TITLE:
Assessing the Skills of Rossby Centre Regional Climate Model in Simulating Observed Rainfall over Rwanda
AUTHORS:
Janet Umuhoza, Lin Chen, Lucia Mumo
KEYWORDS:
CORDEX, RCA4, Rainfall, Rwanda, Simulation Bias
JOURNAL NAME:
Atmospheric and Climate Sciences,
Vol.11 No.3,
May
31,
2021
ABSTRACT: Rainfall over Rwanda is highly variable both in
space and time. This variability leads to chronic food insecurity due to the
overdependence of the economy on rain-fed agriculture systems. This study aims
to evaluate the skills of Rossby Centre Regional Climate Model (RCA4) simulations
driven by 10 GCMs for the period 1951-2005 using the Global Precipitation
Climatology Centre (GPCC v8) as a reference. Different statistical and
geospatial metrics were used to deduce the model’s skills in simulating
seasonal and annual rainfall. Results show that the country received bimodal
rainfall pattern; March-May (MAM) and September-December (SOND). The RCA4
models are inconsistent in simulating the MAM rainy peak. However, the models
are coherent in simulating SOND seasonal peak despite exhibiting wet bias. The
models show reasonable skills in simulating mean annual cycle than interannual
variability as depicted by insignificant correlation and different signs of
rainfall trend. Conclusively, the performance of RCA4 models in simulating
observed rainfall characteristics over Rwanda is relatively weak. The
performance of the models differs at various time scales. Nevertheless, the
models can be ranked from the best performing to the least as; CSIRO, CanESM2,
CNRM, GFDL, MIROC5, ENS, EC-Earth, HadGEM2, IPSL, MPI, and NorESM1. Ranking the
performance of RCA4 historical models acts as a basis for future climate
model’s selection depending on the purpose of the study. The findings of this
study may help in devising appropriate climate adaptation measures to respond
to the ongoing global warming for sustainable economic and livelihood
development. Additionally, modelers may improve the model’s parametrization
schemes and lessen the inherent chronic biases for a better presentation of the
future.