Evaluation of CMIP6 models in reproducing spatially averaged historical temperature and precipitation
Global and Planetary Change, cilt.268, 2027 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 268
- Basım Tarihi: 2027
- Doi Numarası: 10.1016/j.gloplacha.2026.105783
- Dergi Adı: Global and Planetary Change
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Artic & Antarctic Regions, Compendex, Environment Index, Geobase, Greenfile, INSPEC, Zoological Record, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO)
- Anahtar Kelimeler: CMIP6, ERA5, Global climate models, Köppen-Geiger climate zones, Model ranking, Multi-scale evaluation
- Orta Doğu Teknik Üniversitesi Adresli: Evet
Özet
Climate change impact assessments rely on the projections of Global Climate Models (GCMs). The reliability of their projections depends on how well they simulate the historical climate conditions. The most recent phase of the Coupled Model Intercomparison Project (CMIP6) provides a new generation of climate simulations. However, model performance changes with the variables simulated and the spatial domains. Thus, it is crucial to evaluate the performance before selecting them for climate impact assessment studies. In this study, 49 CMIP6 GCMs were evaluated against the ERA5 reanalysis dataset for spatially aggregated monthly averaged near-surface air temperature and monthly averaged daily total precipitation time series over 1940–2014. Model performance was assessed using the Correlation Coefficient (CC), Nash-Sutcliffe Efficiency (NSE), and Kling-Gupta Efficiency (KGE). These metrics were then integrated through a Comprehensive Ranking Index (CRI) framework to provide unified model rankings. The evaluation was performed at global, land-only, ocean-only, and Köppen-Geiger climate zone domains. The historical period was further divided into 1940–1979 and 1980–2014 to assess temporal robustness. Results show that temperature simulations generally exhibit higher skill and greater consistency than precipitation. Model rankings remain stable between the two periods. However, significant differences develop at Köppen-Geiger climate zones, particularly for precipitation, and no single model consistently outperforms others at all climatic regimes. The results also demonstrate that models performing well for global mean time series do not necessarily perform well for climate zone mean time series. The proposed CRI framework provides a systematic approach for evaluating spatially averaged climate time series at multiple domains. It also highlights the importance of considering spatial and climatic context when selecting GCMs for climate change impact studies.