Abstract
Oil spills in multi-island regions with complex coastlines pose severe threats to marine ecosystems. Traditional oil spill models typically use a static “infinite absorption” assumption for shoreline interactions, which does not represent finite shoreline retention capacity or subsequent shoreline release. To address this limitation, this study incorporates a dynamic adsorption-desorption physical mechanism based on substrate types into the Lagrangian particle tracking module of the SCHISM model. By introducing a substrate-specific maximum oil retention capacity, along with exponential desorption and permeation algorithms parameterized by constant, substrate-specific half-lives, the improved model represents the partitioning of stranded oil among surface-stranded, desorbed, and permeated compartments across four shoreline types: bedrock, gravelly, tidal flat, and artificial. Numerical experiments show that the improved model produces a rise-then-fall evolution of surface-stranded oil through the modeled competition among shoreline retention, desorption, and seepage, in contrast to the static mass partitioning produced by the original algorithm. Furthermore, shoreline substrate type is a key factor in determining oil partitioning: tidal flats exhibit extremely high surface retention, while gravelly shorelines cause substantial deep permeation. The improved model is then applied to the Zhoushan Archipelago, where a maximum-envelope methodology is used to extract the full domain cumulative oil concentration distribution across all simulation scenarios. Combined with an environmental sensitivity map, this yields a high-resolution oil spill risk zoning map that reveals the spatial reshaping of risk driven by hydrodynamic transport and ecological vulnerability. This study provides a more process-resolved representation of oil shoreline interactions in complex archipelagic waters and offers scenario based relative information for regional oil spill response planning and ecological protection prioritization.