Accurate estimation of non-seasonal signals (NSSs) of Terrestrial Water Storage Anomaly (TWSA) from Gravity Recovery and Climate Experiment monthly gravity field models is essential for identifying and understanding extreme hydrological phenomena. However, significant north-south striped noise in the models necessitates spectral filtering before estimating NSSs, resulting in signal attenuation and leakage. In this paper, we propose a one-step approach (OSA) that iteratively filters noise and estimates NSSs alongside trends and seasonal signals starting from unfiltered regional TWSA signals, where the covariance matrices of NSSs are populated using distance-based exponential functions. The non-seasonal TWSA signals in Southeastern China, estimated by OSA from April 2002 to December 2024, effectively preserves signal integrity with reduced spatial leakage and enhanced signal strength, aligning closely with those of the RL06 mascon products from CSR (Center for Space Research) and JPL (Jet Propulsion Laboratory), achieving Nash-Sutcliffe Efficiency (NSE) of 0.91 and 0.90. Moreover, we introduce a Standardized NSS (SNSS) index from OSA, which enhances the consistency with the standardized streamflow index, identifying the extreme wetness in pearl river basin (PRB) and Southeastern River Basin (SERB) from August 2015 to June 2016, and the extreme drought in Middle and Lower Yangtze River Basin (MLYRB) from July 2022 to April 2023. SNSS also exhibits enhanced correlations with nine key climate indices, especially for ENSO (El Niño-Southern Oscillation) and TIOS (Tropical Indian Ocean Sea Surface Temperature Anomaly), with cross-correlations of 0.99 and 0.96 for PRB, 0.97 and 0.94 for SERB during extreme wetness, and 0.96 and 0.90 for MLYRB during extreme drought.

Read original article

Read original article