Analyzing groundwater storage anomalies in data‐scarce areas of Ethiopia's Rift Valley Basin using artificial neural network

Abstract

AbstractGroundwater storage anomalies (GWSA) study requires the use of hydro‐climatic models coupled with Gravity Recovery and Climate Experiment climate (GRACE) and GRACE Follow‐On (GRACE‐FO) Terrestrial Water Storage experiments. Understanding groundwater storage anomalies in the research area is difficult due to the restricted hydro‐meteorological data availability, especially, groundwater levels, changes in groundwater storage, and budgetary and technical constraints. No such investigation of GWS anomalies prior to this study in this Basin was conducted. The research focuses on GWSA investigations in the basin employing Global Land Data Assimilation Systems and Artificial Neural Networks‐Back Propagation Algorithm techniques are required. The ultimate objective of this study was to reduce the GRACE GWS from 1° to 0.25° and close the discontinuity between GRACE/GRACE‐FO observations. The results illustrated that GRACE/GRACE‐FO GWS throughout the basin were overall declining. The GWS anomaly significantly decreased between 2011 and 2013, with the highest monthly depletion in March 2011 (−158 mm). In contrast, from mid‐June to mid‐October, the maximum positive value varied from 20.80 mm in the southernmost basin to 39.49 mm in the central‐southeastern basin. The basin's eastern, central‐southeastern, northern, and southern areas displayed positive monthly GWS trends. The entire dry season has been influenced by negative anomalies. Hence, the identification of GWS variations and a comprehensive understanding of their effects provided valuable scientific insights. Identifying and analyzing anomalies in groundwater storage (GWS) can provide valuable information. This data can be efficiently used to develop appropriate regulatory frameworks, which may contribute to the long‐term viability of the availability of groundwater.