SP Processing Tutorial
SP Processing Tutorial
Data correction
Version: 10 April 2012 (contact us to obtain the last update)
After the survey, the entire SP database is reconstructed choosing a unique reference
station. This process requires two corrections of the raw SP data:
- reference correction
- closure correction (or loop correction)
1/18
We use Excel (other spreadsheet applications can be used, like Grapher) to process
the data and Surfer (a contouring and 3D surface mapping program) to interpolate, build maps
and graphics with the results. In a graph, SP data is represented as a function of the distance.
2/18
2- Correct a self-potential profile
2.1- Reference correction
A reference correction must be applied to a SP profile if this one has been performed changing
the position of the reference electrodes.
The reference correction is made to join the different parts of a same SP profile, correcting the
various changes of reference electrode. Indeed, each time the difference of electrical potential is
measured from a new reference, the measurements starts again from 0 mV. Each section of the profile
collected with a single reference must be shifted to the end of the previous section (see figure 2.1).
Only the first section will remain unchanged will the others will be shifted section by section in the
direction in which the measurements were acquired.
3/18
Fig. 2.2. Reference correction of
the SP data in Excel.
The reference
correction is made section by
section of the profile, from one
reference to the next one. In
the Excel sheet, we must add
the last value of the previous
section to all the
measurements of the section
we are currently correcting. In
Figure 2.2, the new references
are evidenced by the orange
rows. These orange rows
correspond to the orange dots
in Figure 2.1).
The initial section (row
3 to 18) depends directly on
the first reference so that no
correction is applied to it.
Section 1 (from row 19 to 33
in this example) is shifted to
the last value of the initial
section; section 2 (row 34 to
48) is shifted to the last value
of section 1. We write the
formula at the first row of each
section and drag it until the
last row of the same section.
In the case of a closed profile, the first point is identical to the last one so the SP value
measured should theoretically be the same. This would be true if no environmental perturbation
occurred between the moments when the first and the last measurements were made. However during
the survey, the measurement conditions can change (e.g. soil moisture, soil temperature, instrumental
error, etc) and a drift will be observed. Using nonpolarizing electrodes drastically reduces the
instrumental drift.
4/18
We consider that the drift increases regularly from the first point to the last one: along the
period of time of the acquisition, the drift accumulates. This drift is considered as parasitic and it must
be corrected. This is the closure correction.
In the example presented in Figure 2.3, we imagine a dataset containing 10 datapoints and
where the first point is geographically the same than the last one (closed profile).
Fig. 2.3a: the drift is the difference of SP between the first and the last point.
Figure 2.3b and c: the correction of this drift is applied on all the datapoints of the profile, increasing
the correction factor from the first to the last point. The initial point is the reference of the profile so
that no correction is applied to it. The first point (after the reference) will be corrected for 1/10th of the
total drift (where10 is the number of data points without counting the reference). The second point will
be corrected for 2/10th of the total drift, etc until the last point which will be corrected for 10/10 th of
the drift (i.e. the total drift) because it must reach the same value than the reference point (see the
orange dots in Figure 2.3c).
5/18
Self-potential - circular profile
120
100
80
60
SP (mV)
40
Serie1
20 Serie2
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
-20
-40
-60
distance (m)
Fig. 2.4. Example of real-case closure-correction on a closed SP-profile. The black line is from the original
dataset (previously corrected for the reference correction) and the orange line is the same profile corrected for
the closure correction. Note that the last point is corrected to reach the same value than the reference point
(geographically the same point). Also note how this correction deforms the graph without eliminating the SP
anomalies (positive and negative peaks).
In the following example Excel-sheet (Fig. 2.5) we can notice that the reference point (row 3)
and the last point (row 48) have the same X and Y GPS coordinates but a different SP value after
reference correction (see cell G3 and G48). As said before, this difference is the drift that we must
correct. The correction is made in column H and the corresponding equation is presented on the right
side of the figure.
First we must create a column with the numeration of the datapoints, beginning at 0 for the
reference (see column A). This column will be used to settle the degree of correction applied to each
point: the reference will need no correction; that is why we assign it the number 0 (cell A3). The
correction begins at the last point and is dragged until the reference point. This allows to see
immediately if the formula has been written properly (by comparing the result with the reference
point). Some elements of the equation are constants but other will vary as the correction is applied
unequally to each point of the profile.
6/18
Fig. 2.5. Closure correction of the SP data in Excel.
In the example shown in Fig. 2.5, the correction equation in Excel is composed as follow:
- Calculus of the drift (D) = $G$48 - $G$3 (1)
This is a constant in the equation for all the profile (that is why we use the $ sign).
- Divide the drift by the number of datapoints in the profile (N; also a constant).
= ($G$48 - $G$3)/$A$48 (2)
7/18
This is a variable A48 is for the correction applied to the last point. When we will drag the formula
in the Excel sheet, this will change as a function of the row (that is why we do not use the $ sign in this
part of the equation).
- Subtract this value to the SP value already corrected for the reference correction (SPr)
= G48 - (($G$48 - $G$3)/$A$48)) * A48 (4)
This is also a variable that will change when we drag the formula.
Remark: note that equation (3) correspond to green arrows in figure 2.3.c and d
As seen before, SP measurements are performed forming loops in order to estimate the drift and
correct it (closed profiles or profiles closed one on the other). When a dataset is composed of several
profiles (in areas covered to build a SP map), each profile must be corrected independently for the
reference correction, as described in section 2.1.
Remark: In all the following description of the correction process we consider that the reference
correction has been performed on all the profiles.
After the reference correction, each profile has got its unique reference at 0 mV. In order to join
all the profiles together, we must choose a first loop which will be the reference closed profile for the
next steps of the closure correction (see step 1 for a, b, c, and d in Figure 3.1). The other profiles will
then be used to form several loops, joined step by step to the data already corrected.
8/18
In the successive steps proposed in Figure 3.1, the black doted lines are the data already
corrected. Steps a4, b4, c5, and d5 correspond to the complete datasets. The green sections are the
loops connected step by step to the data already corrected. A loop can be composed of part of one
profile, a single profile or more than one profile. A valid loop is a section which two extremities are
connected to sections already corrected for the closure correction but that does not include other points
already corrected.
Remark: a rule to keep in mind is that after being corrected for the closure correction, a datapoint is
fixed its value cannot be modified during the following steps of correction.
Remember that after the reference correction, all the profiles have a reference (first point of the
profile) at 0 mV. After step 1 (see section 3.1) we choose a second loop to correct.
9/18
reference loop, at the connection, to all the points of the new loop (Fig. 3.3).
On Figure 3.2.b we observe that, after the shift, the other extremity of the new loop is not equal
to the same geographical point in the reference loop. The difference observed is the drift that must be
corrected following the closure correction equation (see section 2.2).
In the Excel sheet (Fig. 3.3) we apply the equation using the data from the two connected loops
and drag the formula along all the data of the loop to be corrected.
At the end of step 2, the two extremities of the newly corrected loop must be at the same SP
value that the reference loop (Fig. 3.2.c). The data from this loop are now fixed and can be used in the
next steps of the correction (as schematized in Fig. 3.1).
Fig. 3.3. Closure correction of the SP data for the example schematized in Figure 3.2. Loop 1 is the reference
loop (black colour in Fig. 3.2); Loop 2 is the new loop (to be corrected; green colour in Fig. 3.2). The upper part
of the figure refer to the shift of the SP data and the lower part to the closure correction strictly speaking.
10/18
Tutorial for Self Potential data processing II
Data presentation in Surfer
SP data can be presented in graphs as a function of the distance or as a function of the altitude.
The study of this type of graph can be interpreted for localizing discontinuities in the soil (e.g.
circulation of fluids along faults), hydrothermal convection, etc Usually this can be done joining the
results of other geophysical or geochemical methods. With an appropriate dataset, we can also present
SP data in maps.
This part of the tutorial presents the various options you can use to present self-potential (SP)
data with Surfer. It can be applied to other types of data but we will take the example of SP data all
along the tutorial. It does not describe the Surfer functionalities with full details as this information is
clearly described in the Surfer help.
In addition to the basic and intuitive menus (File, Edit, etc), two key menus for creating maps
in a plot document of Surfer are the Grid and Map menus (Fig. 2.1). The Grid menu leads to several
commands used to manage data files (create a grid, extract part of a grid, modify the data into a grid,
11/18
etc). The Map menu allows choosing among several map representations (contour, images with a
colour-scale, shaded relief, surface, etc). Most of the commands available in the Map menu create
maps based on a grid.
The third useful menu is the Help menu where full step by step descriptions and definitions of
the Surfer function are available. You can also follow tutorials for the main functionalities of Surfer.
For creating SP maps we must use the file containing the SP data corrected for the reference
and the closure corrections. We also need the X and Y GPS coordinates.
Data-files contain data acquired in the field, often with an irregular geographical repartition.
This data must be converted into an evenly spaced grid before using many of Surfer's features. A grid
file contains a list of Z values (corresponding to the SP data, in our case) organized in rows and
columns. The position of a data value into the grid corresponds to the X and Y geographical
coordinates of the datapoints.
12/18
- If the Excel file contains more than one worksheet, you will have to pick the one created at step (1).
- A new window opens to adjust the gridding parameters (Fig. 3.1)
- In that window you must select the columns of the Excel worksheet corresponding to the geographical
coordinates (X and Y) and to the SP data (Y).
- Create an output grid file clicking on the button to change the filename and choose an
emplacement to save the file.
- Click on OK.
- Close the Surfer confirmation window and close or save the Surfer gridding report (we usually close
it without saving) appearing on the screen after the end of the process.
- Your grid file is saved and ready to be used.
Remark: creating a grid file you can usually accept all of the default gridding parameters (Gridding
Method and Grid Line Geometry) and generate a grid file that represents your data well.
The repartition of SP measurements in the field is often irregular and the grid can be affected. A
solution for a better result is to perform two successive griddings of the data:
- Use the Grid | Data command to create a grid with a spacing about five times the spacing of the
measurements along the profiles (i.e. enter manually a 100 m spacing value in X and Y for a field
measurement spacing of 20 m).
- The resulting grid must now be added to the original dataset. Use the Grid | Convert command to
convert your .grd file in an ASCII XYZ file. The resulting file is a .dat that you can open in Excel or in
a Surfer worksheet. Copy the data from the converted grid in the same worksheet that the original data.
- Use this new file and the Grid | Data command to create a new grid with a spacing equal to the
spacing of the measurements in the field (20 m if we follow the same example).
The resulting grid is the one you can use to create your SP maps.
13/18
3.2- Create a SP map and enhance its presentation
The SP grid we created can be used to generate various types of maps. In this section, we will
see the useful commands to create them and show some examples of the resulting maps. The process is
described on the left and the result on the right:
14/18
(3) Cut off part of the map
You can now create a new image map with the blanked grid and overlay it with a post map of
the data (as described previously). Alternatively, you can modify the source-file of the image map that
already exists (the one created in section (2)):
- Double-click on the image map
- In General, modify the Input Grid by clicking on and selecting the blanked grid you have just
created.
15/18
(4) Overlay your map on a DEM
Contour maps can help visualizing better the distribution of the SP anomalies highlighting the
SP transitions, not ever easy to visualize with the color-scale. A second interest is to make possible
visualizing the SP data overlaying the contour map on any kind of raster image (e.g. aerial photo,
geological map, etc).
16/18
Overlaying the data on a raster:
- Create a contour map as shown
before and modify the Levels
properties (double-click on the
contour map):
Click on Line, select Gradational,
click on the Color bar and load or
create a colour scale for SP data.
- Validate until all the Properties
windows are closed.
- Double-click on this base map and in Base Map, set the Image Coordinates as a function of the
corresponding GPS coordinates
- Click on OK
- Overlay the contour map together with the base map, the surface of the DEM, and with white
datapoints.
Remark: the Overlay Maps command sometimes gives a bad aspect overlaying contour maps or post
maps on a shaded relief map or a 3D surface (blurry aspect). If this happens you can click-right on the
contour map or the post map (on the left window of Surfer) and click on Break Apart Overlay. This
will separate the selected map. You then can place it manually above the shaded relief map or the 3D
surface (however, this excludes the possibility to change the orientation of the 3D surface.
17/18
- Choose the GPS limits of the DEM to extract (Xfirst-km, Xlast-km, Yfirst-km and Ylast-km; see Fig
4.1).
Remark: last and first are the terms used in the Extract Grid window (see Fig. 4.2)
with Xminkm et Yminkm the minimum coordinates of the original DEM ( bottom left end corner
of the DEM; see Fig. 4.1). To see it, double-click on the original DEM in the main window and check
the Limits; and for a 25 m DEM, spacing = 0.025 (km)
- Enter the four X and Y limits (just calculated in nodes) in the lower part of the window (see Fig. 4.2)
- Click OK
Surfer allows extracting a profile from any grid file. We can create topographic profiles from a
DEM, SP profiles, temperature profiles, etc
18/18
- Choose a name for your resulting .bln and .dat files (see the Surfer help Slice section for details on the
files content)
19/18