Tutorial 10: Anomaly Detection

Database and Grid Setup

Visit the Minty Geophysics website at www.mintygeophysics.com to get background information on the anomaly detection (GAMMA_Target). The Anomaly Detection tool requires a database containing final potassium, uranium and thorium radioelement concentrations, as well as ERMapper grids of these database fields. Put the grids in a sub-directory of the working directory.

Use either Intrepid or Oasis Montaj to set up all the input database fields required by the tool. You will see which fields are required at Step 3 below. Of course, for an Intrepid database, the line number, flight number and fiducial fields must also always be present in the database – with prescribed names as indicated in the "Preparing The Database" section.

Processing Steps – Offshore-Onshore Lines

    1. Open a database from the Main Window – File|Open. For the demonstration dataset select "demo_data\airborne_spectrometry\anomaly_detection\ANOMALY_DETECTION..DIR" (or the Oasis Montaj equivalent).
    2. Launch the tool – Enhancements|Anomaly Detection.
    3. Set the database aliases – File|Set Aliases. Click the Load From File button and select the "aliases_anomaly_detection.pc" control file from the working directory. For an explanation of the aliases click the Help button. Make sure that the Grid Directory is set to the full path to the grids in "demo_data\airborne_spectrometry\anolmaly_detection\GRIDS". If necessary, click the file icon button to navigate to the grid directory to set the path. Click Save And Return.
    4. Set the tools parameters – File|Set Parameters. Click the Load From File button and select the "parameters_anomaly_detection.pc" control file from the working directory. For an explanation of the parameters click the Help button. These parameters cover all the anomaly types. Click Save And Return.
    5. Initiate anomaly detection. Click the Run button. We are searching for Spectral anomalies using the gridded data (Grid Spec anomalies). 
    6. View the log file – File|View User Log File to see the full results. The main output is to a shape file in the working directory.

Try other anomaly types. The Anomaly Type, Shape File name and target threshold (Tgt%Thresh) can conveniently be set in the tool window without having to return to the Set Parameters window. The anomalies are displayed as a color plot with the colour coded according to anomaly type: K – red, Th – green, U – blue (and blends of these for mixed anomalies). You will have to load the anomaly shape file into Oasis Montaj, or a GIS, to see what anomaly attributes are available for anomaly assessment.

Note: Cesium anomaly detection requires raw spectra. To run the Cs anomaly detection demo, navigate to the "C:\GammaSpec\demo_data\airborne_spectrometry\cesium_detection_maralinga" directory and load the "CESIUM_DETECTION..DIR" database. The tool creates two new database fields: "csRatio" is the ratio between the counts in a window targeting the Cs-137 peak at 0.662 MeV and an adjacent window, and "csDifference" is the difference between the two window count rates. These fields can be gridded and imaged for analysis/interpretation. Anomalies are detected where deviations from the mean "csRatio" values exceed "csThreshold" standard deviations.