This work introduces a novel method for inverting hydrological loading displacement using 3D Convolutional Neural Networks (3D-CNN). This approach utilizes vertical displacement time series data from 41 Global Navigation Satellite System (GNSS) stations across Yunnan Province, China, and its adjacent areas, coupled with spatiotemporal variations in terrestrial water storage derived from the GRACE satellite. The 3D-CNN method demonstrates markedly higher inversion accuracy compared to conventional load Green's function inversion techniques. This improvement is evidenced by substantial reductions in deviations from GNSS observations across various statistical metrics: the maximum deviation decreased by 1.34 millimeters, the absolute minimum deviation by 1.47 millimeters, the absolute mean deviation by 79.6%, and the standard deviation by 31.4%. An in-depth analysis of terrestrial water storage and loading displacement from 2019 to 2022 in Yunnan Province revealed distinct seasonal fluctuations and a rising trend, primarily driven by dominant annual and semi-annual cycles. These cycles accounted for over 90% of the variance, with an annual increase of 1.83 millimeters. The spatial distribution of water load displacement is strongly associated with regional precipitation patterns, showing smaller amplitudes in the northeast and northwest and larger amplitudes in the southwest. This pattern underscores the significant impact of precipitation on changes in terrestrial water storage. This research findings underscore the efficacy of deep learning techniques in inverting Earth geophysical parameters and offer fresh perspectives on regional water cycle dynamics. This has profound implications for water resource management and adapting to climate change.