Multivariate variational mode decomposition to extract the stripe noise in GRACE harmonic coefficients

Jian, Guangyu1; Zou, Fang1; Xu, Chuang1,2,3; Yan, Zhengwen4

2024-07-30

10.1093/gji/ggae241

GEOPHYSICAL JOURNAL INTERNATIONAL   影响因子和分区

0956-540X

1365-246X

In this work, a novel method has been developed to remove the north-south stripe noise in the Level-2 spherical harmonic coefficient products collected by the Gravity Recovery and Climate Experiment (GRACE) mission. The proposed method extracts the stripe noise from the equivalent water height (EWH) map via the Multivariate Variational Mode Decomposition algorithm. The idea behind our method is to extract the cofrequency mode in multiple-channel series in the longitude direction. The parameters of our method are empirically determined. The investigation in a closed-loop simulation proves the improvement of our methods compared with the Singular Spectrum Analysis Spatial (SSAS) filter. Subsequently, the spatial-domain and spectral-domain investigations are conducted by using real GRACE data. Our method only suppresses stripe noise at low latitudes (30 degrees S-30 degrees N) and imposes an order-dependent impact on spherical harmonic coefficients but with potential oversmoothing. Meanwhile, the well-documented water level proves that our method further reduces outliers in a time-series of localized mass variations compared with the SSAS filter. More importantly, users are allowed to reduce the filtering strength of our method to preserve small-scale strong signals while suppressing stripe noise. Moreover, noise levels over the ocean at low latitudes are evaluated as well. The noise level of our method using empirical parameters is 32.48 mm of EWH, with 31.54 and 53.52 mm for DDK6 and SSAS, respectively. Our work introduces a novel method to address the issue of north-south stripe noise in the spatial domain.

Satellite geodesy Satellite gravity Time variable gravity Spatial analysis

SCI ; EI

英语

其他

National Natural Science Foundation of China[42274004] ; National Natural Science Foundation of China[42204012] ; Young Scientists Fund of the National Natural Science Foundation of China[2022A1515010396] ; Guangdong Natural Science Foundation[2023A04J1527] ; Guangzhou Science and Technology Plan Project[PGMF-2024-P011]

Geochemistry & Geophysics

Geochemistry & Geophysics

WOS:001280642000003

OXFORD UNIV PRESS

GEOSCIENCES

Web of Science

1.Guangdong Univ Technol, Sch Civil & Transportat Engn, Dept Geodesy & Geomat Engn, Guangzhou 510006, Peoples R China
2.Huazhong Univ Sci & Technol, Natl Precise Grav Measurement Facil, Wuhan 430074, Peoples R China
3.Guangdong Univ Technol, Cross Res Inst Ocean Engn Safety & Sustainable Dev, Guangzhou 510006, Peoples R China
4.Southern Univ Sci & Technol, Dept Earth & Space Sci, Shenzhen 518055, Peoples R China

Jian, Guangyu,Zou, Fang,Xu, Chuang,et al. Multivariate variational mode decomposition to extract the stripe noise in GRACE harmonic coefficients[J]. GEOPHYSICAL JOURNAL INTERNATIONAL,2024,238(3).

Jian, Guangyu,Zou, Fang,Xu, Chuang,&Yan, Zhengwen.(2024).Multivariate variational mode decomposition to extract the stripe noise in GRACE harmonic coefficients.GEOPHYSICAL JOURNAL INTERNATIONAL,238(3).

Jian, Guangyu,et al."Multivariate variational mode decomposition to extract the stripe noise in GRACE harmonic coefficients".GEOPHYSICAL JOURNAL INTERNATIONAL 238.3(2024).