Development and testing of a 2D offshore oil spill modeling tool (OSMT) supported by an effective calibration method

Abstract

Abstract Spill trajectory prediction (STP) is essential to decision-making support for oil pollution preparedness and response. An Oil Spill Modeling Tool (OSMT) has been developed in this study to predict the transport and fate of oil spills from surface releases. Particularly, the Kullback-Leibler (KL) divergence method is adopted as a performance metric for the first time to formulate a calibration framework for STP from the Lagrangian transport model (LTM). By finding the candidate with minimal KL divergences from modeling scenarios using designed parameter combinations, the prediction discrepancy between simulated trajectories of the LTM and oil slicks detected from satellite images are reduced. The developed approach has first been evaluated through a numerical comparison analysis between OSMT and Operational Oil Modeling Environment (GNOME) model to examine model validity in a hypothetical case. Subsequently, a real case study is conducted to examine the applicability and effectiveness of the KL divergence-based oil spill trajectory modeling method. The study results indicate that OSMT can provide reliable spill trajectory simulations, and the KL divergence-based calibration method is effective in calibrating the oil spill LTM.