1D along-track pre-processing of the GOCE gravity gradients and nonlinear filtering of the radial components Vzz in spatial domain

Abstract

The paper presents pre-processing and nonlinear filtering of the GOCE gravity gradients (GGs) in spatial domain. At first, a multiple fully automatic 1D along-track pre-processing of the GOCE GGs in the gradiometer reference frame (GRF) is introduced. This tool is based on the nonlinear diffusion filtering in 1D while treating issues like a detection of ‘jumps’ in time series, outliers and unforeseen systematic tendencies. We focus on the 1D along-track pre-processing of the GOCE GGs in GRF for the last period of the GOCE satellite mission (from 2012-12-01 to 2013-10-19) that are later transformed into the local north-oriented frame (LNOF). Afterwards, the radial components V_zz are reduced on a reference surface chosen as a mean value of all altitudes of the GOCE satellite orbits from the processed period. Here they are filtered using the nonlinear diffusion filtering on a closed surface. Finally, a grid of the filtered radial components V_zz in LNOF on the reference surface is obtained.