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.
