8.1 Data validation
The data sets need to be carefully inspected to validate the usability of the measurement results. Data validation shall incorporate different actions, including:
- Coordinate validation: The data sets have to be thoroughly examined to correct positioning errors if any, using PDOP
(quality of GPS) or any accuracy information in combination with background map or visual localization.
- Incorrect values: The data sets have to be thoroughly examined to exclude the cases presenting missing or wrong values
(probe saturation or probe failure for instance).
- Detection limits (DL): Values reported as below the detection limits shall be considered as valuable information. Depending
on the intended processing algorithm the values can be replaced by either the detection limits, or by different fractions
of the detection limits. It is conventionally adopted in Geochemistry to use half of the DL value for data interpretation.
- Duplicates: Results that can be considered as duplicates (e.g. measurements performed within a distance less than the
targeted spatial resolution in a grid plan) must be removed or replaced by a single value (the minimum, the maximum or
the average of the defined cluster), depending on the targeted interest. If the interpretation intends to be conservative,
the average of such two neighboring measurements can be adopted. In case of radiation safety concern, it is preferred to
use the maximum of the two values.