Use of GRACE time-variable data and GLDAS-LSM for estimating groundwater storage variability at small basin scales: a case study of the Nzoia River Basin

Abstract

This study integrates time-variable Gravity Recovery and Climate Experiment (GRACE) gravimetric measurements and Global Land Data Assimilation System (GLDAS) land surface models (LSM) in order to understand the inter-annual variations and groundwater storage changes (GWSC) in the Nzoia River Basin in Kenya, using the water balance equation and parameters. From averaged GRACE and GWSC data, the results showed that over the 10-year period, the basin experienced a groundwater depth gain of 6.38 mm year−1, which is equivalent to aquifer recharge of 298 million cubic metres (mcm) year−1. The deseasonalized groundwater variation analysis gave a net gain in groundwater storage of 6.21 mm year−1 that is equal to a groundwater recharge gain of 290 mcm year−1. The observed results are comparable to the groundwater safe yield of 330 mcm year−1 as estimated by the Water Resource Management Authority in Kenya. Through cross-plotting and analysis with averaged satellite altimetry data and in situ measurements from rainfall and streamflow discharge, the total water storage change (TWSC) and GWSC in the basin were consistent and closely correlated in variation trends. The inter-annual standard deviation of groundwater change was determined as ±0.24 mm year−1, which is equivalent to 85% degree of confidence in the obtained results. The results in this study show that GRACE gravity-variable solutions and GLDAS-LSM provide reliable data sets suitable for the study of small to large basin groundwater storage variations, especially in areas with scarce and sparsely available in situ data.