Evaluation of GEOS-CF V1.0 against Surface Observation in Accra: Fine Particulate Matter Concentration

Abstract

Air pollution is one of the major environmental issues due to increasing anthropogenic activities and its negative impact on human health. Accurate forecasting is crucial in reducing exposure to poor air quality as well as mitigating against the effects of air pollution. For this reason, the performance of available models need to be evaluated to enhance the development of effective early warning systems. The Goddard Earth Observing System composition forecast (GEOS-CF) system is an operational product from NASA's Global Modeling and Assimilation Office (GMAO).The system provides near-real time forecast of atmospheric composites. This study evaluates the GEOS-CF representation of PM2.5 concentrations over Accra against the BAM-1020 ground station measurements. The performance of the model was described using index of agreement (IOA), coefficient of determination (R2) and fraction of prediction with a factor of 2 (FAC2) while the error was assessed with the normalized mean bias (NMB) and root mean square error (RMSE).The results showed good agreement between the GEOS-CF and surface observation with IOA, R2 and FAC2 recording 0.58, 0.52 and 0.81 respectively. The NMB and RMSE were found to be -0.03 and 15.82 respectively. The model skill improved after the application of the XGBoost regression model with an R2 of 0.65 and RMSE of 6.13. The end goal of this study is to apply a bias-correction factor to the model output for air quality forecasting over Ghana.