AN APPRAISAL OF THE ECMWF REANALYSIS5 (ERA5) MODEL IN ESTIMATING AND MONITORING ATMOSPHERIC WATER VAPOUR VARIABILITY OVER NIGERIA

Abstract

This study research the performance of the ERA5 reanalysis model in estimating and monitoring the variability of atmospheric water vapour content over Nigeria. The ERA5 is a fifth-generation reanalysis model recently released by the European Centre for Medium-Range Weather Forecasts (ECMWF). The ERA5 model comes with excitingly high spatial and temporal resolution when compared to earlier models like the ERA-Interim and ERA-40. However, like the previous models, the ERA5 comes with numerous modelling uncertainties arising from data fusion methods and observation schemes, which often affects its performance at the different regions of the Earth. In this study, ERA5 precipitable water vapour (PWV) was validated with GNSS PWV from permanent GNSS stations in Nigeria NIGNET for the period of 2012–2013. The performance of ERA5 was investigated at sub-daily, diurnal, and seasonal scales in relation to KöppenGeiger climate classification using standard statistical metrics (namely, coefficient of correlation (r), Root mean square error (RMSE), Reliability index (RI), Mean absolute errors (MAE) and mean bias). The r, RI, RMSE, MAE and mean bias values at sub-daily, diurnal and seasonal scales were computed as, (0.8670, 0.882, 0.979), (3.697 mm, 3.400 mm, 7.014 mm), (1.015, 1.019, 1.008), (2.769 mm, 2.706 mm, 1.939 mm) and (0.826 mm, 2.033 mm, 1.739 mm), respectively. The results indicate the strongest performance of ERA5 at seasonal scale with more than 95% agreement. The pattern of variability of ERA5 within the different climate zones of Nigeria showed good consistency with GNSS PWV and Köppen-Geiger climate classification. The study recommended the use of ERA5 in the retrieval of historic PWV records and near real-time GNSS applications.