Answers are grounded in this paper and cite their sources.
Introduction Oil spills in the marginal seas surrounding the Korean Peninsula pose major ecological and socioeconomic threats, yet proactive spatial risk assessment remains limited because spill occurrence, transport-mediated exposure, and receptor sensitivity are often conflated. Methods We developed a high-resolution probabilistic Oil Spill Risk Assessment (OSRA) framework that explicitly separates spill likelihood from modeled oil exposure probability and integrates exposure with environmental and socioeconomic receptor sensitivity. Spill likelihood was mapped on a 300 m grid using 20 years of spill records, AIS-derived vessel traffic and collision likelihood, oil transport volume, and coastal oil storage capacity. A total of 310 representative spill locations were identified and 31,000 oil spill fate and trajectory simulations were conducted using 100 seasonally stratified release times. Results The resulting simulations were used to estimate oil exposure probability, which was subsequently combined with receptor sensitivity to generate an integrated spatial risk map. The Yellow Sea exhibited the highest oil exposure probabilities and overall risk, particularly in shallow semi-enclosed coastal areas where tidal flats, wetlands, and fisheries coincide with persistent oil retention. The South Sea showed moderate risk with localized hotspots near major industrial and port areas, whereas the East Sea exhibited generally low risk because receptor sensitivity was lower and open-coast circulation promoted rapid dilution and offshore advection. Discussion By distinguishing spill likelihood, transport-mediated exposure, and receptor sensitivity, the framework explains why areas with frequent spill sources do not necessarily correspond to the highest shoreline risk. The proposed OSRA framework provides an operationally relevant basis for prioritizing prevention, preparedness, and response planning and is adaptable to other heavily trafficked marginal seas under comparable data conditions.
No takes yet. Share an insight, caveat, or question.
Jeong et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: