SPATIAL-TEMPORAL ANALYSIS OF CHANGES IN GROUNDWATER STORAGE USING GRACE DATA AND A HYDROLOGICAL MODEL IN EWASO NYIRO RIVER BASIN

Abstract

The focus of this study was to determine the applicability of using GRACE satellite data in studying the Groundwater storage fluctuations. To do so the Total Water Storage anomaly from GRACE mascon JPL was used. The computation of this anomalies was key in determining very accurate results and for this to be achieved, the state of reliable preprocessing of the gravity data from GRACE was key. We obtained the preprocessed GRACE mascon with a 300km smoothing radius for it provided the best results as compared to other available smoothing radii. The aim of the project was to evaluate the potential of Gravity Recovery and Climate Experiment (GRACE) satellite data in assessing rate of Groundwater depletion on a medium scale river basin that is Ewaso Nyiro river basin The application of Soil Water Analysis Tool SWAT model provided a great insight into the testing of the reliability of the GRACE satellite data. Despite the small size of the ENNB compared to GRACE spatial resolution, good agreement between GRACE solutions and hydrological model TWSA has been found (maximum correlation coefficient 0.61 and Nash-Sutcliffe Efficiency, NSE, 0.45). These datasets showed that TWSA in the ENNB is mainly due to water stored in the first few meters of soil and in the shallow aquifer. And on other parts of the basin in the deep aquifers i.e. Merti aquifer. The precipitation was key contributor to the changes in TWSA observed. The drought periods observed after hydrological year 2007 showed a steep decrease in the amplitudes of the plotted graphs This study also showed that GRACE and models Mismatches for example the monthly delays in phases between the model graph and the GRACE graph should be due to GRACE poor spatial resolution, but also to its monthly time resolution that is in the calculation of the anomalies. However, the good agreement in the amplitudes and phases of the GRACE and SWAT models data showed that it was viable to use the medium sized ENNB river basin and thus further studies can be made on similar basins.