Accurate error covariance information of the Spherical Harmonic Coefficients (SHCs) of the GRACE (Gravity Recovery and Climate Experiment) monthly gravity field model is crucial for the precise data post-processing. Unfortunately, most GRACE products only provide formal errors of SHCs, disregarding their high correlations due to massive datasets, especially in high degrees and orders. Therefore, we propose a semiparametric covariance reconstruction approach for simplifying covariance structure and reconstructing time-varying covariance matrices using common eigenvectors and individual formal errors. The covariance matrices of Tongji-Gace2018 and GFZ RL06 products from April 2002 to December 2016 are used as training data to derive the common eigenvectors and covariance reconstruction model, which are utilized to reconstruct the covariance matrices of any GRACE product.

Contact: 2011480@tongji.edu.cn

Data and codes linkage: https://github.com/LinZhangTJ/Covariance_Reconstruction