your GPS does not have enough resolution you should use a local reference system Acton Plot Raw Data E j Plot Filtered Data 21 Data mode HCP Plot Decimated Data f Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI 0 0 0 0 0 0 0 0 0 0 Cond mS m Distance m t Data mode PRP Cond m8 m
15. aio e a a Cond im HAMA Ose oO 5 ha ch bi Th wT m rc i P mg o ma z D CH Mm TN Cond ims a T e nim megi I Ext Suv npt Daza Processng Iwersion Savc Mirt Hdp Abot ina namnam reweE 99 4 196 6 298 3 298 3 97 7 397 7 Data mede HCP 4971 596 5 Diszance ra Dala uuede PRP 4971 Diszance m 695 8 7959 6948 9942 2 An E D2G Figure 5 5 14 Transect of DUALEM 421 which has negative values in the 4mPcon Figure 5 5 15 shows how the negative o has been corrected and using the Correction of negative values option B T Soil 0 EM Later ally Constrained Invers Ext Suw Inpst Daza Processnq_ Tversion Save Prk Hap Abot Actinn Plot Raw Data lot Filtered Data Plul Decinialed Dala Plot Noise Analysis gt PlatWNeg Carrreted Plot Data Shifted lot Initial Mode Pluthiverted Mudel Plot Data Response Pint DOI Quit 196 6 298 3 397 7 397 7 Corrected Peta mods ACP 49 1 5965 695 5 THistanee ra Correcled Dela urdea PRP MA ptt page 4971 0965 69689 795 9 694 8 994 2 Diszance m Figure 5 5 15 Correction of DUALEM 421 which had negative values in the 4mPcon 23 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion If for some reason some of the o data has a wrong base line value Figure 5 5 16 shows where you can apply an base line correction to corr
16. batch You can introduce all the necessary information clicking in make the batch x In this option vou can input all the necessary information data Files inversion parameters and other information to invert a number of lines automatically It must be noted that before inverting the lines you must check the data All the data files you want to invert should be imported one by one after you press OK Finish clicking in the Cancel button A table contained the selected files and the default parameters will be displayed Figure 5 6 14 4 Mil Batch mode 10 x Exit 4 Figure 5 6 14 Making a batch file 35 yo EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion You must modify the wrong or not appropriated information before push the OK The program will do a basic check of your data and files where problems have been detected will be highlighted Figure 5 6 15 Batch mode NE Error in reading the Data File Please confirm the File you are reading or the Format in the File Mf Batch mode Exit Figure 5 6 15 Checking the batch file In this case the sensors assigned to the first two files are wrong After correct these mistakes the program will check again the information If every thing is well the batch can be run Table Codes to be used in the batch D a D 42 D 421 J o o D 642 J EM38 J o EMI J EM34 J GEM J PROF NOTE The code
17. codes for the sensors and the data type are summarized below Xeasting and Ynorthing are UTM coordinates or other linear system od coordinates Z is the elevation m Ca is the apparent conductivity mS m In phase and Quadrature and magnetic susceptibility components can be in ppt unit 1 or ppm unit 2 the frequency is in Hz and the mode the acquisition can be VDM mode 1 HDM mode 2 HCP mode 3 or PPR mode 4 Latitute and Longitude must be in decimal degrees Time can be expressed in any format for example 2 12 32 Example of a data file 3 sensors PROFILER PROF 40 measurements for each sensor only displayed 3 mode the acquisition is VDM the data type is 5 and the units of In phase and quadrature are ppm GSSIsurvey PROF 40 0O 2000000 1 220000 1000 000 1 5 1 X Y Z Ca Inp Quad Sus Lat Long Time 1 320000000000000E 002 0 500000000000000 1 250000000000000 002 1 00000000000000 1 180000000000000E 002 1 50000000000000 0000000E 00 45 02300 2632 000 132 0000 0000000E 00 518 5630 1378 000 1519 000 0000000E 00 1007 727 663 0000 2951 000 0000002E 03 0 00 0 00 Q 0000002E 03 0 00 0 00 OQ 0000001E 03 0 00 0 00 O oon he PROF 40 O 2000000 1 220000 9000 000 1 5 1 KX Y Z Ca Inp Quad Sus Lat Long Time 1 320000000000000E 002 0 500000000000000 1 250000000000000E 002 1 00000000000000 1 180000000000000E 002 1 50000000000000 0000000E 00 588 3160 2823 000 15505 00 0000000E 00
18. depth investigation in EM is a maximum depth at which a given target in a given host can be detected by a given sensor The depth of investigation depends on several factors such as sensor sensitivity operating frequencies target and host properties noise level etc In LIN instruments like those from GEONICS or DUALEM the depth of investigation is quite well established However the issue is not clear for multifrequency instruments like GEM and PROFILER For that reason this tool is only available for multifrequency sensors Clicking in Depth of investigation entrance will open the new module Figure 5 6 17 E EM4Soil Depth of investigation module Ioj x Exit Display Input Calculation Priok About Figure 5 6 17 The Depth of investigation module menu bar Clicking Input will open the menu Figure 5 6 18 for input of the testing model and relevant parameters Figures 5 6 19 and 5 6 20 37 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion W EM45oil Depth of investigation module Exit Display Input Calculation Print About Model Parameters Figure 5 6 18 Choosing the testing model input E Testing Model Ioj x Figure 5 6 18 Defining the testing model Depths are in meter and conductivity in mS m 15x Exit frequency 1 frequency 2 detect threshold 7 depth increasing 7s OK Figure 5 6 20 Choosing the maximum and minimum frequencies the detecting
19. few layers 2 or 3 and the inversion is performed at each measuring site without any constraint This option is useful when dealing with data collected with one sensor only in both modes of acquisition The inversion algorithm is based on a least square method using singular value decomposition SVD to solve the equation system The solutions is stabilized using the Levemberg Marquardt algorithm Johansen 1977 A description of the algorithm can be found at Johansen 1977 In this method we try to find a set of earth parameters pi p2 p gt m that minimize an error function This problem is nonlinear and its solution requires a linearization procedure near the correct solution in order to obtain an improved solution The nonlinear problem is linearized using the Taylor series expansion d 5 J P 1 where d is an N dimensional vector containing the data y is an N dimensional vector containing the theoretical responses of a model with parameters specified in the M dimensional vector p and p is a M dimensional vector containing the unknown correction of the parameter J is an N x M Jacobian matrix defined J 2 The Pj parameter corrections are calculated by solving equation 1 using the Singular Value Decomposition SVD technique Golub and Reinsch 1970 ACKNOWLEDGMENTS Several people have contributed indirectly for this work We are particularly grateful to Richard Taylor Dualem Inc Scott Holladay Geosensors Inc and
20. of layers and depth of the different initial models are as follows the conductivities are defined according the data For DUALEM 2I1s 7 0 3 0 6 1 0 1 5 2 3 3 3 For DUALEM 421s 9 0 3 0 6 1 0 1 5 2 3 3 3 4 5 5 7 For DUALEM 642s 8 0 6 1 5 2 7 3 8 4 2 5 7 8 0 For EM34 10 3 6 9 12 18 24 30 40 50 NOTE it must be noted that these models are very general and might not be the best approach for some particular data set The user should define his models according the information available for the area in investigation 72 10 Output File Format 10 1 Data Model response file The containing DUALEM GEONICS the EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion data model response has a format Dist rhlo rhlc rplo rplc rh2o rh2c rp20 rp2c rh4o rh4c rp4o rp4c S hyo hic pio pic h20 hoc p20 pec hyo hyc p4o pac S2 hyo hic p10 pic h20 hzc p20 prc hyo hyc p40 pac Sn h10 hic p10 pic h20 hzc p20 pac hyo hyc p4o p4c aS follow Where S is the distance along the line Subscript o indicates an observed value measured and the subscript c indicates a calculated apparent conductivity value h and p represent the HCP and PRP modes in DUALEM sensors for GEONICS they correspond to VDM and HCM respectively Example of a file from EM34 dist vdmo vdmc hdmo hdmc vdmo 00 00 47 50 69 89 73 20 81 49 56 80 10 00 53
21. y OK 18 0 22 5 270 31 5 36 0 40 5 Distance m Figure A1 Apparent conductivity over a two layer model conductivity of 10 and 85 mS m and thickness of 1 m 85 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Inverting the data using a 1D model program the solution depends on the initial model Table 1 These results show that the ambiguity of the inversion of these data can be important Table 1 Results if the 1D inversion for two different initial models Initial model Initial model CI mS m Figures A2 and A3 shows the result obtained by EM4Soil using the same data set The results of the 1D LCI inversions are quite good Due to the lateral constraints the solution at the ends of the profile 1s worst Both solutions show a conductivity contrast at a depth of 1 1 1 m However the conductivities are not well resolved Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Calculated Model Plot Data Response Plot DOI Quit Elevation m l Pe 0 9 I am L x 0 0 4 5 3 0 13 5 180 225 270 315 360 405 465 0 Distance m Figure A2 Model obtained by EM4Soil using the linear approach misfit of 4 3 after 10 iterations Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Calculated Model Plot Dat
22. 00 0 010 0 060 0 090 0 000000 0 000000 0 0 0086 1 0000 0 00 2601 390 1467 510 1166 490 2526 000 21795 000 33378 000 7618 000 38677 000 51238 000 0 010 0 060 0 090 0 000000 0 000000 0 0 0093 1 5000 0 00 2640 600 1497 530 1192 360 2671 000 22137 000 33846 000 7733 000 39468 000 52375 000 0 010 0 060 0 090 0 000000 0 000000 0 Data for EM sounding For GEONICS instruments EM31 EM38 the data format for EM soundings is as follow soundingName N ho H1 VDM1 HDM1 H2 VDM2 HDM2 HN VDMN HDMN For DUALEM instruments S1 S2 S21 the format is this is an example for DUAL EM S21 soundingName N ho H1 HCP1 1m PRP1 1m HCP1 2m PRP1 2m H2 HCP2 1m PRP2 1m HCP2 2m PRP2 2m HN HCPN Im PRPN Im HCPN 2m PRPN 2m Where N is the number of measurements Hi is the height of the sensor m and ho must be 0 0 HCPi and PRPi or VDM and HDM are the measured apparent conductivity at each sensor Lat and Lon are the Latitude and the Longitude in decimal degrees and Time is the time in any format not exceeding 14 characters For GEM or PROFILER the format should be as follow soundingName N ho Nf Nop fl f2 fNF 67 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion H1 cond1f1 cond1f2 condl NF H2 cond2f1 cond2f2 cond2 NF HN condNf1 condNf2 condNfNF Where N is the number of measurements Hi is the height of the sensor m Nop is 1 if the data was acqui
23. 1 h 3 Fi m Cond m ACUUIL Plot Raw Data Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Cond mS Plot DOI Quit 159 3 j 142 3 T 125 3 108 3 913 74 3 57 2 40 2 D 1 Nit 20 ate e 16 710 8 Data mode HCP 5li 2 639 0 7668 884 6 1022 4 1150 24 1278 0 Distance m Data mode PRP 5112 6390 7668 8946 10224 1150 2 Distance m Data Response mode HCP yea naaa oO mph OSEBAG AAA Ns LA Fai Pa Eira a 200 4 8826 5104 638 0 7656 893 2 1020 8 1148 4 1276 0 Distance m Data Response mode PRP of F a p 4 cae plies t ste in werent A oe 55 2 382 8 510 4 638 0 765 6 Distance m Figure 12 6 Decimated apparent conductivity curves and model responses Figure 12 8 shows the model calculated using the full solution The model response is shown in Figure 12 9 79 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion EM4Soll v0 EM Laterally Constrained In Joj x Fxil Srey mal DaaPu si Teveisie See Oral Hel Ahal Action Ilot law Data Plut Fillered Dala Plot Decimated Data Pini Noise Analysis Plot Neq Corrected Hoi Data thitted Plut Initial Mudel pace aE Caleululed Mudal Plot DOI Quit im gu 382 A F104 68A0 7666 ADAB 1020 8 1148 4 1276 0 Distarce fm Figure 12 7 Conductivity
24. 25 09 26 46 27 63 33 66 33 04 33 29 34 04 rp4c 16 61 24 59 25 88 27 05 32 48 31 73 32 08 32 53 73 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 10 2 Processed data Filtered This format is common to all processed data and is the same of the input data file to allow to be read as input file for DUALEM 421 for example is as follow Name of line N h X Y Z HCPIlm PRPIm HCP2m PRP2m HCP4m PRP4m Lat Lon Time 10 3 Initial Model The file of the initial model generated automatically used in 1D laterally inversion has the format as follow N NL S Di Ci S2 D2 C2 Sni gt Dyu CNL Where N is the number of sites along the line NL is the number of layers D and C are the depth m and the conductivity mS m of the ith layer S is the distance along the line of the ith site having the first measurement as an origin NOTE this file cannot be used as input file 10 4 Final Model The file contains an interpolation of the final model The format of the file is as follow Xcoord Ycoord Distance Elevation mS m ohm m X Yi Si Zi Ci f X2 Y2 82 Z2 Co r2 XNL gt YNL SNL ZNL gt CNL INL Where X Yj represent the easting and northing coordinates and Zj the elevation or thedepth in a flat earth S is the distance between measuring sites along the line having the first measurement as the origin Cj and r are the conductivity and resistiv
25. 50 72 51 73 50 80 71 63 00 20 00 58 90 70 56 73 90 73 52 69 20 30 00 65 50 69 09 67 60 66 76 67 80 40 00 68 40 68 01 61 50 62 47 72 70 50 00 63 90 72 69 71 40 71 83 72 40 vdme hdmo hdmc vdmo vdmc hdmo hdmc 53 77 99 80 71 42 34 10 36 62 68 80 57 76 56 50 95 50 72 55 33 80 36 22 64 10 58 87 57 47 83 30 68 93 35 90 36 95 48 40 57 57 59 33 74 50 65 87 39 80 39 02 45 20 57 13 59 84 64 60 63 72 36 50 39 28 49 10 56 28 60 50 90 40 69 55 38 60 38 23 42 80 58 95 Example of a file from DUALEMA42 The value 9 means that there is no data for the lm sensors dist rhlo 0 9 2 9 4 9 6 9 8 9 9 9 11 24 9 13 24 9 rhlic 9 9 9 9 9 9 9 9 Example of a file from PROFILER dist 0 00 0 19 1 24 1 80 3 09 4 46 rh4o 20 16 22 16 23 08 23 92 29 48 30 12 31 46 31 98 rh4c 18 5 21 93 22 12 23 71 29 44 29 77 30 77 31 33 Cflo 134 06 140 18 149 33 155 70 164 35 165 78 rplo rple rh2o0 rh2c rp2o rp2c 9 9 15 27 17 52 17 91 16 82 9 9 22 92 23 68 27 9 27 88 9 9 24 09 24 89 28 7 28 8 9 9 25 06 26 08 29 39 29 22 9 9 31 54 31 87 32 76 33 51 9 9 29 95 31 52 31 59 31 72 9 9 30 37 32 13 31 5 31 57 9 9 30 68 32 63 31 73 31 88 Cflc Cf20 Cf2c Cf30 127 44 113 76 118 96 124 74 143 05 124 02 133 20 133 75 150 39 133 92 139 98 142 25 157 15 141 39 146 21 148 59 163 76 146 30 152 25 154 00 168 39 151 41 156 44 158 04 Cf3c 115 34 128 98 135 51 141 51 147 28 151 28 rp4o 16 32
26. 9 00 99 6 23 6 69 9 0 0 0 0 0 0 For GEM PROFILER LinePROF N h Nop Nf f1 2 Nf X Y Z condfl condf2 condfNf InPf1 InPfNf Qf1 QfNf Susfl SusfNf Lat Lon Time Sequence of values row by row Where N is the number of measuring sites h is the height of the sensor m Nop is 1 if the data was acquired in VDM mode or 2 if acquired in HDM mode Nf is the number of frequencies and fl f Nf are the frequencies of the survey in Hz X Y and Z are the easting and northing linear coordinates and the elevation m of each measuring site respectively condfi InPi Qfi and Susfi are the measured apparent conductivity mS m InPhase ppm Quadrature ppm and Susceptibility for the 1 th frequency 66 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion NOTE Unlike the LIN instruments if you have all data corresponding to one frequency with negative values it is advisable do not use this frequency Example of a data file only the first rows of the file are shown corresponding to a line of 20 sites measured with a PROFILER instrument at 0 2 m above the surface The data was collected at three frequencies 1000 9000 and 15000 Hz in VDM mode linelm 20 0 2000000 1 3 1000 000 9000 000 15000 00 X Y Z Cafl Caf2 Caf3 Inpf1 Inpf2 Inpf3 Qudf1 Qudf2 Qudf3 susf1 susf2 susf3 Lat Long time 0 0078 0 5000 0 00 2616 530 1460 490 1160 610 2761 000 22007 000 33481 000 7662 000 38492 000 50980 0
27. DUALEM 421 o data prior to filtering The Running Average filter is a three point moving average filter whereby a moving average filter averages a number of input o and produces a single output value Figure 5 5 5 indicates how the DUALEM 421 o data shown in Figure 5 5 4 appears after filtering and using the Running Average 16 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion Ml EM4Soil v0 EM Laterally Constrained Inversio 10 Exit Survey Input Data Processing Inversion Save Print Help About Action Plot Raw Data Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response a 9 98 3 397 7 4971 5965 6959 795 3 8948 994 2 Distance m Data Filtered mode HCP Figure 5 5 5 Filtered DUALEM 421 generated using the Running Average The Weighted Average filter is a three coefficient filter 0 3 0 4 0 3 The weight is calculated taking into account the distance and value of the closest non negative values The Weighted Average filter should be used carefully and cannot be applied if the number of negative data exceeds 30 of the total values per channel Figure 5 5 6 shows an example of DUALEM 421 o post filtering using the Weighted Average It should be noted that the correction of a large sector of negative values will produce poor results 15 Action Plot
28. EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion EM4Soil A program for 1 D laterally constrained inversion of EM data Version 1 05 This Software is produced by EM TOMO with the collaboration of John Triantafilis Email emtomog gmail com May 2013 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion CONTENTS 1 Introduction 2 EM4Soil Packed Items 3 EM4Soil Installation 4 Running EM4Soil 5 Menu operations bar 5 1 Exit 5 2 Survey 5 3 Open 5 4 Input 5 5 Data Processing 5 5 1 Editing 5 5 2 Filtering 5 5 3 Resampling 5 5 4 Decimation 5 5 5 Data Statistics 5 5 6 Noise Analysis 5 5 7 Pseudo section E Ca Pz 5 5 8 Data corrections 5 5 9 Converting Lat Lon to UTM 5 5 10 Medium properties 5 6 Inversion 5 7 Display Settings 5 8 Save and Print 5 9 Help 6 The Action Display zone 7 The EM4Soil Map module program 7 1 Display 7 2 Data Processing 7 3 Profile mode 7 4 Inversion 7 5 Zoom 7 6 Save 7 7 Print 7 8 Help 8 EM Vertical Sounding module program 9 Input Files Format 10 Output Files Format 11 Common Errors that must be avoided troubleshooting 12 Examples References and useful bibliography APPENDIX A Inversion of EM data APPENDIX B Noise analysis APPENDIX C Correction of the sites location APPENDIX D The Simulated Annealing Method APPENDIX E The 1D inversion Using OTHER sensors EMA4Soil v1 05 a program fo
29. Initial Model F Figure 8 6 Inputting the initial model 6l EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion momoa Exit Manually 1 D C Open File 1 0 l0 x Figure 8 8 Defining the initial model manually This version of EMSound module only has implemented the SA Simulated Annealing algorithm Figure 8 9 Please read APPENDIX D about Simulated Annealing method W EM4Soil Sounding module Exit Display Input Inversion Save Print About Inversion parameters 54 i Inversion 54 Figure 8 9 The Inversion options The user needs to define the parameters to use in SA inversion Figure 8 10 the number of iterations in the external loop the temperature decrease with the number of iterations the number of iteration of the internal loop the iterations of the internal loop are done with a constant temperature the initial temperature the number of models these models will be used to build the correlation of the parameters and an average model This information can be displayed as tables in the Display entrance of the menu bar the target misfit and the approach to be used for forward calculations 62 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion O x Eyit Nit Ext Loop Nit Int Loop Temperature Num Wodels arit Target Misfit Linear Solution C Full Solution OK ee
30. The Metropolis algorithm iterates Over a sequence of models at a constant T value Internal loop This renders the solution independent of the initial model and allows the algorithm to escape from local minima Looping over the Metropolis algorithm while T decreases there is also a decrease in the acceptance probability eq 1 and only perturbations decreasing the objective function are accepted Therefore it is expected that the accepted models will concentrate in the vicinity of the global minimum of the objective function A slowly decreasing T parameter External loop is important in the efficiency of the SA algorithm allowing a representative sampling of the parameters space The simulate annealing SA algorithm was adopted to perform the 1 D inversion of EM soundings The objective function based on norm and involving M conductivity data is defined as N 2 gt ld y i Pm WO ON did y 1 di ld y i l i 2 where N is the number of measuring points The value of the parameter T was taken as 10 times the initial value of the objective function The number of iteration external and internal loops is usually of 100 allowing each parameter to change more than 20 000 times 93 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion APPENDIX E THE 1D INVERSION EMA4Soil allows the inversion of the data collected along a line assuming a layered model In this case the model may have a
31. a Response Plot DOI Quit B m w m ONO GD YS f n Yn F QUI puo 80 v4 67 Elevation m i t F omuw oo 36 0 405 465 0 Figure A2 Model obtained by EM4Soil using the non linear approach misfit of 4 8 after 3 iterations 86 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Inversion algorithm The discrete inverse problem can be formulated as follows calculate M parameters of the model values of conductivity and thickness of the layers in 1D models knowing N values of apparent conductivity measured on surface The nonlinear smoothness constrained inversion algorithm described by Monteiro Santos 2004 was adopted in EM4Soil The earth model used in the inversion process consists in a set of 1D models distributed according to the locations of the measurement sites All the models have number of layers whose thickness is kept constant Two forward modeling subroutines one based on the cumulative response McNeill 1980 Wait 1962 and another based on the full solution of the Maxwell equations Kaufman and Keller 1983 are used optionally to calculate the o responses of the model Two inversion algorithms are available in the S1 algorithm the optimization equations are represented as follows Sasaki 1989 I S AC C p J b 1 In the second algorithm S2 the equations are Sasaki 2001 J J AC C p J b C C p po 2 where p is the vector containi
32. a Statistics Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI Quit Figure 4 1 Screen shot of the EM4Soil main screen Figure 4 2 shows the welcome page of EM4Soil has 10 operational menu buttons Me EM4S5oil 104 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Save Print Help About Figure 4 2 Screen shot of the EM4Soil menu operation buttons In addition Figure 4 3 shows the 11 plotting actions which appear along the left hand side of the welcome page of EM4Soil The rest of the welcome page allows the user amongst other things to plot data e g original and filtered and modelling results e g initial and inverted to be displayed The application s window can be resized using the option Window Size in the Display Settings menu bar entrance Please see details in 5 7 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Action Plot Raw Data Plot Filtered Data Plot Resampled Data Plot Decimated Data Plot Data Statistics Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI Quit Figure 4 3 Screen shot of the EM4Soil plotting action buttons vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 Menu operations ba
33. ase ppm 20 Distance m Apparent Susceptibility 20 Distance m Figure 3 2 2b Data resampled spacing 1 5 m 3 3 Inversion To invert OTHER data use the option Inversion OTHER Figure 3 3 1 To invert the data it is necessary to input an initial model using the appropriated entrance in the menu Figure 3 3 1 Note that in this case the Automatic option is not available due to the variability of sensors that can be used trained Inversion ji a ef anc YL j T Jata Processing Inversion Display Settings Save Print Help About Parameters Depth of investigation Input initial model Q25 Inversion Q2D Inversion Calculation CF iD Inversion Calculation FS Parameters Inversion OTHER Input initial model 1D p Inversion 1D Appraisal of Inversion Batch mode Figure 3 3 1 Inversion of OTHER data with Q2D approach To construct the initial model the information gave by the Depth of investigation tool can be useful The initial values of the magnetic susceptibility can be calculated taking into account that the susceptibility in ppt is gave by 99 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Xa 9 002 Un phasein ppm The program will ask to choose the data REE Exit Choose data for inversion RAW DATA FILTERED DATA NEG CORRECTED RESAMPLED DATA Figure 3 3 2 Data to invert
34. be used if data are not uniformly sampled but the coordinates are good erally Constrained Inversion BEW pai version foom Display Setting Edit Editi MotePad Filter Resampling Resampling GPS Decimation Local coordinates Data Statistics Noise Analysis Section ECa Pz Correction of Lat Lon to UTM Medium Properties Resampling DAR Resampling options mmg ns pont Exit Xx local m Resampling options Y local mJ Linear spacing C Inverse distance spacing EALA spacing m aui OK O OK e In the second option Local Coordinates the coordinates in the data file are not considered New coordinates will be constructed and data and elevation values will be associated to this new system This option should be used when the coordinates are not credible and the measurements have been done in positions referred in a tape for example 18 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 5 4 Decimation Decimation allows the EM4Soil software to ignore o data measured along a transect Decimation may be required to allow a user to account for a high sampling rate or redundancy in the o data The user will be asked about the decimation factor which represents the number of samples that will be skipped Figure 5 5 8 shows an example of EM34 o data prior to and post Decimation and using a Decimation factor of 2 This means that only every third EM34 sens
35. choices about how to display your maps W EM45oil Map module Exit Display Data Processing Profile mode Invers SSS ne Survey layout Raw Data Cond Resistivity Filtered Data Topography Sensor O2D Model Slices OSD Model Settings agaa window size Figure 7 1 1 Displaying data from a survey Figure 7 1 2 shows an example of the display of a PROFILER EMP400 survey with three profiles The first site is marked in red Sites with negative values of apparent conductivity are displayed in blue Figure 7 1 3 shows an example of the display of a map of apparent conductivity that was acquired with a multifrequency equipment PROFILER To move through the other maps you must click on the right button of your mouse The Window size option is very similar to that explained in the first part of this manual The Q3D Model option is devoted to display Q3D inversion results and it is explained in the Manual for the Q3D Module gt We thank Mr Michael Gehrig P G P E President of the GEHRIG Inc for kindly allowed us to use some of his PROFILER data in this Manual 45 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion E EM41Soil Map madule oj x Exi Dis lay Profile mode Save Prnt Help About Quii 2 0 T A Distance X m Figure 7 1 2 Displaying your survey E TH Soil Map module iDa Ext Depla Molene Scye rint cla Aut Quit Raw Data FMD Map of cordus
36. d Technical Note TN 6 Monteiro Santos F A 2004 1 D laterally constrained inversion of EM34 profiling data J Appl Geophys 56 123 134 Monteiro Santos FA Triantafilis J Taylor R Holladay S and Bruzgulis K 2010 Inversion of conductivity profiles using full solution and a 1 D laterally constrained algorithm Journal of Envirnmental and Engineering Geophysics 15 3 163 174 83 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Monteiro Santos FA Triantafilis J Bruzgulis K and Roe J A E 2010 Inversion of multiconfiguration electromagnetic DUALEM 421 profiling data using a one dimensional laterally constrained algorithm Vadose Zone Journal 9 1117 125 doi 10 2136 vzj2009 0088 Sasaki Y 1989 Two dimensional joint inversion of magnetotelluric and dipole dipole resistivity data Geophysics 54 254 262 Sasaki Y 2001 Full 3 D inversion of electromagnetic data on PC Journal of Applied Geophysics 46 45 54 Triantafilis J Roe J A E Monteiro Santos F A 2011 Detecting a landfill leachate plume using a DUALEM 421 and a laterally constrained inversion model Soil Use and Manamgement in review Triantafilis J Monteiro Santos F A 2010 Resolving the spatial distribution of the true electrical conductivity with depth using EM38 and EM31 signal data and a laterally constrained inversion model Australian Journal of Soil Research 48 434 446 Triantafilis J Mon
37. e user must take into account that this pseudo section is not a model and can not be used for interpretation 21 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 5 8 Data corrections Individual measurements along a transect inputted into the EM4Soil software can be corrected using the Correction of values option Figure 5 5 12 In the first instance the 6 data to be corrected should already be plotted and visible on the screen Me EM4Soil 103 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Save Edit Editi NotePad Plot Raw Data Filter Resampling Plot Filtered Data Decimation Data Statistics Plot Resampled Da pjojse Analysis Action values manual Correction of __Plot Decimated Da Lat Lon to UTM negative values base line Plot Data Statistic ____ edi Properties F elevation d sites location GPS Plot Noise Anavsis i Figure 5 5 12 Drop down box menu for data Correction of o data options Once the Correction of values option is selected the mouse arrow icon will change to a cross Go to the sites along the transect and select the o data you want to correct Click on the left mouse button A second third and fourth click will allow the user to select multiple sites In order to finish the selection of sites click on the right mouse button As shown in Figure 5 5 13 a Table appears Negative and n
38. ect any channel E Base line correction Ioj x Exit Correction value mS m Data to be corrected a HCP E PRP Iw 1m 2m 4m Figure 5 5 16 Drop down box menu for base line correction option Example is for DUALEM 421 The Elevation correction option allows a user to display the elevation along the transect line of interest and active within in EM4soil and thereby smooth it moving average filters of 3 7 and 11 points are available It 1s also possible to disable the elevation When this option is chosen subsequent inversion modelling will be assumed data to have been collected across a flat terrain Me EM4S5oil 103 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Save Print Help Edit Editi NotePad Plot Raw Data Filter Resampling Plot Filtered Data Decimation Data Statistics Plot Resampled Da _ Analysis Action Correction of values manually Lat Lon to UTM negative values Plot Decimated Da base line Plot Data Statistic al eae display smooth Flat Medium Properties elevation sites location s Plot Noise Analysis Figure 5 5 17 Drop down box menu for elevation altitude corrections A GPS receiver is not usually installed near the centre of an EM instrument and during a survey Therefore the position of the EM o should be corrected The correction of site location option allows doing t
39. enu 5 9 Help You can read a summary of the main function of the program including a Short Guide on how to use this program using the Help menu W EM4Soil v1 EM Laterally Constrained Inversion Exit Survey Input Data Processing Inversion Display Settings Save Print Help About erry Short Guide Sensors Plot Raw Data Data Initial model Plot Filtered Data Initial Model File Formats Data File Formats Plot Decimated Data pista braeessine g Inversion Plot Noise Analysis Input Inversion Results Display Settings Plot Neg Corrected a Save Plot B L Corrected Print Figure 5 9 1 The Help menu 6 The Action Display zone Your data and your results can be displayed on the screen using the buttons available in the Action zone The program will be closed clicking in the Quit button Closing the program using the Exit option in the Main top menu bar will delete all temporary files saved during the use of the program Anyway these files will be deleted next time you start the program or when you read a new data file 43 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 7 The EM4Soil Map module program If your data was acquired in a survey constituted by several profiles covering an area you can import them using the Input Data Area option Figure 7 1 The format of the data file is the same as that for profiles but including all measuring sites Please see the Data Fo
40. eply will include the required input KEY which once entered will enable EM4Soil to run EM4Soil 0 EM Laterally Constrained Inve rs x WRONG KEY ASK FOR THE KEY YOUR CODE IS Exo ot o send an email with this code to emtomog agrmail com Figure 3 1 Screen snapshots which show the EM4Soil CODE corresponding to your software package and where to input the key once the CODE has been sent NOTE The key links the program to a particular computer A new key is necessary to run EM4Soil on another computer NOTE Save the Key in a safe place You may need it in the future NOTE Make a copy of all files in a safe location NOTE All screens displaying the OK button must be closed after clicking on it NOTE To run the EM4Soil in Windows7 could be necessary to redefine Compatibility properties of the program Please go to section 11 of this manual Avoid folder names with spaces like C EMTOMO EM4S vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 4 Running EM4Soil In order to run the EM4Soil software double click on the EM4Soil icon EM4Soil will start up and the screen snapshot shown in Figure 4 1 will appear This is the welcome page E EM4Soil v103 EM Laterally Constrained Inversion o x Exit Survey Open Input DataProcessing Inversion Display Settings Save Print Help About Action Plot Raw Data Plot Filtered Data Plot Resampled Data Plot Decimated Data Plot Dat
41. epth of investigation depends only on the transmitter receiver separation and not on the frequency or o Kaufman and Keller 1983 It is then possible to construct a mathematical function of depth that describes the relative contribution to 87 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion the secondary magnetic field measured at the receiver due to the homogeneous material within a thin horizontal layer at a depth d McNeill 1980 Kaufman and Keller 1983 Gomez Trevino et al 2002 Callegary et al 2007 It is worth mentioning that this approach is only valid for fairly resistive environments For highly conductive structures the instrument response is not linear and the use of the cumulative function will produce biased models the conductivity of deeper layers will be underestimated if covered by highly conductive layers In the presence of a layered earth model the relative contribution to the secondary magnetic field from all material up to a depth d below the sensor can be expressed by the cumulative function R for HCP or PRP configurations used in the DUALEM instruments as defined the McNeill 1980 and Wait 1962 Rp l 4 HCP ae a a 2 Rorre 5 4z 1 where z d s represents the depth normalized by the coil spacing s Taking into account these definitions the response of an M layer earth is calculated by adding the contribution from each layer independently weighted according
42. erted Mudel i Calerlatec Medel Plot Data Response Pint DOI Quit 192 9 6C2 4 104 4 102 4 4 a an ka J pe gt ot E 19I IRI A 16a 0 0 98 5 1970 295 5 3940 4925 59311 6896 709 1 886 6 935 1 Distanze m Figure 1 2 Inversion result using the linear approach Q2D The inversion procedure used in EMA4Soil is a 1 dimensional laterally constrained technique 1 D LCI It is also known as a Quasi 2D Q2D inversion The forward modelling of the EM4Soil software is based upon the cumulative function McNeil 1980 or on the full solution of EM fields in a layered earth Keller and Frischknecht 1996 The inversion algorithm is based upon the Occam regularization method e g DeGroot and Constable 1990 Sasaki 1989 Figure 1 2 shows the result of inversion using the data shown in Figure 1 1 The package has a Map Module that allows the user to display the survey and to choose the profiles for inversion vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 2 EM4Soil Packed Items The EM4Soil package is a 32 bit application that can run on Windows XP Vista or even in the 64 bits Windows seven EM4Soil has a graphical user interface based on the DISLIN graphics library http www dislin de The full package contains the files EM4Soil v105 exe EMSmap exe EMSound exe EMDepinv exe only for GEM and PROFILER InvEMQ2D vSp0 exe InvFDEMQ2D v0 Inv1D exe InvEM1D v0
43. exe JointEM1D v0 exe JointEMQ2D vSp0 exe InvEMQ3D vSp exe InvEMQ3D 2L vSp0 exe JointEMQ3D vSp exe JoiIntEMQ3D SMvsP exe INSTEM4Soil BAT PathEMSGen exe disdll dll and disdll_d dll these are the dynamic link libraries for the DISLIN graphics this Manual Instructions to run EM4Soil program and some example of the input files txt files required Files related to the Q3D marked with are only available if the Q3D module was purchased DISLIN is written by Helmut Michels Max Planck Institut fuer onnensystemforschung Katlenburg Lindau 1986 2009 All rights reserved EMTOMO Licence No 201118100101 l See section 11 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 3 EM4Soil Installation In order to install the EM4Soil software take the following steps 1 Open a new folder for the program on your computer for example C EMTOMO 2 Copy all package files into the new folder 3 Run the batch file INSTEM4S BAT 4 Run the EM4Soil software by clicking twice on the icon The first time a user runs the EM4Soil software the user will be asked to provide an alpha numeric key see left hand panel of Figure 3 1 In order to obtain this key the user will need to send an email to emtomog gmail com In the email the CODE indicated by the pop up box generated once EMA4soil is first run needs to be provided In the example shown in Figure 3 1 the alphanumeric key is EX8C545 The email r
44. he same used for selection of sites for EMsounding inversion see 5 4 4 Select the sites using the left mouse button and finish clicking on the right button Display Model Profiles Select a site rt 1 07 oO a r a iy Calculated Model m R e W gT pueqjo de rt ky Elevation im a He NNN ga co ga pa ROR P PO P Oe OOo e b en 32 fal 10 12 Distance m Figure 5 6 13 Vertical conductivity profiles and their location in the Q2D model The results can be saved in a file To do that go to the menubar save save EC profiles 34 yo EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Each time an inversion is calculated a file with extension INVr is saved in the folder inverse These files can be imported using the option Inversion results in the Open entrance The Batch mode allows you to define a batch for inversion of several lines automatically The Batch mode is not available for the Q3D inversion module Inversion zoom Display Settings Save Print Help 4 amp 4 Parameters Depth of investigation Input initial model 2D Inversion Q20 a 1D Parameters Input initial model 10 Inversion 1D Appraisal of Inversion Batch mode make the batch run the batch retrieve batch results save batch read batch There are five options in the batch mode 1 make a batch 2 run a batch 3 retrieve results 4 save a batch and 5 read a
45. his correction as well the correct positioning of the o of the different sensors This is particularly the case in a multisensor instrument such as the DUALEM 421 24 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Figure 5 5 18 will appear when the GPS correction option is selected This allows the instrument orientation the GPS location and the method of GPS correction to the o data collected by each sensor This last option is valid only for DUALEM instruments Please read APPENDIX C for further details NOTE Site correction should be applied at profiles It can be used also in mapping surveys if data was collected in continuous parallel profiles bidirectional acquisition 7 Note that the correction assumes that the transmitter loop is behind receivers during the acquisition lox Exit Instrument orientation in line GPS position gut of instrument C builtin GPS DUALE hi Se d h Interpolation DUALE M Inverse distance C Nearestneighbor OF Figure 5 5 18 Drop down box menu for GPS correction option 5 5 9 Converting Latitude and Longitude to UTM The Lat Long to UTM option tool allows the conversion of GPS data acquired in Latitude and Longitude into UTM coordinates This conversion is possible for five datum selecting the appropriate central meridian of the zone The calculations are based on the formulas presented by Steven Dutch http www uwgb edu dutchs usefuldata
46. ic field measured at a site with coordinates x y h over a N layer model are given by Keller and Frischknecht 1966 m H T A B And AB m x L T A B And r it XS 13 where m is the magnetic moment of the source and r is the transmitter receiver distance A B and are given by A h o B rl lo 0 2 0 UU 14 and T R 8 g e J 8B dg 0 T R g8 ge J gB dg 0 15 89 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Here Jo and J are Bessel functions of the first kind of order O and 1 respectively Ro g is calculated recursively taking into account the conductivity and thickness of each layer Integrals in equations 9 are evaluated using a subroutine developed by Anderson 1979 The predicted values of the Quadrature component for HCP and PRP configurations at height h above the ground are given by Q ucr E Im 7 H O prr IMH H 16 90 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion APPENDIX B Noise analysis Everett and Weiss 2002 have shown that the Power Spectral Density PSD of apparent conductivity measured with LIN instruments follows a power law of the form Acw w where w is the spatial wavenumber and b is the slope of a best fit straight line in a log log graphic Those authors shown that a values of lbl gt 1 0 indicates a non Gaussian fractal signal b
47. igure 3 1 NON N or n can be used to say that a channel is not to be filtered F1 F2 and F3 are used to define the filter to be applied Running Average Weighted Average and Sheppard respectively C io x Figure 3 1 1 Selecting the filter and the sensor to be filtered 97 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 3 2 Resampling This option allows the resampling of the measured line at a different spacing from that used in data acquisition Figure 3 2 1 and 3 2 2 Lines acquired at different time and at slightly different positions can be put together in the OTHER format However the joint inversion of this data set only can be done if a common reference exists for all lines This can be done resampling the data with spacing 0 0 fi Data Resampling Ioj x Exit Resampling options Linear opacing ic Figure 3 2 1 Resampling data lines no name Resampled Data Apparent Conductivity j P N ond m3 m mi an E 2m 20 Distance m In Phase and Quadrature 2600 000 1606 1000 600 0 600 1000 1600 2000 In Phase ppm 20 Distance m Apparent Susceptibility 20 Distance m Figure 3 2 2a Original data 98 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion lins no name Resampled Data Apparent Conductivity 20 Distance m In Phase and Quadrature In Ph
48. ion around the profile defined through the search radius wish is defined by the user The value of this radius should be less than the distance between measuring sites MS EM4Soil Map module Exit Inversion Display Data Processing oii yee Add profile Add profile to Plot Survey Delete profile See ee Bee ee ee ee ee Figure 7 3 1 Adding Deleting a profile to the survey E EM4Soil Map modu Exit 101 x Profile ends xe ve Profile Search radius m i New Profile YES C NO Figure 7 3 2 Adding a profile to the survey After close the small menu clicking OK the screen will present the sites along the profile in red and an arrow indicating the sense of the profile Figure 7 3 3 It means that a coordinate 0 will be assigned to the site at the begging of the arrow The other sites will have coordinates according its distance from the first site The user must take into account the distance between sites in his survey for correctly choosing the search radius The search radius that appears in the menu is proposed by the program and roughly corresponds to the half value of the average distance between sites of the survey However the user should verify if this value is correct for his application 52 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E EM1Soil Map madule B b Ioj x Exi Dis lay Profile mode Save Print Help Abo
49. ission of Drs Rick Taylor DUALEM and Scott Holladay GEOSENSORS Inc TI yy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E TH Soil Map module 10 xj Ext Dapy Moiemce Scye rint cl gt Aut Quit Raw Data oO m jm jan 5 Ue Distance f im Ur IFAD 276 2 a7AS BALI lhistance X ra Figure 12 3 Apparent conductivity map for sensor PRP 6 m Figures 12 2 and 12 3 show the apparent conductivity maps for two sensors The apparent conductivity values corresponding to the profile shown in Figure 12 4 are shown in Figure 12 5 Caii an I hH S on i G ci rei i Cw ee eS aro areg meh liistance X im Figure 12 4 Apparent conductivity map for sensor HCP 2 m The north south profile has 1279 sites For a first interpretation there is no need of such density of sites Figure 12 6 shows the decimated n 10 curves together with the 78 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion response of the model shown in Figure 12 7 This model was calculated using the linear approach algorithm 20 iterations and lambda of 0 3 Action Legend Plot Raw Data z 2 m Plot Filtered Data 4m A 6m Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI Quit 159 9 142 8 125 7 108 6 91 5 74 4 57 93 40 2 23 1 6 0 11
50. it Gridding Method Grid geometry Inverse Distance Mean distance Input Data min 0 0000 Raw data C Rotated xmax 34 0000 OK xlines 3 Ymin 0 0000 Ymax 17 0000 rlines 5 omoothing lines y lines Weight Figure 7 2 1 The Data Processing menu The inverse distance option used the GETMAT subroutine in DISLIN The value at the grid point j k is calculated by 49 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Where j k are the indices from 1 to Nx lines and 1 to Ny lines respectively d is the distance of the grid point i k from the point P w is a weighting number default is 2 0 and N is the number of data points lying in the area controlled by the parameters in the smoothing section around the grid point j k The grid file can be filtered The available filters are linear low pass filters and should be applied to reduce the high frequency noise They calculate weighted averages of the neighbouring input grid nodes In the Moving Average SxT filter the weight are equal to one In the Inverse Distance SxT filter the weights fall off with increased distance The role of the distance is controlled by the Power The higher the power the more rapidly the weights fall off with distance The filter can be applied to all data set or only to specific channels Tox Exit Filter options I Moving average Distance Weighting Filter Height S and idth T i Power
51. ity i e o and resistivity p The default option is that the results will be presented in o Secondly these values of o or p can be presented using either a linear or logarithmic contour The default option is that the results will be presented using a linear scale Finally the user can also choose to display the inversion model alone and also the original o data The default option is that only the modelled results will be presented eae Save Print Conductivity Resistivity Log Lin ar contour Data Model section Grid lines Color scale Window size Figure 5 7 1 Choosing display settings The option Window size allows the user to adjust the size of the EM4Soil application The defaults size is the normal size 100 The program should be restarted to validate the option REE Size 5 Cancel Figure 5 7 2 Window sizes 40 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E Color Palette E E Ed Exit Color Palette TEMPERATURE GREY Minimum Maximum 90369 Data values Minimum Maximum 90369 E Use data limits OK Figure 5 7 3 The color palette menu 5 8 Save and Print The Save menu allows the user to save files and to print figures of Inversion and Data Processing Figure 5 8 1 shows that the Save menu options include Save Initial Models of either 1D or Q2D modelling Processed Data and results from inversion calculations Print Help Abou
52. ity at X Y Z or S Z in mS m and ohm m respectively 10 5 Inverse Models Q2D The file contains all the models imported in the Map module The format is as follow Xcoord Ycoord Altitude Distance Elevation Cond mS m Xi Yi Z11 Si Z1 Ci X2 Y2 ZT2 S2 Z2 C2 Xn_ gt Yn ZINL SNL ZNL CNL Where X Yi represent the easting and northing coordinates ZT the altitude topography and Z the elevation or the depth in a flat earth S is the distance between 74 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion measuring sites along the line having the first measurement as the origin C is the conductivity and resistivity at X Y Z or S Z in mS m 10 6 DOI The file contains the calculated values for the DOI and the format 1s X y S Z dol Xi Y1 S Zi doy X Y2 S2 Zo doiz Xn Yn Sn Zy doin Symbols have the usual meaning 10 7 Example files The package contains examples of input files and others that can be used to clarify the formats and to help the user to explore the different tools of the program Some of the files are generated synthetically Example of data files corresponding to data acquired in a line EM31 txt data from EM31 EM38 txt data from EM38 EM34lineO txt data from EM34 D21 data txt data from DUALEM 2Is D421 data txt data from DUALEM 421s D421Line0 txt data from DUALEM 42 1s Line4Profiler txt ProfilerSurvey txt 1s
53. l v1 05 a program for 1D Laterally Constrained Inversion APPENDIX D THE SIMULATED ANNEALING METHOD The name SA comes from the analogy with the process of physical annealing in thermodynamics once the undetermined parameters of the geophysical model are analogue to the particles of the physical system and the objective function of the inverse problem is analogue to the energy of the physical system Similar to the annealing process which is controlled by an initial temperature and a cooling schedule the estimation of the solution in the inverse problem is also controlled by a positive parameter T which limits the perturbation of the parameters values to acceptable values The acceptability of a perturbation is based on the algorithm presented by Metropolis et al 1953 According to this algorithm perturbations of the parameters leading to a decrease in the objective function are systematically accepted When an increase in the objective function is verified the changes in the parameters are not systematically rejected Instead the acceptance of the new parameters depends on the value of the function w exp AE T 1 where AE represents the objective function variation that is compared with a randomly generated number y between 0 and 1 The changes in the model are accepted if w gt x and rejected otherwise For high values of T corresponding to high temperatures of a melted physical system almost all changes are accepted
54. llow you to perform either 1D or Quasi 2D inversions These options are indicated in a drop down box shown in Figure 5 6 1 ained Inversion Usemi zoom Display Settings Parameters Depth of investigation Input initial model Q2D Inversion 2D b 1D Parameters Input initial model 1D Inversion 1D Appraisal of Inversion d Display model profiles Batch mode b Figure 5 6 1 Drop down box menu for Inversion selection If the o data does not have any negative value an initial model is automatically constructed from your data and during the data input stage However a user can choose an initial model using the option Input initial model IE Initial Model Ioj x Exit C Manually 1 0 Automatic C Open File 1 D Figure 5 6 2 Menu for input of the initial model As shown in Figure 5 65 2 a user has three options to input an Initial model a Choosing Manually 1 D the user can input a 1D model and as shown in Figure 5 6 3 defines the number of layers the depth of the lowest interface of the layers in meters and provide an estimate of the true electrical conductivity o in mS m b A uniform constant o initial model can be constructed Automatically inputting the initial conductivity this option should be used if you want to study the resolution of your model through DOI or your inputted data has negative values c Finally you can read the initial model from an ASCII file previously
55. lp About Legend 1m Plot Filtered Data am Data mode HCP 4m Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI 93 4 46 298 3 397 7 4971 3596 99 9 799 50 894 8 994 2 i Distance m Cond mS m Figure 5 5 9 Schematic representation of DUALEM 421 sensor data collected using a mobile EM sensing platform E EM4Soil v0 EM Laterally Constrained Inversion l m Exit Survey Input Data Processing Inversion Save Print Help About Action Plot Raw Data Plot Filtered Data 7 Decimation mode HCP 6 AF P Plot Noise nalysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI f 98 5 197 295 6 3942 492 7 5914 Distance m m wy qt SEH od oS T5 r 2 2 2 r FF 1 1 1 1 Figure 5 5 10 Schematic representation of DUALEM 421 sensor data decimated using a value of 2 5 5 5 Data Statistics A basic statistic of the data is possible using this option Histogram for each channel will be displayed together with a table showing the mean value and the standard deviation Also the number of negative values in each channel is shown The program uses 13 bins to create histograms Figure 5 5 11 20 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Exit Survey open Input DataProcessing Inve
56. luation will be done through the DOT index Oldenburg and Li 1999 Before using the DOI option the user will need to perform two different inversions using two different uniform initial models for example of 30 and 300 mS m and save the inversion model results use Save Inverse Model in the Save menu When selecting the DOI option the program will ask the user to input the files with those models previously calculated The Fitting option allows the user to see the correlation between experimental and calculated response Figure 5 6 12 Plot Filtered Data Plot Decimated Data Plot Data Statistics Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model A Appraisal of the inversion Ddal Plot Data Response 40 4 80 9 1213 1617 2022 2426 283 0 32935 363 9 4043 Distance m Action Plot Raw Data line line Plot Filtered Data 2 m Re 0 959 Plot Decimated Data Plot Data Statistics Plot Noise nalysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI Quit 76 7 121 5 166 3 Data mS m Figure 5 6 12 Correlation between observed and calculated data 33 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion The Display Model Profiles option allows the plot of vertical profiles of electrical conductivity through the model at some specified sites The proceeding to select sites is t
57. m for 1D Laterally Constrained Inversion NOTE you can follow the inversion steps in the DOS windows opened when EM4Soil starts Figure 5 6 7 Figures 5 6 8 and 5 6 9 show the models obtained from EM34 data using the Quasi 2D and the 1D unconstrained algorithm respectively line em34survey e 19 0 e Nit 20 Calculated Model 40 4 121 3 161 7 2021 242 6 283 0 323 4 363 9 404 3 Distance m Figure 5 6 8 Q2D model obtained from EM34 data line em34survey e 22 0 Nit 20 Calculated Model 235 8 611 7 536 0 460 3 Q 384 6 2 302 0 T 185 1 233 3 U2 157 6 5 81 9 151 3 6 3 0 0 40 4 80 9 121 3 161 7 2021 242 6 283 0 323 4 363 9 404 3 Distance m 218 9 202 0 cd as a oO _ ina gt g ea 168 2 Figure 5 6 9 1D model obtained from EM34 data assuming a two layers initial model One or more parameters can be fixed during the 1D inversion To do that the user must change the F to C in the table used in the definition of the initial model Figure 5 6 10 E Initial 1D Model Exit Figure 5 6 10 Defining an initial model for the 1D inversion F means that the parameter is free to be modified C means that the corresponding parameters will not be modified during the inversion 32 yy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion The Appraisal option will allow you to evaluate the calculated model In the present version of the program this eva
58. mber electromagnetic induction instruments Vadose Zone J 6 158 167 DeGroot Hedlin C and Constable S C 1990 Occam s inversion to generate smooth two dimensional models from magnetotelluric data Geophysics 55 1613 1624 Dualem Inc 2008 DUALEM 421S user s manual Dualem Inc Milton ON Canada Goldstein N E Benson S M and Alumbaugh D 1990 Saline groundwater plume mapping with electromagnetics In Geotechnical and Environmental Geophysics Ed By Ward S H Investigations in Geophysics n 5 SEG Golub and Reinsch 1970 Singular value decomposition and least squares solution Num Math v 14 no 3 403 420 Gomez Trevino E F J Esparza and S M ndez Delgado 2002 New theoretical and practical aspects of electromagnetic soundings at low induction numbers Geophysics 67 1441 1451 Huang H 2005 Depth of investigation for small broadband electromagnetic sensors Geophysics 70 6 G135 G142 Johansen H K 1977 A man computer interpretation system for resistivity soundings over horizontally stratified earth Geophysical Prospecting 25 677 691 Kaufman A A and G V Keller 1983 Frequency and transient soundings Methods in Geochem and Geophys 16 Elsevier New York Keller G V and F C Frischknecht 1996 Electrical Methods in Geophysical Prospecting Pergamon Press Inc 513 p McNeill J D 1980 Electromagnetic terrain conductivity measurement at low induction numbers Geonics Limite
59. model calculated with the linear approximation Plot Raw Data Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model ___Plot Inverted Model Plot Data Response Plot DOI Quit Calculated Model _ _ E Q sl PE w ma wu caj 510 4 638 0 Distance m Figure 12 8 Conductivity model calculated using the full solution 80 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion Action Legend Plot Raw Data 2m l Plot Filtered Data 4m Data Response mode HCP 6m Plot Decimated Data Plot Noise Analysis pcre l Plot Neg Corrected itt sauenee BN Plot Data Shifted x Plot Initial Model Plot Inverted Model Sn 7 6 2 3828 5104 638 0 7656 8932 10208 11484 1276 0 Distance m Quit Data Response mode PRP Q h t g 4 eet a R A dn wee wre t s nappat BET epeenegne sae z w gel 7 gt a SORE Tan T a baetae 4 ptt BO Es agi gitt a 510 4 638 0 765 6 Distance m Figure 12 9 Data and model responses from the full solution model PROFILER SURVEY Figures 12 10 and 12 11 show the model and its apparent conductivity responses obtained by inverting the data collected along the north profile of the survey shown in Figure 7 1 2 Plot Raw Data Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot I
60. n adding profiles in a survey with closer sites To apply it select Zoom In in the menu bar and select the area of interest with the left mouse button pressed Disable the zoom selecting Zoom Out tofile mode Inversion Zoom Save Print Help About oom In oor 1k Figure 7 5 1 Printing survey maps Exit Display Profile mode Zoom Save Print Help About Quit 698 0 Distance X m Figure 7 5 2 Zooming in survey maps 55 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E EM4Soil Map module Exit Display Profile mode Zoom Save Print Help About Quit l0l m Y Distance 686 6 690 6 Distance X m Figure 7 5 3 The Zoom In result 56 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion 7 6 Save Mee Save Print Help About Save Profiles Save Corrected Data Save Conductivity Maps Save Inverse Models 2D Save Inverse Model Q3D Save Hor Slices Q30 Save Vert slices OSD Save Model Response OSD Save DOT 3D Save Map Results MOP Figure 7 6 1 Options in the Save menu After select all your profiles they must be saved in separated files Use the option Save in the menu bar to do that Select the profile you want to save and proceed choosing the folder and file name Figure 7 6 2 These files can after be input by EM4Soil program to be inverted NOTE The file of an added profile will con
61. nes The water content and bulk resistivity can be measured in laboratory Alternatively the resistivity calculated by inversion can be used The data can then be used to determine the parameters of the equations Figure 5 5 22 0 OF Microhigh Mel O Mbcrolow Ut A intermediate Fitted curve p4 0 2 EA Fon o 01 Irth i i 20 4 Gi a0 10i Resistivity ohm m Figure 5 5 22 Example of soil moisture versus resistivity data and fitting curve a 6 0145 b 0 12 and m 1 536 from Amidu and Dunbar 2007 In addition water content variation can be estimated from two EM surveys made at different times This assumes that Archie s law is valid and that the variation in water content is responsible for the variation in 6 and between the two survey times The user will be asked to provide the n parameter and the files corresponding to the two models The user must remember that the use of the Archie s law requires a clay free medium Plot Initial Model Plot Inverted Model Water Content Variation Plot Data Response Plot DOI Quit QO onl w a 4 a 7 5 10 0 12 5 50 175 Distance m Figure 5 5 23 Example of a pseudo section of water content variation 28 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 6 Inversion After carrying out various Data Processing options the o data is ready for inversion Selecting the Inversion option the program will a
62. ng the corrections applicable to the parameters logarithm of block conductivities p of an initial model po is a reference model b is the vector of the differences between the logarithm of the observed and calculated oa b In o In o J is the Jacobian matrix whose elements are given by 0 Oai 06a Oo the superscript T denotes the transpose operation and is a Lagrange multiplier that controls the amplitude of the parameter corrections and whose best value is determined empirically The value can be determined empirically by comparing the models calculated using different values with the available information The elements of the matrix C are the coefficients of the values of the roughness in each parameter which is defined in terms of the four neighbors parameters The elements of C are 4 1 or 0 An iterative process allows the final model to be obtained with its response fitting the data set in a least square sense The misfit between data and model response is measured through the rms defined by rms Ee o 3 Cumulative Response of a Multilayered Earth At low induction numbers the magnetic coupling between ground current loops induced by the primary field is negligible and for this reason the secondary magnetic field measured at the receiver is the sum of the independent magnetic fields from each individual induced current loop McNeill 1980 In this case and properly normalizing the measured fields the d
63. nitial Model Plot inverted Model Plot Data Response Plot DOI Quit Calculated Model 260 0 Fa 207 1 o 154 3 101 4 234 29 2 350 409 46 7 Distance m Figure 12 10 Conductivity model obtained by inversion of a PROFILER data line 81 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion Exit Survey Input Data Processing Inversion Save Print Help About Action Legend Plot Raw Data 1000 Plot Filtered Data 9000 Data Response mode VDM 15000 MN Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted 171 2 Plot Initial Model 152 9 Plot Inverted Model su ot Inverted Mode oO 116 24 Plot Data Response 97 9 Saat 5 175 234 292 9350 409 Distance m Quit Nw jee Ne D Cond mS m Figure 12 11 PROFILER data and model responses corresponding to the model shown in Figure 56 82 yo EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion References and useful bibliography Sikiru A Amidu S A and Dunbar J A 2007 Geoelectric Studies of Seasonal Wetting and Drying of a Texas Vertisol Vadose Zone Journal v 6 n 3 511 523 Anderson W L 1979 Numerical integration of related Hankel transforms of order 0 and 1 by adaptative digital filtering Geophysics 44 1287 1305 Callegary J B T P A Ferr and R W Groom 2007 Vertical spatial sensitivity and exploration depth of low induction nu
64. ofiles Filter Corr Delete Negative values Figure 7 8 1 The Help menu 59 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 8 EM Vertical Sounding Vertical EM soundings stands for measurements made at a range of heights over a given location see e g http www dualem com vertsnd htm The EMSound module was designed to invert such data However you can also invert the data collected at a particular site of your survey line Let s see this option first Start displaying the raw data After that go to Select Site EMSounding option in the Input entrance in the menu bar This allows you to choose the sites which data you want to invert The selection of the sites is made clicking on the left button of the mouse similarly to the process explained in 5 4 5 1 Correction of individual values Figures 8 1 and 8 2 shows some spots of the selection process O gt Action sensor Plot Raw Data line lined 1m Plot Filtered Data 2m Data mode HCP 4m Plot Decimated Data 170 4 159 5 Plot Noise Analysis 148 6 Plot Neg Corrected he 137 8 126 9 Plot B L Corrected 4 116 1 105 2 Plot Initial Model J o4 5 Plot Inverted Model 83 5 Data selection for EM SOUR Plot Data Response Select a site anD is 161 7 2022 2426 283 0 3235 3639 4043 Distance m t Data mode PRP 172 1 156 0 2139 9 123 7 2 107 6 zj 91 5 753 5 43 0 26 9 Figure 8 1 Selecting
65. on 5 5 2 Filtering Prior to inversion EMA4Soil allows a user to Filter the o data As shown in Figure 5 5 3 the various options include Running Average Weighted Average and Sheppard filtering NOTE Selections not desirable must be disabled These three filters can be applied to the entire o data set collected 1 e All or to selected o data In Figure 5 5 3 the options available are for the EM34 o data whereby o data collected in either the vertical dipole G e VDM or horizontal dipole i e HDM modes can be individually selected for filtering Fim Oataritering auPTTES Exit Filter options Iw Running average Weighted average E Sheppard Data to be filtered E All Iw YOM a HOM I 10 m E 20 m E 40 m Figure 5 5 3 Pop up box showing Filtering options and Data to be filtered In order to describe the various filters i e Running Average Weighted Average and Sheppard filtering and the effect on o data Figure 5 5 4 shows an example set of DUALEM 421 o data prior to filtering EM4Soil 0 EM Laterally Constrained Inversion i i 10 x Exit Survey Input Data Processing Inversion Save Print Help About ece Plot Filtered Data 7 Data mode HCP Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot DatafResponse Plot DOI 993 4 497 7 497 1 596 5 695 9 799 9 894 8 334 2 Distance m Figure 5 5 4
66. on Quasi 2D model algorithm or a 1D inversion without constraints The user chooses to use the linear or the non linear approach for the model response and derivatives calculations In any case you must define the o data to invert Figure 5 6 6 MiInversion afl ess Exit Choose data for inversion RAY DATA FILTERED DATA DECIMATED DATA NEG CORRECTED B L CORRECTED RESAMPLED DATA Figure 5 6 6 Choosing data for inversion NOTE Quasi 2D inversion of data collected with EM instruments having only one sensor like EM31 EM38 DUALEM 1 or DUALEM 2 only should be done if the data set is complete and even in such case the models should be analysed with care In such cases it is preferable to perform a 1D inversion Action Plot Raw Data Plot Filtered Data Plot Decimated Data 9RGta pet ge sete E the Linear Approach Maximum iterations if Plot Noise Analysis Humber of parameters 728 Sensor BUALEMN 421 _ Plot Neg Corrected Aa ot 1988008 76 68339 nit Phib Phi 76 60339 24 10653 Plot Data Shifted nit Phib Phi 24 10653 2158361 nit Phib Phi 21583681 18 92857 Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI Quit Figure 5 6 7 Following the inversion process NOTE During the inversion which can take a significant time dependent on the number of sites you have the program will not process any instruction 31 vy EMA4Soil v1 05 a progra
67. or data is used in inversion modelling E EM4Soil v0 EM Laterally Constrained Inversion E l loj x Exit Survey Input Data Processing Inversion Save Print Help About Legend 10 m Plot Filtered Data 20 m Data mode YDM Plot Decimated Data cau Plot Noise Analysis Plot Neg Corrected Plot Data Shifted Plot Initial Model Plot Inverted Model Plot Data Response Plot DOI 20 AQ jl BO0 0 100 0 120 0 1400 160 0 180 0 Distance m ait Data mode HDM Cond mS m Cond m8 m 20 0 400 600 800 1000 1200 1400 160 0 180 0 Distance m Figure 5 5 8 Screen snapshot of EM34 sensor data that will be ignored if a Decimation factor of 2 is selected Owing to the large amounts of EM o data that can be collected using mobile EM sensing platforms the decimation option also allows a user to experiment with a smaller set of EM o in order to determine a suitable value of and the number of iterations to carry out without having to wait for long periods In this regard and once an optimal set has been identified the entire set of EM sensor data can then be inverted for optimal results Figure 5 5 9 shows an original set of DUALEM 421 sensor data Figure 5 5 10 shows the effect of selecting a decimation factor of 2 on this data 19 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion EM4Soil 0 EM Laterally Constrained Inversion Exit Survey Input Data Processing Inversion Save Print He
68. pen Input DataProcessing Inversion Display Settings Save Print Help About Action lins no name Plot Raw Data Raw Data Apparent Conductivity Plot Filtered Data Plot Resampled Data Plot Decimated Data Plot Data Statistics ate Plot Noise Analysis 2 4 6 Distance m Plot Neg Corrected In Phase and Quadrature Plot B L Corrected epa R000 1606 Plot Initial Model 1000 Plot Inverted Model Plot Data Response 1000 1600 2000 Plot DOI 2 4 6 Distance m Quit Apparent Susceptibility a an 5 g a a f T 5 4 6 Distance rm Figure 2 2 Data from a line 96 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion Exit Display Data Processing Tools Profile mode Inversion Zoom Save Print Help About Survey survey_70 Plot Raw Data Sensor Sites Distance X m Quit aaa t 3 2 9 9 316 Qe 4618504 00 461850100 4616498 00 4616404 50 tl q be Z bo m 1 eowe4zsig we payit merit eae ST ee on ae i eee a g Call paa Pt un pe ea J oe s gaged hp vi i rik af APA ee l 4 ra ad gta 461849156 i i Porai men peat i 4618488 00 wren a a 7 a k eee T ii ample wi 8 ae PTE a4 ir i 4616485 00 8299 7 9B8 583009 06 583020 75 583032 44 583044 13 583055 81 x Easting Figure 2 3 Data from a survey area 3 Data Processing 3 1 Filtering The sensors and components to be filtered can be selected in a table F
69. r 5 1 Exit Selecting the Exit button will cause EM4Soil to shut down A user s data processing and inversions results will not be saved unless the user has previously saved them 5 2 Survey When the Survey option is selected a small drop down box will appear with two options Figure 5 2 1 W EM4S5oil 103 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Save Print Help About Figure 5 2 1 Screen shot of the Survey button options 5 2 1 Survey New Selecting the New option enables meta data to be entered and pertaining to the o survey data This meta data will be displayed in the resulting output files NOTE Entering meta data is optional As shown in Figure 5 2 1 the meta data that can be entered includes information pertaining to the survey title 1 e transect number company the survey was done for or undertaken by area country date of survey or processing sensor type and operator can be entered Fm New Survey TIE EE ar Survey M Company i a Area St Country a Date Sensor fT apene O OK Figure 5 2 1 Screen shot of the input meta data table to enter information about the survey This information will be saved upon clicking the OK button vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 2 2 Survey Open Selecting the Open option allows EM4Soil to read from an ASCII file previously
70. r 1D Laterally Constrained Inversion DISCLAIMER The programs software included in any version of the EM4Soil packed are provided as are without any express or implicit warranties including their suitability for a particular purpose The authors and EMTOMO LDA will not assume any responsibility for any indirect or consequential damages or any loss caused by the use of these programs Efforts will be made to correct any program bug that appears during the usage of the package vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 1 Introduction EMASoil is a software package that has been developed in order to enable the inversion of electromagnetic EM induction data acquired at low induction numbers LIN A review of the LIN theory as well practical applications of the theory with regard to geology and hydrogeology can be found in Geonics Technical Note 6 McNeill 1980 The input data for EM4Soil is the measured apparent soil electrical conductivity Ga as measured by EM instruments As an example of the type of data that can be inverted by EMA4Soil is shown in Figure 1 1 It shows DUALEM 421 og data that was collected using the horizontal co planar HCP array at 1 red 2 green and 4 m blue coil spacing The lower panel shows the equivalent og data that collected using the perpendicular PRP array and at 1 red 2 green and 4 m blue coil spacing Me FMi FM ed erally Constrained lave rsions
71. re are two predefined models one for the frequency range 19950 to 1710 Hz and a second model for a more wide range of frequencies The users are advised to build their own models taking into account their experience and data characteristics For further information in this regard refer to Section 5 5 Input Data x Figure 5 4 5 When the elevation values are too noisy the user can smooth it The elevation seems to have significant variations Do you want bo smooth it n_ Figure 5 4 5 shows a pop up box which may appear in the event elevation data which is scanned by EM4soil shows a large scattering This scattering might distort the output inversion models In this case it is advisable to consider filtering the elevation data 13 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Selecting this option will allow the user to conduct this type of filtering see Section 5 5 5 5 4 4 EM Sounding You can choose a set of EM sensor sites made at particular sites to perform individual 1 D inversions Figure 5 4 6 This will be explained in a next paragraph EM Laterally Constrained Inversio m Input Data Processing Inversion Input Sensors Input Data Select sites EMSounding a0 to Map Figure 5 4 6 Choosing a site a set of measurements for 1D inversion 14 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 5 Data Processing Once the o data has been succe
72. red in VDM mode or 2 if acquired in HDM mode Nf is the number of frequencies and fl f Nf are the frequencies of the survey in Hz condifl are the measured apparent conductivity at high i for each frequency mS m 68 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion The format for OTHER sensors This format is to be used when the sensor is OTHER The format allows the user to input in EM4Soil program data acquired from different instruments In this format the data of each sensor 1s defined individually The format 1s Nameofthesurvey character Number of sensors integer Code of the sensor character Number of readings height integer real Coils spacing frequency mode acquisition type of data units real real integer integer integer X Y ZCa Lat Long time labels corresponding to the data in the following rows The following rows contains the measured data and the format depends on the type of data For data type 1 Xeasting Ynorthing Z Ca lat long time all real For data type 2 Xeasting Ynorthing Z Ca In phase lat long time all real For data type 3 Xeasting Ynorthing Z In phase Quad lat long time all real For data type 4 Xeasting Ynorthing Z Ca In phase Quad lat long time all real For data type 3 Xeasting Ynorthing Z Ca In phase Quad Susc lat long time all real The information in blue must be repeated for each sensor The
73. rger survey and carried out across an area can also be imported This last option will be explained later in this Manual see Section 7 EM Laterally Constrained Inversion m Input Data Processing Inversion Display Settings Save Input Sensors wl area Select sites EM Sounding EM Sounding fel 0 to Map Figure 5 4 3 Screenshot of the drop down box which allows the appropriate data to be selected In order to prepare the EM o data refer to Section 9 Input File Format If a user s data has successfully inputted o data into EM4Soil the following pop up box will appear on the screen Figure 5 4 4 Input Data x Data Read DUALEM s4 1 Figure 5 4 4 Pop up box which indicates that the data collected along the transect to be inverted was successfully read into EM4Soil type of number of measurements 400 sensor the o data was collected with number of o measurement sites that were collected and that the construction of an initial model for inversion initial model saved has been saved An initial multilayer model based on the EM sensor data selected is generated from the c values inputted In this regard a pre determined number of layers and depths will be constructed by EM4Soil The number and depths of layers is based on the type of EM sensor selected and the theoretical depth of investigation see Section 9 For PROFILER the predefined model is based on the frequency range of 1000 to 15000 Hz For GEM the
74. rmat Section to read about formats EM Laterally Constrained Inversion n Input Data Processing Inversion Display Settings Input Sensors Area EM Sounding Select sites EMSounding 50 fo Map Figure 7 1 Inputting data from a survey Inputting data with this option you will run the EM4Soil Map module program which allows you to see your data as a set and to select profiles for inversion Figure 7 2 shows the screen of the program The main EM4Soil program will be inactive during the use of the Map module Clicking in Quit you return to EM4Soil NOTE For the same data set you can go to Map module program using the Go to Map option E EM4Soil Map module Exit Display Data Processing Profile mode Inversion Zoom Save Print Help About Oj x Action Plot Survey Plot Raw Data Quit Figure 7 2 The EM4Soil Map module screen Figure 7 3 shows the default menu bar of the program vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion W EM45oil Map module Exit Display Data Processing Profile mode Inversion zoom Save Print Help About Figure 7 3 The menu bar of EM4Soil Map module NOTE Some options related to the Q3D inversion in the MAP module are explained in the Q3D Manual 7 1 Display Click in the Display entrance to see the location of your sites Survey Layout and maps of your data Data Conductivity Resistivity and elevation In Settings you can make
75. rsion Display Settings Save Print Help About cee Plot Raw Data Line lined Exit Plot Filtered Data Sensor Mean Std HCP Tm 10123 10 85 Eee CPAP Im aa 7 58 Plot Data Statistics 9 d 3 HCP 2m 135 87 12 69 ts 4 by N n PRP 2m 89 29 10 53 Plot Noise Analysis i 9 J R ou D Plot Neg Corrected m Plot B L Corrected r NIN N w Np JIN z A 72 6 86 4 100 2 114 0 94 6 113 3 132 0 150 7 103 3 120 6 137 9 155 2 Plot Inverted Model oe z f X ieee K nie z E pe ae Conductivity mS m Conductivity m35 m Conductivity m5 m Plot Data Response Plot DOI PRP 1m PRP 2m PRP 4m Quit u A tab a z T 6 o i H H Fy Foy 35 9 450 540 63 3 764 895 102 5 89 3 109 8 130 3 150 9 Conductivity m5 m Conductivity m m Conductivity m5 m Figure 5 5 11 Presenting results from data statistic e e 5 5 6 Noise Analysis The Noise Analysis option is an experimental tool based on Everett and Weiss 2002 When using this option you can generate information about the noise of your o data see Appendix B NOTE This tool must be used with some caution given its experimental nature 5 5 7 Pseudo section ECa Pz This option allows the display of a pseudo section calculated from measured apparent conductivity This is only a different way to present the raw data The pseudo section is calculated assuming a pseudo depth which is roughly the depth of investigation theoretically associated at each sensor Th
76. s for sensors models and the approach to use in the inversion should be written in capitals The software is key sensitive to the codes NOTE For PROFILER and GEM the approach must be FS NOTE Do not put in the same batch data files acquired with instruments from different constructers Do not mix DUALEM data with GEONICS ones NOTE The check made by the program is very basic so the user should verify all the files before start a batching process 36 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion A table with a resume of the inversion results will be presented Figure 5 6 16 after running the batch The results of each inversion are saved in the respective INVr files which were saved in the folder inverse and can be recovered using the Open Inversion results entrance in the main menu bar ixl Exit Data File Misfit 26 5 09386 Nit a A 5 90758 se i 5 79033 Figure 5 6 16 After running the batch The batch file should be saved with the extension BATm to avoid misunderstandings with other type files These files can be read with the read batch option and run See the format section about the batch file format NOTE sometimes the right mask for files does not appear in the windows The user should look for that The Depth of Investigation entrance will open a new module that will allow the user to study the depth of investigation of his data According to Huang 2005 the
77. s not stand in the center of the instrument at where the measured values correspond For this reason a correction of the sites coordinates is necessary The correction can be done knowing the GPS position relatively to the center of the instrument as well as the instrument position relatively to the line survey Figure C1 DUALEM instruments included optionally a built in GPS which location in a local referential is shown in Table 1 Table 1 Built in GPS location in the local reference system s 0 These values are averaged from the distances corresponding to HCD and PRP sensors DUALEM IS _ 0 325 DUALEM 28 _ 0 175 northing northing easting easting Figure C1 Local reference system X Y to localize the external GPS assuming that the survey is carried out in the Y direction The figures show the instrument in the in line position left and in the board side position right O represents the center of the instrument and Tx the transmitter position DUALEM instruments When using multisensors like DUALEM 421S the values measured by the different sensors are not referenced to the same coordinate The program allows you to refer all measurements to the same coordinate the center of the instrument using two different algorithms by inverse distance to a power interpolation or by nearest neighbor The second algorithm is probably more adequate for continuous measurements with short distance between sites 92 vy EMA4Soi
78. saved In order to prepare the o data please refer to Section 9 Input File Format 5 3 Open This entrance allows the user to open input results obtained in previous inversions as well as results obtained in EM4Soil and Map module session Me EM4S5oil 105 EM Laterally Constrained Inve Exit Survey Open Input Data Processing Inver Inversion results INVr Inversion resulks INY3 Action Plot EM45oil results PRI Plot F Map module File MAP Plot Re Merge Data Files Convert Data Format Pint 1 Figure 5 3 1 Inputting previous inversion results generated by EM4soil 5 3 1 Input Inversion Results INVr It is also possible to import and see the results of previous Q2D inversions Figure 5 3 2 Files with Q2D inversion results have the INVr extension INV3 is reserved for results of Q3D inversions and are named according to the name of the line survey Zox Results of your inversion Please select the one you want to display rms iteration O 31 27 rms iteration 1 2 85 rms iteration 2 2 51 NONE Figure 5 3 2 Inputting previous inversion results generated by EM4soil 5 3 2 Input EM4Soil results files PRJ Input files with results obtained during a session in the EM4Soil and saved with extension PRJ see item 5 8 5 3 3 Input Map module files MAP Input the files saved in the Map module and containing the processing results MAP extension 10 vy EMA4Soil v1
79. saved see the Format Section 29 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Figure 5 6 3 The Table for input of a 1D model i Initial Model joj xj Exit conductivity mS m K Figure 5 6 4 Inputting the conductivity of the uniform initial model option Automatically Selecting the Inversion Parameters option will display the following menu Figure 5 6 5 which allows the user to define some of the inversion parameters E Inversion Parameters Ioj x Exit Damping factor Number of iterations Data error C Algorithm 1 Algorithm 2 OK Figure 5 6 5 Input of inversion parameters The damping factor controls the roughness of the model see the appendices for details The larger the damping the smoother the resulting inversion model A suitable damping factor value should be determined empirically and by performing inversions with different values e g from 0 03 to 30 NOTE When written values follow the format displayed in the menu The program allows the use of two different inversion algorithms named here as S1 and S2 They are based on the work published by Sasaki in 1989 and 2001 see appendices 30 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion for details In general S2 algorithm produces smoother models when compared with those from S1 EMASoil allows the user to invert o data using a 1D Laterally Constrained Inversi
80. sites for 1 D inversion lox You have selected 2 sites to be inverted Please select the site you want to invert site 60 5 site 152 3 Figure 8 2 Selecting a site from the chosen sites 60 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion After select a site the EMSound module will be activated Figure 8 3 The screen of this module is divided in two parts the left part is for data display and the right one for models The menu bar Figure 8 4 allows to Display data models and inversion results Figure 8 5 E EM4Soil Sounding module i O gt Exit Display Input Inversion Save Print About Sounding site 60 5 BUAL 4218 EM Sounding Data H oo oo ff Oo Cond m m 2o re Q 0510 15 20 4 5 3 0 35 4 0 4 5 Coila Spacing m Figure 8 3 The EMSound module screen W EM4S5oil Sounding module Exit Display Input Inversion Save Print About Figure 8 4 The menu bar of the sounding module Me EM45oil Sounding module Exit Display Input Inversion Save Print About Data Models Inversion results Correlation table Models table Figure 8 5 Display options The entrance Input is used to define the layered initial model This can be done by reading an external file see the data format section or filling a table Manually W EM45oil Sounding module Exit Display Input Inversion Save Print About gt
81. ssfully entered into EM4Soil the software allows the user to carry out some basic data processing The raw data are preserved during processing Therefore any new processing is done over the raw data NOTE The users must save any processed data if they want to keep it WE EM4Soil v105 EM Laterally Constrained Im Exit Survey Open Input GRE eee Inversion Action eat Editi WotePad Plot Raw Data Filter Resampling d Plot Filtered Dat Decimation Qaka Statistics Plot Resampled D Noise Analysis Section ECa Pz Correction of Lat Lon to UTM Plot Decimated Dz Plot Data Statist Medium Properties Plot Noise Analys Figure 5 5 1 The Data Processing menu and the Correction option 5 5 1 Editing Selecting Edit will open a table with the selected raw EM sensor data Figure 5 5 2 If changes are made the user will be prompted to save the data in a new file This will need to be selected in order for this edited data to be inverted EE Zox Exit sensor height Z VDM ADM oo ooo 231 50 70 00 135 00 C oen aooo f 23160 105 00 140 00 2060 30 00 231 70 7000 Taso C sem 30 00 231 80 5500 140 00 x amw 8050 sooo 23 40 50 00 135 00 Figure 5 5 2 Pop up box indicating the layout available to Edit raw data The Edit NotePad will start the NotePad program if available 15 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversi
82. tain the Lat Lon and Time of the raw data corresponding to sites included in the profile CESE l0 x Quit 46 1 CEE ixi Exit Added profiles no name 2 0 72 fafa Distance X m Figure 7 6 2 Saving the data of each added profile NOTE The option Save Inverse Models Q2D will save all imported models in one ASCII file This file can be used in other graphical programs See Section about the format of output file 57 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion The options referring the Q3D outputs will be explained in the respective Manual The option Save Map Results is of particular interest because it allows the user to save all the processing done in the Map module in one file MAP This file can be read by EMA4Soil program in the Open menu bar entrance 7 7 Print The entrance Print in the menu bar will allow you to print figures of the survey layout and of the apparent conductivity and elevation maps in the selected metafile format Print Help About Metafile Format OSD Results d CondiResistivity Topography Model Slices Filtered data Figure 7 7 1 Printing survey maps 58 7 8 Help EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion An overview of the different tools is given in the Help menu Print Help About Short Guide Display survey Display data Model Slices Data Processing F Gridding Pr
83. teiro Santos F A 2009 2 dimensional soil and vadose zone representation using an EM38 and EM34 and a laterally constrained inversion model Australian Journal of Soil Research 47 809 820 Wait J R 1962 A note on the electromagnetic response of a stratified earth Geophysics 27 382 385 84 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion APPENDIX A Inversion of EM data Inversion of geophysical data is a mathematical procedure that seeks to obtain the distribution of one or more physical property in the survey area or volume In the EM case electrical conductivity or resistivity is the property of interest Therefore from a finite number of data apparent conductivity values one want to know the true conductivity distribution that justify such data set Having a discrete sampling of the apparent conductivity it is only possible to calculate discrete conductivity distributions that is conductivity values representative of some limited zones of the survey earth s zone It means that an earth model should be adopted There are three candidates a layered earth 1D an earth model allowing the variation of conductivity in two directions 2D or a more realistic model allowing that the conductivity vary in the three directions 3D The parameterization of each model is obviously different and the number of unknown parameters increases from 1D to 3D models The model to be adopted depends on several fac
84. threshold and the factor for depth increasing The depth of investigation is calculated using the ratio A Sa So where Sa is the apparent conductivity of the two layer model and So the apparent conductivity of a uniform model host The depth of investigation is the depth of the second layer top at which A gt 1 T 100 or A lt 1 T 100 where T is the detecting threshold After model definition and parameters selection the calculation will be performed clicking in the Calculation entrance The output will be displayed as shown in Figure 5 6 21 38 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E EM4Soil Depth of investigation module Ioj x Exit Display Input Calculation Print About Fregi Hz 15000 1000 Instrument EMP 400 Depth of investigation T 16 4 T O Depth fer 15000 Hs isi 10 m 2 Depth for 1000 Hz ix 26 m OG 0 4 C1 O 0 1 4 oh 4 4 6 4 Thickness Lat layer m0 Figure 5 6 21 Showing the depth of investigation for selected testing model and frequencies Note All calculations are considering the VDM mode which is probably the most used mode in acquisition 39 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 7 Display Settings Figure 5 7 1 shows the Display Settings menu options here allow the user to choose how the inversion models are presented In the first instance the user can choose between conductiv
85. tic H i i ERE p Plot Noise Analysis water Content Variation Figure 5 5 20 Drop down box menu for Medium Properties option As an example and by selection water content this property will be estimated As shown in Figure 5 5 21 there are two approaches both of which are based on experimental data for fractional m m water content estimation E EM4Soil 103 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Save Print Help About Edit Editi NotePad Plot Raw Data Filter Resampling Plot Filtered Data Decimation Data Statistics Noise Analysis Correction of r Lat Lon to UTM line emd4 Linel e 1 Medium Properties Water Content Experimental Data Water Content Variation algorithm 2 Action Plot Resampled Da Plot Decimated Da Plot Data Statistic Plot Noise Analysis Figure 5 5 21 Drop down box menu for water content option The first algorithm is based on the equation Epe The second one is based on 0 ap b 21 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Where is the fractional water content p is the bulk resistivity 9 1s an experimental parameter correlated to the water resistivity a b and m are experimental parameters These parameters can be determined from samples collected at different depths in boreholes located at selected sites preferentially coincident with EM li
86. tivily mF q ce C pe e te am 20 F Iba 2 0 lhistance X ra Freqtency 1000 0 Hz Figure 7 1 3 Displaying a map of apparent conductivity PROFILER data at a frequency of 1000 Hz 46 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion E EM4Soil Map module q Exit Data Grid geometry Raw Data min 765139 812 Si max 767851 312 Plot sites C NO sites Ymin 6758557 000 Display Ymax 6760393 000 Conductivity C Resistivity Contour type Linear C Logarithmic Rescale axis NO C YES OK Figure 7 1 4 Settings menu You can also import the models calculated from each profile and display horizontal slices at select elevations or depths Figure 7 1 5 to 7 1 7 ainai 2 s 4 Horizontal resistivity slices a D o 2 co 202 0 E x OTL 3 FO 5 am wp a b a a if ml TU UE i 5 lt n i 670 2 679 5 688 7 98 9 TIe 716 5 Distance X m Depth 5678 Figure 7 1 5 Displaying a model slice from the inverted models elev 157 8 m 47 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion REE Exit Data Grid geometry Raw Data min C Plot sites NO sites mtn Conductivity Slices Resistivity Elevim Contour type C Depth m ig aac Rescale axis f NO C WES OK 24 6 3 2 E 22 0 Minimum elev 21 1 Kdaximum elev 26 9 Figure 7 1 6
87. to its conductivity and depth as o o R z Yo R 5 Rs 94 1 R zy1 The derivatives of the apparent conductivity oa with respect to the layer conductivities are Co a R 0O 2 2o lt 7 R z R z4 i 2 M 1 8 Co S Oo 56 UR Gu 7 Similar equations can be written for GEONICS instruments EM31 EM38 and EM34 McNeill 1980 Ry lt lt 10i 88 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Rap V42 1 2z 11i and O O I R z So Rza E R z Oyu Rma 121 i 2 for the apparent conductivity calculations Usually the initial thickness of the layers is kept constant in the inversion process The use of the cumulative response to calculate the model response at each measuring site means that we are not considering the EM interaction between constituent blocks of the model however the smooth inversion algorithm constrains each block o to be somewhat dependent on its neighbors That is the method represents a one dimensional laterally constrained approach and the final model is a rough representation of a two dimensional model Quasi 2D Full Solution of the Maxwell Equations Forward calculations based on the work of Keller and Frischknecht 1966 Wait 1982 and Anderson 1979 are used to calculate sensor responses at each measuring site assuming a 1 D model For a vertical dipole as primary source the secondary components of the magnet
88. tors but the most important are the geology and the geophysical array used in the data acquisition In any case the information about the conductivity distribution will be limited by the number of data and model correctness One cannot expect to have more information than that contained in the data set Because this information is partial several models can fit the data The decreasing of this ambiguity can only be done using additional information from boreholes or other geophysical indirect methods Let s look closely to an example Consider the synthetic data shown in Figure Al acquired with EM38 The data consists of VDM and HDM apparent conductivity values measured over a two layer earth 10 and 85 mS m and 1 m thick With these two data values at each site what kind of information do we expect to obtain In the best situation a rough two layer model can be reached The parameters will be the thickness of the top layer and the conductivity of both layers The ambiguity of this model will be also a function of the geoelectrical profile The model resolution of a low high conductivity layer sequence will be different of a high low conductivity sequence www Plot Filtered Data Data mode VDM Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Response 2 Plot DOI A 9 Oo 18 0 22 9 ai Distance m t Data mode HDM Cond m8S m e roa
89. ull values will appear with a red background You can modify these o values Selecting OK allows any and all subsequent inversions that will be carried out by EMA4soil to use these corrected o values as raw data However the original data file will not be modified iol x Exit 102 30 231 40 120 00 125 00 60 00 5500 115 00 2715 50 231 40 90 00 130 00 150 00 18 00 115 00 Figure 5 5 13 Drop down box menu for correction option Sometimes negative values are present in the oy data Such values can not be interpreted by EM4Soil and will therefore not be considered during inversion A user may try to correct these values by interpolation This can be done within EM 4soil and using the Correction of negative values option Negative values can have different origin electromagnetic noise weak induction or high conductivity bodies e g metallic objects Correction of these values can only be useful if the negative values are due to noise 22 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Figure 5 5 14 shows an example of a transect of DUALEM 421 o which has negative values in the 4mPcon E MH Soil v0 M Laterally Cor Artinn i __PlotRawData lot Filtered Data Plul Decimate d Dala Plot Noise Analysis Plat Neg Corrected Plot Data Shifted l lot Initial Mode Pluthiverlted Mudel Plot Data Response Pint DOI Quit a S m pa pm j d p DS S mae
90. ut Quit 2 0 TA Distance X m Figure 7 3 3 Screen showing the sites included in the profile which sense is indicated by the arrow If this procedure is followed with the second option all sites collected in sequence between the selected sites will be incorporated in the line This option must be used 1f the user is confident about the sequence of the measurements otherwise the program can select wrong data This option is useful if the survey was acquired in parallel profiles NOTE The most problematic situation occurs when the acquisition paths cross each other In such case the added profile can show a significant scatter and the data should be filtered probably manually A similar situation can appear if the survey lines are very close distance between lines of the same order of the distance between sites In those cases the sampled profile must be carefully analysed before inversion NOTE The user can sample two or more short lines along the same direction that will be merged posteriorly into a file NOTE There is a limit for the number of profiles saved in a session 35 profiles 53 yo EM45Soil v1 05 a program for 1D Laterally Constrained Inversion 7 4 Inversion Q3D Please see the Manual for Q3D Module 54 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 7 5 Zoom The user can zoom in the display of the survey and maps This operation is particularly useful whe
91. utmformulas htm 25 EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Ef Input Data for e al ed Exit Austi 965 UTH zone K Figure 5 5 19 Drop down box menu for Lat Long to UTM option UTM coordinates will be used by EM4soil but the raw data file will be not changed If you want to use it in the future you must save it Save Processed data To convert coordinates of a survey area proceed as follows 1 select the sensor 2 go to the conversion and input the data 3 do the conversion and save the result in a file which will be input appropriately after 26 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 5 10 Medium properties The Medium Properties menu option allows the user of EM4soil to estimate values for some properties of the earth connected to the bulk electrical conductivity resistivity like water content and water content variation Figure 5 5 20 This option works on inversion models calculated previously and saved in the format X YZ Me EM4Soil 103 EM Laterally Constrained Inversion Exit Survey Open Input Data Processing Inversion Display Settings Sawe Print Action ae EdittNotePad Plot Raw Data Filter Resampling Plot Filtered Data Decimation Data Statistics Plot Resampled Da oise Analysis Correction of Plot Decimated Da Lont to UTM line ema e 0 1 Medium Properties Water Content Plot Data Statis
92. uttons with Input Sensors is selected As shown in Figure 5 4 2 DUALEM e g DUALEM 1 GEONICS EM e g EM38 GEM or PROFILER sensors can be selected 11 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion A new option of sensor named OTHER is made available in this release not shown in Figure 5 4 2 that allows the user to put together data acquired with different sensors DUALEM and GEONICS or different GEONICS sensors and jointly invert all data set see the Manual for this option However there are a few functions in the data processing that are not available for OTHER sensors EE EM4Soil v0 EM Laterally Constrained Inversion i Exit Survey Input Data Processing Inversion Display Settings Sawe Print Help About Action MM Input Sensor les Plot Raw Data Choose sensor Plot Filtered Data Plot Decimated Data Plot Noise Analysis Plot Neg Corrected Plot B L Corrected Plot Initial Model Plot Inverted Model Plot Data Resoonse Figure 5 4 2 Screenshot of the drop down box which allows the appropriate sensor to be chosen This option is a module sold separately 12 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion 5 4 2 Input Data Selecting the Input Data option enables the user to select the appropriate sensor line i e transect of o data A set of profiles mapping that is a series of transects collected as part of a la
93. w by row For EM34 Name of line N h X Y Z VDM10m HDM10m VDM20m HDM20m VDM40m HDM40m Lat Lon Time Sequence of values row by row Where N is the number of EM sensor sites e g 20 40 etc his the height of the sensor in m e g 0 1 0 2 etc X Y and Z are the coordinates in Easting and Northing and the elevation m of each EM sensor sites respectively In addition HCP and PRP or VDM and HDM are the measured at each EM sensor site Lat and Lon are the Latitude and the Longitude in decimal degrees and Time is the time in any format not exceeding 14 characters NOTE The data can be separated by comma or by space NOTE If N is negative the program will input all sites in the file assuming them as belonging to the survey NOTE If there is no data for a particular channel mode of acquisition then a value of 9 should be entered This is because the program assumes that you have used both modes for each coil spacing during acquisition NOTE If only Lat and Lon coordinates are available fill with the X and Y fields with values of 0 0 After input EM4soil will use the conversion Lat Lon to UTM option to calculate X and Y NOTE If Lat and Lon coordinates are not available fill the respective fields with 0 0 65 vy EMA4Soil v1 05 a program for 1D Laterally Constrained Inversion Example of an EM34 data file complete data set corresponding at a line of 11 sites measured in a flat zone
94. with coordinates in a local referential Lat and Lon fields are filled with 0 0 values because these coordinates are not available Line EM34 11 0 X Y Z VDM10m HDM10m VDM20m HDM20m VDM40m HDM40m Lat Long time 0 0 0 47 5 73 2 56 8 99 8 34 1 68 8 0 0 0 0 0 0 10 0 0 53 5 73 5 63 95 5 33 8 64 1 0 0 0 0 0 0 20 0 0 58 9 73 9 69 2 83 3 35 9 48 4 0 0 0 0 0 0 30 0 0 65 5 67 6 67 8 74 5 39 8 45 2 0 0 0 0 0 0 40 0 0 68 4 61 5 72 7 646 36 5 49 1 0 0 0 0 0 0 50 0 0 63 9 714 724 904 38 6 42 8 0 0 0 0 0 0 60 0 0 66 6 80 1 69 1 78 6 33 5 46 9 0 0 0 0 0 0 70 0 0 71 8 80 9 68 9 81 5 28 1 50 2 0 0 0 0 0 0 80 0 0 760 3 85 67 2 86 7 27 2 62 7 0 0 0 0 0 0 90 0 0 78 8 78 7 67 4 94 2 25 3 70 7 0 0 0 0 0 0 0 0 76 5 88 1 69 4 996 23 6 69 9 0 0 0 0 0 0 Example of an EM34 data file corresponding at a line of 11 sites measured in a flat zone but not containing the VDM20m data Line EM34 11 0 X Y Z VDM10m HDM10m VDM20m HDM20m VDM40m HDM40m Lat Long time 0 0 0 47 5 73 2 9 00 99 8 34 1 68 8 0 0 0 0 0 0 10 0 0 53 5 73 5 9 00 95 5 33 8 64 1 0 0 0 0 0 0 20 0 0 58 9 73 9 9 00 83 3 35 9 48 4 0 0 0 0 0 0 30 0 0 65 5 67 6 9 00 74 5 39 8 45 2 0 0 0 0 0 0 40 0 0 68 4 61 5 9 00 64 6 36 5 49 1 0 0 0 0 0 0 50 0 0 63 9 71 4 9 00 90 4 38 6 42 8 0 0 0 0 0 0 60 0 0 66 6 80 1 9 00 78 6 33 5 46 9 0 0 0 0 0 0 70 0 0 71 8 80 9 9 00 81 5 28 1 50 2 0 0 0 0 0 0 80 0 0 76 3 85 9 00 86 7 27 2 62 7 0 0 0 0 0 0 90 0 0 78 8 78 7 9 00 94 2 25 3 70 7 0 0 0 0 0 0 100 O 0 76 5 88 1

EM4Soil - uplink

Contents

Download Pdf Manuals

Related Search

Related Contents