Summary

Pursuant to the requirements of DAGCAP SOP-4 v1.0 Software Validation, this document provides default input parameters for AGC processing in BTField, and guidance on when to change them, how changing them impacts the results, and documents limitations of the software.

We will first describe how to build a post-processing flow, capable of cleaning out-of-spec data according to the requirements of typical project QAPPs. The following processing steps will be configured to run in the P Flow1 processing flow of BTField.

Processing - P Flow1

Figure 1 below shows a typical P Flow1 in BTField:

Jekyll
Figure 1 - P Flow1 Processing Flow


Interp Fix

Interpolates GPS channels wherever fix quality falls outside of the specified range. The interpolation is performed over a maximum specified distance:

Jekyll
Figure 2 - Interp Fix Cleaner

Interp Fix - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the POS type because the Fix data member is found within it.

  • Default Mask: Select the Fix data member.

  • In: Specifies the data state to operate on. Typically the Fix Interp function operates on the UnFilt data state, before the data has been merged in the Raw to Merged step of the flow.

  • Max interp dist: Specifies the maximum distance to interpolate out of range Fix values. Values that are out of range beyond this distance are not interpolated and are left in the data. We recommend using the 3.0 meter default. Using larger distances may hide areas that likely have poor positioning.

  • Min/Max accept value: Together these values define the range of acceptable Fix values in the data. Values outside of this range are interpolated up-to the Max interp dist. We recommend using Min = 4.0 and Max = 4.1. Using values outside this range risks including poorly positioned data in the processed survey.

  • Remove Spikes / Max jump: Enabling the Remove Spikes allows this function to detect jumps or changes in the sequential data of a channel beyond the Max jump threshold. This function is not enabled by default, and it is typically not used for Fix data.

  • interp all members when bad data found: This setting will apply the interpolation to all members of the POS type. For example, if out-of-spec Fix data is found, the interpolation will also be applied to the Easting, Northing and Elevation channels. It recommended that this setting is always enabled for any Interp Fix functions.



Interp IMU

Removes spikes and interpolates Pitch, Yaw and Roll channels wherever Max jump is exceeded. The interpolation is performed over a maximum specified distance:

Jekyll
Figure 3 - Interp IMU Cleaner

Interp IMU - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the ATT type because the Pitch, Roll and Yaw (PRY) data members are found within it.

  • Default Mask: Select the Pitch, Roll and Yaw (PRY) data members.

  • In: Specifies the data state to operate on. Typically the IMU Interp function operates on the UnFilt data state, before the data has been merged in the Raw to Merged step of the flow.

  • Max interp dist: Specifies the maximum distance to interpolate out of range PRY values. Values that are out of range beyond this distance are not interpolated and are left in the data. We recommend using the 3.0 meter default. Using larger distances may hide areas that likely have poor positioning.

  • Min/Max accept value: Together these values define the range of acceptable PRY values in the data. Values outside of this range are interpolated up-to the Max interp dist. We recommend using Min = -5.0 and Max = 5.0 radians. Using values outside this range risks including poorly positioned data in the processed survey.

  • Remove Spikes / Max jump: Enabling the Remove Spikes allows this function to detect jumps or changes in the sequential data of a channel beyond the Max jump threshold. This setting is typically enabled for the Interp IMU function, with a recommended Max jump default of 0.1 radians. Using a lower Max jump may result in interpolating over valid PRY data. Using a higher Max jump may result in leaving PRY spikes in the processed data.

  • interp all members when bad data found: This setting will apply the interpolation to all members of the ATT type. For example, if out-of-spec Roll data is found, the interpolation will also be applied to the Pitch and Yaw channels. It recommended that this setting is always enabled for any Interp IMU functions.



Re-Calc GPS Heading

Fixes GPS heading flips when the platform has little or no forward motion:

Jekyll
Figure 4 - Re-Calc GPS Heading

Re-Calc GPS Heading - Recommended Defaults

  • Name: Use a name that best describes the operation.
  • Wnd: The length (meters) of the window used for the heading calcuation. It is recommended to use a 2.0 meter window. Using a shorter window length will risk calculating erratic heading values. Using a longer window length will risk calculating heading values that are unresponsive to actual movement of the platform.



Raw to Merged

Merges all raw data to the time base of the EM sensors:

Jekyll
Figure 5 - Raw to Merged

Raw to Merged - Recommended Defaults

  • Name: Use a name that best describes the operation.
  • apply ATT mounting correction: The option to apply a mounting correction for the ATT data is only used in rare cases when the ATT sensor has been mounted incorrectly.



Snip bad fix

Snips all data outside of the fix quality range of 4.0 to 4.1. This removes all data that could not be interpolated in step 1, due to the maximum range of 3 meters constraint:

Jekyll
Figure 6 - Snip bad fix

Snip bad fix - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the POS type because the Fix data member is found within it.

  • Default Mask: Select the Fix data member.

  • Min/Max accept value: Together these values define the range of acceptable Fix values in the data. Fix values found outside this range are snipped from the processed data. We recommend using Min = 4.0 and Max = 4.1 . Using a different range risks including poorly positioned data in the processed survey.

  • Snip all sensors: This setting will apply the snip to all sensors (all data). For example, if out-of-spec Fix data is found, the snip will be applied to all data types and their members. It recommended that this setting is always enabled for any Snip Fix functions.

  • Use slope: When selected all the Min / Max accept values will apply to the value of the slope of the data. This option is typically not used when snipping bad fix data.

  • Slope wnd: The length of a moving window that will calculate the slope. This option is typically not used when snipping bad fix data.

  • snip zero slope: This gives the user the option to ignore zero slopes when detecting bad decays. This option is typically not used when snipping bad fix data.

  • Apply to decays: This option applies only to Rx data, therefore it is not used when snipping bad fix data.

  • Remove spikes: Remove spikes that exceed the Max jump value. This option is typically not used when snipping bad fix data.

  • Max jump: Threshold for the Remove spikes. This option is typically not used when snipping bad fix data.



Snip low current

Snips all data below a transmitter current threshold:

Jekyll
Figure 7 - Snip low current

Snip low current - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the Tx type because the Current data members are found within it.

  • Default Mask: Select all Tx Current data members.

  • Min/Max accept value: Together these values define the range of acceptable Current values in the data. Values outside this range are snipped from the processed data. We recommend using Min = 15 and Max = 1000000 Amps. Using a Min less than 15 Amps risks including poorly illuminated anomalies in the processed survey.

  • Use slope: When selected all the Min / Max accept values will apply to the value of the slope of the data. This option is typically not used when snipping low current.

  • Slope wnd: When selected all the Min / Max accept values will apply to the value of the slope of the data. This option is typically not used when snipping low current.

  • snip zero slope: This gives the user the option to ignore zero slopes. This option is typically not used when snipping low current.

  • Apply to decays: This option is not used when snipping low current.

  • Remove spikes: Remove spikes that exceed the Max jump value. This option is typically not used when snipping low current.

  • Max jump: Threshold for the Remove spikes. This option is typically not used when snipping low current.





Snip low velocity

Snips all data outside an acceptable velocity range:

Jekyll
Figure 8 - Snip low velocity

Snip low velocity - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the POS type because the Velocity data member is found within it.

  • Default Mask: Select the Velocity data member.

  • Min/Max accept value: Together these values define the range of acceptable Velocity values in the data. Values found outside this range are snipped from the processed data. We recommend using Min = 0.2 and Max = 100 m/s. Using a Min less than 0.2 m/s risks calculating erratic GPS heading data in the processed survey.

  • Use slope: When selected all the Min / Max accept values will apply to the value of the slope of the data. This option is typically not used when snipping low velocity.

  • Slope wnd: When selected all the Min / Max accept values will apply to the value of the slope of the data. This option is typically not used when snipping low velocity.

  • snip zero slope: This gives the user the option to ignore zero slopes. This option is typically not used when snipping low velocity.

  • Apply to decays: This option is not used when snipping low velocity.

  • Remove spikes: Remove spikes that exceed the Max jump value. This option is typically not used when snipping low velocity.

  • Max jump: Threshold for the Remove spikes. This option is typically not used when snipping low velocity.





Detrend

This function applies a standard de-median filter to the Rx type data. A background is estimated by computing the median over a time window, while rejecting data outside of the upper/lower cutoffs. The user has the option to filter all lines or a subset of lines. When filtering a subset of lines, it is recommended to create an additional detrend function in the flow.:

Jekyll
Figure 9 - Detrend

Detrend - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • Type: Select the Rx type because all Tx-Rx pairs and time gate members are found within it.

  • Default Mask: Select the all data mask to detrend all Rx data.

  • In: Specify a named data state to detrend. Typically we detrend the unfilt data state.

  • Out: Specify a named data state to assign the result to. Typically we assign the result of the detrend filter to the -a data state.

  • Process all lines / Select lines: Configure this function to apply filtering to all lines or a subset of lines. Typically, the default of Process all lines is used.

  • Filter Wnd: The length (m) of the moving window used to estimate the background. It is recommended to use the default value of 10 m. Using shorter window lengths in dense areas will risk inclusion of actual anomaly response in the background estimation. Using longer window lengths will lead to unecessarily long execution times.

  • Upper CutOff: The upper percentile cut-off to exclude data from being used to compute the median background for the window. The default value of 70% is recommended.

  • Lower CutOff: The lower percentile cutt-off to exclude data from being used to compute the median background for the window. The default value of 30% is recommended.

  • Recalculate cumulative distance: Option to re-calculate the cumulative distance. It is recommended to enable this option. This option is enabled by default.




Target Picking - Auto-Pick Flow

Figure 10 below shows a typical Auto-Pick flow in BTField:

Jekyll
Figure 10 - Auto-Pick Flow

Detect Targ

This function performs standard threshold target picking using the Blakely Peak Detector Algorithm (Blakely, 1986):

Jekyll
Figure 11 - Detect Targ

Detect Targ - Recommended Defaults

  • Name: Use a name that best describes the operation.

  • In: The input data state for target picking. This is typically set to the -a data state that has been detrended.

  • Use noise threshold: When selected, this option will extract the noise estimate from the active noise polygon, then multiply it by the Noise scalar value to determine the detection threshold.

  • Noise scalar: The default is 5.0 times the active noise estimate. This will auto-calculate the detect threshold uV/A. Values as high as 10.0 can be used in higher background response areas. values higher than 10.0 risk missing targets of interest (TOI). Setting values lower than 5.0 will typically detect an unnecesarily large number of targets.

  • Use custom threshold: When selected, this option will set the detect threshold directly, and ignore the active noise estimate. To determine the appropriate threshold, it is recommended that a Detection Modeller View be inserted into the BTField project and then used to determine a worst-case detection threshold for the required detection depth of the smallest TOI at the site.

  • Kernal Radius: Defines the circular area around the evaluation point where a 2D peak must be present in the data. It is recommended to use the default of 0.75 m. Using larger values will risk under-picking (with longer processing times). Using smaller values will risk over-picking.

  • Grid Sep X: The easting resolution of the grid (i.e. the distance in meters of one grid pixel in the easting direction). The default value is 0.05 m. Using larger values will risk missing small TOI. Using smaller values will lead to unnecessarily long processing times.

  • Grid Sep Y: The northing resolution of the grid (i.e. the distance in meters of one grid pixel in the northing direction).

  • Blanking: The max distance to grid beyond the collected data. It is recommended to use the default of 0.30 m

  • Merge Dist: The distance between two or more targets where a decision is made to select only the target with the highest peak amplitude. The default value of 0.60 m is recommended. Using smaller values may generate more picks. Using larger values will risk missing TOI.

  • SUM: The mask indicating the range of time channels that will be summed and used to create the grid for the target detection. This is typically set to a mid-range of time gates (from 400 us to 850 us). Using an earlier range will risk including saturated responses. Using a later range will risk a loss of response in the generated grid used for the target detection.

  • Tx-Rx pairs: This button will display a top-down view of the platform and allow the user to select the Tx-Rx pairs used to create the grid for the target detection.

  • Pick on Lines: An option to pick target along the Rx profile lines only (i.e. a grid is not used for the target detection).

  • Positive peaks: When enabled, positive peaks are detected. This is enabled by default.

  • Inside boundaries only: When enabled, targets will only be detected inside the site’s boundary inclusion polygons. This option is disabled by default.

  • Ignore SU boundaries: When enabled, Survey Unit (SU) boundaries will be ignored during the target detection. This option is disabled by default.

Tags: