ASSESSMENTS OF SUB-SEASONAL TO SEASONAL WEATHER FORECASTS AND APPLICATIONS TO PREDICTING MALARIA INCIDENCES IN NIGERIA

Abstract

The increasing needs for reliable weather forecasts for enhanced living, protection of property and socio-economic sustainability in Nigeria motivated this study. This study assesses and evaluates the skills of China Meteorological Administration (CMA), European Center for Medium-range Weather Forecasting (ECMWF), and UK-Met-Office (UKMO) Sub-Seasonal-to - Seasonal (S2S) models in the ability of the S2S models to forecast monsoon rainfall dynamics with emphasis on the peak of the monsoon season this includes the skill of the S2S model at forecasting rainfall anomalies distribution on an inter-annual time-scales during the peak of the monsoon and the ability of the S2S model to reproduce global drivers modulating the monsoon and it variability. It also investigates the S2S models skill in predicting monsoon onset, its variability and teleconnections modulating global drivers with rainfall onset anomaly. Because of the need to develop a robust malaria early warning system(MEWS) in a time-scale just right for effective action, this study also, uses two hierarchical evaluation technique to investigate the skill of VECTRI model using the simulated Entomological Inoculation Rate (EIR) and evaluate the skill of three accessed S2S models in driving the VECTRI model. Here, the simulated EIR from observed driven VECTRI is evaluated with the simulated EIR from the three S2S models used. All the models, their ensemble members, and the observations were subjected to quantitative statistical analyses from 1998 to 2017. Results show that all the three models reproduced the evolution and variability of the global drivers modulating the monsoon during its peak of the monsoon and monsoon onset with disparities. Furthermore, Results also show that the three models are able to simulate the Northwards migration of the onset dates adequately with biases and unique characteristics. They captured the vii evolution and variability of the global drivers modulating the monsoon onset. While CMA and the ECMWF models improve towards the Sahel. The VECTRI model, driven with observed station rainfall and temperature data can simulate the hyper-endemic characteristics of malaria occurrence in Nigeria. It suggests that simulated Entomological Inoculation Rate (EIR) could be used to interpolate reported cases of malaria over Nigeria. Again, all the three S2S models used in driving the VECTRI-Model reproduced the EIR that signifies the hyper-endemic nature of malaria in Nigeria, with different characteristics over different regions. In conclusion, the results suggested that out of the three S2S models used the CMA model has the least skilled. However, all the S2S models, despite the inherent biases can be a good tool for forecasting monthly rainfall amounts in Nigeria, particularly in areas from latitude 10◦ north-ward. They can also predict rainfall onset over Nigeria, within the sub-seasonal time scale and predict hyper-endemic malaria occurrences in Nigeria , within the sub-seasonal time scale. Also, the reproducibility of the atmospheric dynamics by the S2S model may not transform into skillful prediction of the anomaly of rainfall and the anomaly of rainfall onset in Nigeria. Finally, the improvements in multi-model ensembles could be a value-added information, able to significantly improve model performance.