ORIGINAL PAPER Evaluation of CMIP5 models for west and southwest Iran using TOPSIS-based method Reza Zamani 1 & Ronny Berndtsson 2 Received: 1 February 2018 /Accepted: 29 August 2018 /Published online: 6 September 2018 # Springer-Verlag GmbH Austria, part of Springer Nature 2018 Abstract GCMs (general circulation models) are main tools for generating climate projections for climate change research in hydrology and water resources. Accordingly, evaluating the performance of these models in simulating future climate is very important for choice of proper models. In this study, performance of 20 Coupled Model Intercomparison Project Phase 5 (CMIP5) model series was assessed using a technique for order performance by similarity to ideal solution (TOPSIS)- based approach together with normalized root mean square error (NRMSE), the Taylor skill score (S Taylor ), and two probability density function (PDF) skill scores. Precipitation and temperature data during 1976 to 2005 from three river basins including Zard River (ZR), Bakhtegan (BKH), and Ghareso (GH) in west and southwest Iran were used to select the best model. In general, models showed superiority in simulating temperature over precipitation. Based on the GCM ranking results for the ZR Basin, MIROC-ESM and IPSL-CM5A-LR were selected as the best and the weakest model, respectively. For the BKH Basin, the best model was BCC-CSM1.1 and the weakest IPSL-CM5A-MR and CCSM4. In other words, BCC-CSM1.1 had the maximum relative closeness to ideal solution. Based on the TOPSIS results, BCC-CSM1.1 and CanESM2 were the best models and IPSL-CM5A-MR the weakest model with a minimum relative closeness to the ideal solution in simulating temperature and precipitation for the GH basin. The approach presented in this study can be utilized to select appropriate climate models in other regions for future studies of climate change. 1 Introduction During the last decades, climate change has become a main issue in environmental and sustainable development studies. Investigation of adaptation solutions to mitigate future climate change impacts is of great importance. In studies of climate change, general circulation models (GCMs) are the main tools for future climate change projections in analyzing impacts of climate change in hydrology and water resources. These models represent physical processes for the three main water components of the global climate system (oceans, atmosphere, and cryosphere) as well as land surface. Although GCMs have been applied in climate change stud- ies for a long time, the incomplete knowledge regarding the complex dynamics of the climate system and the computation- al limits in simulation of small-scale physical processes are causing uncertainty in using the outputs. Therefore, character- izing and assessing this uncertainty and evaluating GCMs are important to provide more accurate information for re- searchers and decision-makers and have been addressed in many studies (e.g., Wójcik 2015; Choudhury et al. 2016; Ahmadalipour et al. 2017; Lee and Kim 2017). Perkins et al. (2007) evaluated the performance of 14 GCMs from the Fourth Assessment Report (AR4) of the Intergovernmental Panel on Climate Change (IPCC) for daily minimum and maximum temperatures and daily precipitation over Australia on the basis of the probability density function (PDF) skill score (SS). The results depicted higher skill of MIROC-M, CSIRO, and ECHO-G models with respect to the other models. Brown et al. (2011) examined the ability of 24 coupled model simulations for the twentieth century from AR4 of IPCC over the South Pacific Convergence Zone (SPCZ). They concluded that a large number of models are able to capture the major properties of SPCZ climatology * Reza Zamani rzamani.am@gmail.com 1 Water Resources Research Center, Shahrekord University, Shahrekord, Iran 2 Division of Water Resources Engineering & Center for Middle Eastern Studies, Lund University, Box 118, SE-221 00 Lund, Sweden Theoretical and Applied Climatology (2019) 137:533–543 https://doi.org/10.1007/s00704-018-2616-0