Comparative analysis of satellite and reanalysis data with ground‐based observations in Northern Ghana
Josephine Thywill Katsekpor,Klaus Greve,Edmund Ilimoan Yamba,Ebenezer Gyampoh Amoah
DOI: https://doi.org/10.1002/met.2226
2024-08-21
Meteorological Applications
Abstract:CHIRPS and PERSIANN‐CDR outperform ERA5, TRMM and ARC2 in simulating rainfall at Wa, Tamale, Walewale, Bawku and Navrongo (GMet) gauge stations. For mean temperature, ERA5 performs better than CFSR and MERRA‐2 when compared with the gauge data. ERA5 and CSFR outperform MERRA‐2 in simulating soil moisture in Bimbilla, Navrongo, Tamale, Pusiga and Bongo ISMN stations. Accurate predictions of streamflow and flood events are contingent upon the availability of reliable hydrometeorological data. In regions characterized by scarcity of ground‐based hydrometeorological observations, satellite and reanalysis data assume prominence as alternative predictors. Floods and droughts have emerged as a significant concern in Northern Ghana, yet the scarcity of ground‐based hydrometeorological data impedes effective prediction of these hydrological events. Consequently, the identification of suitable surrogate hydrometeorological data holds paramount importance in addressing these challenges. This study, therefore, assessed the accuracy of satellite and reanalysis data against ground‐based data in Northern Ghana. Rainfall and mean temperature spanning from 1998 to 2019 and soil moisture datasets from 2019 to 2022 were collected from GMet, ISMN (ground‐based), CHIRPS, PERSIANN‐CDR, ERA5, ARC2, MERRA‐2, TRMM and CFSR (satellite and reanalysis). Employing rigorous statistical measures, namely standard deviation, mean absolute error (MAE) and mean bias error (MBE), the accuracy of these datasets was thoroughly evaluated. The results revealed that CHIRPS and PERSIANN‐CDR exhibited superior accuracy in rainfall simulation, with CHIRPS demonstrating particularly consistent congruence with observed data. In terms of mean temperature prediction, ERA5 surpassed MERRA‐2 and CFSR. Regarding soil moisture assessments, both ERA5 and CFSR offered satisfactory simulations. Hence, our findings advocate for the preference of CHIRPS (for rainfall data), ERA5 (for temperature data) and a combination of CFSR/ERA5 (for soil moisture data) as dependable primary data sources for streamflow modelling, drought analysis, flood prediction and water resource management in the context of Northern Ghana.
meteorology & atmospheric sciences