Comparative Analysis of Groundwater Recharge Simulated Using Historical Observed and Projected Atmospheric Forcing Data

Abstract

This study focuses on a comparative analysis of groundwater recharge over Africa using atmospheric data from the reanalysis dataset originating from GSWP3 project and projections dataset originating from the CORDEX covering the common period from 2006 to 2014. The aim is to understand the reasons for the discrepancies between the recharge estimates obtained from these two types of datasets based on the output of the CLM5 and examine their implications for water resource assessment. The first step in the analysis was to estimate groundwater recharge using reanalysis data and projected data separately. This comparison revealed significant differences between the two sources. In order to better understand the origin of these differences, a study of the components of the water balance was conducted. This showed that precipitation and evapotranspiration are the main determinants of groundwater recharge. The differences observed between the results are therefore largely due to differences in these two hydrological components between the datasets considered. The variability in precipitation can be explained by the intrinsic nature of the data, as it was directly incorporated into the recharge calculation without first being used by CLM5. This characteristic contributes to accentuating the differences between reanalyses and projections. Furthermore, examination of the meteorological variables used as model inputs revealed significant differences between the data from reanalyses and those from climate projections. These discrepancies raise questions about the reliability of reanalysis data and highlight the high degree of uncertainty associated with them. A further analysis of the characteristics of the two datasets also showed that they differ in terms of spatial and temporal resolution. As the model was run without harmonising these resolutions, this methodological difference is likely to be an additional factor explaining the extent of the discrepancies observed. From this study, further recommendations are observed. Firstly, a detailed verification process and validation of the weather atmospheric dataset, to further continue the investigation related to meteorological variables influencing the evapotranspiration, and secondly, to analyse the impact of using the same resolution datasets for the input of the CLM5.