Evaluation of ocean wave power utilizing COWCLIP 2.0 datasets: a CMIP5 model assessment

Global Climate Models (GCMs) are very essential and crucial for projecting future climate scenarios under different greenhouse gas emissions, incorporating uncertainties in the global warming projections. The present study evaluates the seasonal performance of 32 Coupled Model Intercomparison Project Phase 5 (CMIP5) models obtained from the Coordinated Ocean Wave Climate Project phase 2 (COWCLIP 2.0) in simulating the global and regional wave power (WP) from 1979 to 2004 using historical data, and comparing them against the ERA5 reanalysis. Three skill metrics, such as Root Mean Square Error (RMSE), Interannual Variability Skill (IVS), and M-Score were used to assess the model performance across three clusters (CSIRO, JRC, and IHC). In addition, Intra-seasonal and probability distribution is also employed to determine the cluster's performance, including individual models. The IHC cluster, employing statistical techniques, exhibited the lowest RMSE and highest M-Score values with the least variation among models over the global as well as regional ocean basins such as the North Atlantic (NA), North Pacific (NP), Indian Ocean (IO), and Pacific Ocean (PO. Results from intra-seasonal variability and probability distribution indicate that the IHC cluster demonstrates the most stable performance in simulating intra-seasonal variability of WP as compared to other clusters.