Study of different filtering techniques applied to spectra from airborne gamma spectrometry.

One of the features of the spectra obtained by airborne gamma spectrometry is the low counting statistics due to a short acquisition time (1 s) and a large source-detector distance (40 m) which leads to large statistical fluctuations. These fluctuations bring large uncertainty in radionuclide identification and determination of their respective activities from the window method recommended by the IAEA, especially for low-level radioactivity. Different types of filter could be used on spectra in order to remove these statistical fluctuations. The present work compares the results obtained with filters in terms of errors over the whole gamma energy range of the filtered spectra with the window method. These results are used to determine which filtering technique is the most suitable in combination with some method for total stripping of the spectrum.