Proceedings of the International Conference on Climate Change, Vol. 1, 2017, pp. 36-50 Copyright © 2017 TIIKM ISSN 2513-258X online DOI: https://doi.org/10.17501/iccc.2017.1204 Corresponding Author Email: *met.herath@gmail.com EVALUATION OF DOWNSCALED CMIP5 CLIMATE MODELS TO SELECT THE BEST MODELS TO DEVELOP FUTURE CLIMATE SCENARIOS FOR SRI LANKA H.M.R.C. Herath 1* and I.M.S.P. Jayawardena 2 Department of Meteorology, Sri Lanka Abstract: Due to coarse resolution of global climate models, techniques are often needed to generate finer scale projections of variables affected by local-scale processes such as temperature and precipitation. Downscaling is widely used to improve spatial and/or temporal distributions of meteorological variables from regional and global climate models. This downscaling is important because climate models are spatially coarse (100– 200 km) and often misrepresent extremes in important meteorological variables, such as temperature and precipitation.NASA Earth Exchange Global Daily Downscaled Projections data with 25km grid spacing and Coordinated Regional Climate Downscaling Experiment (CORDEX) data with 50km grid spacing of 6 GCM models were used to evaluate models suitability for Sri Lanka. Annual mean precipitation and seasonal mean precipitation for four climatic seasons namely Southwest Monsoon, Northeast Monsoon, First Inter monsoon and Second Inter-Monsoon, of model historical runs (1975-2005) were compared with observed climatological average of precipitation to evaluate the models’ performance. CORDEX downscaled models were unable to capture the bi-modal pattern of annual cycle of precipitation in Sri Lanka and the spatial pattern of precipitation. NEX-GDDP downscaled models captured the bi-modal pattern of annual cycle of precipitation in Sri Lanka and the spatial pattern of precipitation of annual average as well as seasonal average.Based on the model performance of historical runs NASA Earth Exchange Global Daily Downscaled Projections of 6 GCM models (CanESM2, CNRM- CM5, CSIRO-MK3-6-0, GFDL-CM3, MRI-CGCM3and NCAR-CCSM4) with 25km grid spacing were selected to develop future projections for Sri Lanka. Keywords: downscaling, temperature, precipitation, seasonal, monsoon, annual, NEX- GDDP, CORDEX Introduction Introduction of general circulation models (GCMs) General circulation models (GCMs) are the primary tool for understanding how the global climate may change in the future but local scale precipitation processes are poorly represented due to the coarse resolution of the GCMs For climate change impact assessments finer scale information is required (Maraun et al, 2010).