COMPARISON OF DIFFERENT REANALYSIS METEOROLOGICAL DATA WITH GROUND OBSERVATIONS OVER NIGERIA

Abstract

In this study, a worldwide overview on the expected sensitivity of reanalysis choice is provided. Surface air temperature and precipitation data from ten (10) synoptic stations obtained from the Nigerian Meteorological Agency (NIMET) were compared with data from various datasets. This reanalysis is important for the development of both the dynamic and statistical schemes. The evaluation was based on a comparison of both temporal and spatial variability which includes several assessment criteria such as; root mean square error (RMSE), mean bias (MB), principal component analysis (PCA), correlation coefficient (r). The result showed that, the two reanalysis datasets, Global Precipitation Climatology Centre (GPCC) and University of Delaware (UDEL) dataset capture the temporal and spatial variability of the observed precipitation and mean air temperature (MAT) over Nigeria respectively. The result generally showed that the two reanalysis datasets performed better for precipitation and mean air temperature (MAT) when compared with GPCP, ERA-40, ERA-INTERIM, and UKMO. The two reanalysis dataset revealed the existence of a reasonable agreement with observed datasets according to the evaluation. UDEL was better at simulating the temporal and spatial distribution of temperature than other reanalysis dataset such as; ERA-40, ERA-INTERIM, UKMO and CRU of MAT, while GPCC was better at simulating the temporal and spatial distribution of precipitation than other reanalysis dataset, such as; GPCP, UDEL, CRU and UKMO. Of precipitation. The two reanalysis performed better in the Sahel zone of the study area then in the other zones by simulating spatial and temporal variables of precipitation and MAT. UDEL was poor at capturing the temporal variability of MAT. in the tropics and coastal zones of Nigeria. Based on time series trends analysis, the two reanalysis showed higher trend for precipitation and MAT than UKMO, GPCP, ERA-40, ERA-INTERIM and CRU. However UDEL and GPCC have larger explained variance for the first two PCAs of observed precipitation and observed MAT. This implies that both reanalysis dataset tend to simulate a more uniform spatial and temporal distribution of precipitation and MAT in the study area.