Quantification of precipitation and temperature uncertainties simulated by CMIP3 and CMIP5 models

• Published in: Journal of geophysical research. Atmospheres 2016-01, Vol.121 (1), p.3-17
• Main Authors: Woldemeskel, F. M., Sharma, A., Sivakumar, B., Mehrotra, R.
• Format: Article
• Language: English
• Subjects: Assessments; Atmospheric precipitations; Climate; Climate change; Climate models; CMIP3; CMIP5; Coastal zone; Cold regions; Computer simulation; Datasets; Decomposition; Elevation; Emission; Environmental assessment; Environmental impact; GCM uncertainty; General circulation models; Geophysics; Global climate; Global climate models; Heavy rainfall; Identification; Intercomparison; Loads (forces); Mathematical models; Meteorology; Mountains; Natural variability; Precipitation; Projection; Radiative forcing; Rain; Rainfall; Resolution; Resources; Spatial discrimination; Spatial resolution; Temperature; Temperature effects; Uncertainty; Variability; Variance; Water; Water resources
• ISSN: 2169-897X; 2169-8996
• DOI: 10.1002/2015JD023719

Assessment of climate change impacts on water resources is extremely challenging, due to the inherent uncertainties in climate projections using global climate models (GCMs). Three main sources of uncertainties can be identified in GCMs, i.e., model structure, emission scenario, and natural variability. The recently released fifth phase of the Coupled Model Intercomparison Project (CMIP5) includes a number of advances relative to its predecessor (CMIP3), in terms of the spatial resolution of models, list of variables, and concept of specifying future radiative forcing, among others. The question, however, is do these modifications indeed reduce the uncertainty in the projected climate at global and/or regional scales? We address this question by quantifying and comparing uncertainty in precipitation and temperature from 6 CMIP3 and 13 CMIP5 models. Uncertainty is quantified using the square root of error variance, which specifies uncertainty as a function of time and space, and decomposes the total uncertainty into its three constituents. The results indicate a visible reduction in the uncertainty of CMIP5 precipitation relative to CMIP3 but no significant change for temperature. For precipitation, the GCM uncertainty is found to be larger in regions of the world that receive heavy rainfall, as well as mountainous and coastal areas. For temperature, however, uncertainty is larger in extratropical cold regions and lower elevation areas. Key Points GCM projection uncertainty is quantified for CMIP3 and CMIP5 data sets Uncertainty is visibly reduced in CMIP5 for precipitation but not for temperature Uncertainty is large in heavy rainfall regions, mountainous, and coastal grids for precipitation