A boundary compensation approach based on bemd for geomagnetic anomaly field data

Boundary effect is one of the most concernful problems in analysis of geomagnetic anomaly field data, it introduces great error in analytical result and results in serious thread in engineering practice. This problem reflects on many aspects, such as reference map preparation, multi-resolution decomposition, etc. Improved bi-dimensional empirical mode decomposition (BEMD) is firstly proposed and used to decompose experiment data into bi-dimensional intrinsic mode functions (BIMFs). For all components, scale-feature is defined to quantitatively measure their scale, and then apply total Hilbert transform to compute instantaneous amplitude and frequencies for all BIMFs. By dividing the components into blocks of certain size, a boundary compensation database with tree structure is established as a general compensation resource for any datasets in need. As this approach fully respect the distribution law of geomagnetic anomaly field data, this approach could well improve the boundary compensation precision and possesses high efficiency and wide applicability.