CMIP5 model performance of significant wave heights over the Indian Ocean using COWCLIP datasets

Abstract

Wind-generated surface gravity waves forms an integral part in modulating the air-sea exchange processes. Information of wave parameters is also very essential in planning marine- and coastal-related activities. It is now well recognized that wind-wave activity shows changing trends over the global ocean basins. Numerous studies have addressed the projected changes in significant wave height for the Indian Ocean (IO) region, and there is a need to conduct thorough performance evaluation of global climate models (GCMs) over this region for futuristic planning. With this motivation, the present study examined the performance of historical dynamical wave climate simulations generated under the Coordinated Ocean Wave Climate Projections (COWCLIP) experiment. The simulations utilized near-surface wind speed datasets from 8 CMIP5 (Fifth phase of Coupled Model Intercomparison Project) GCMs to force a spectral wave model. The skill level of individual GCM forced wave simulations and multi-model mean (MMM) in reproducing the significant wave height (SWH) over four different sub-domains in the IO was evaluated with reference to the ECMWF Reanalysis 5th Generation (ERA5) datasets. Several performance metrics such as the Taylor Skill, M-Score, Model Climate Performance Index (MCPI), and Model Variability Index (MVI) are employed to establish the skill level of model simulations. The study deciphers that model performance is highly reliant on the region and its characteristics. Representation of the historical wave climate over the Arabian Sea (AS) and the Bay of Bengal (BoB) regions is remarkable in the COWCLIP datasets. However, there are discrepancies noticed in SWH distribution over the South Indian Ocean (SIO) attributed to model limitations in adequately reproducing swell wave fields over that region. The MMM constructed using the best-performing models (MRI-CGCM3, ACCESS1.0, INMCM4, HadGEM2-ES, and BCC-CSM1.1) is found consistent at all the sub-domains. The study signifies that the performance evaluation of GCM forced wave simulations is crucial before employing them for practical applications. Best-performing models listed from this study can be used to establish futuristic scenarios of SWH in a changing climate for the IO region.