Airborne geophysical data levelling based on variational mode decomposition

Abstract

ABSTRACTA new technique is proposed to level airborne geophysical data based on the assumption that the level errors along the flight‐line direction have a distinguishable centre frequency. In the levelling method, airborne geophysical data from the entire survey area are corrected automatically after sorting the data into a group according to the measurement order. Variational mode decomposition is applied to the profile data to adaptively extract the level‐error component with distinct spectral bands. In airborne geophysical surveys, data in the anomalous region show visible peaks with strong gradients along the flight‐line direction, which would impact the decomposition. To minimize their effect on variational mode decomposition, the higher amplitude anomalies are grouped and discarded by K‐means clustering to obtain the real data level. The levelling method extracts the level errors of the entire survey area simultaneously, thus avoiding the regional error caused by strong noise, missing data or error transfer in the common levelling process. Moreover, the approach is automatic and applicable to both irregular and regular line patterns without the participation of staff members or tie‐line control. The levelling method may require much time when decomposition is performed on the extended survey area. We have applied the method to the airborne electromagnetic and magnetic data acquired by Geotech Limited to confirm its validity and have compared the obtained results with those based on the flight‐line correlation levelling algorithm.