Overview

The latest version of the HYbrid Coordinate Ocean Model (HYCOM) and an ensemble based data assimilative framework are used to estimate the ocean state variables necessary for oil spill risk analysis in the Gulf of Mexico. The system assimilates along-track sea level anomalies, gridded sea surface temperature, profiles of salinity and temperature and gridded climatology to provide data constraints to the ocean model.

An assessment of the data assimilative hindcast shows robust model skill in reproducing the various features of the circulation in the Gulf of Mexico. There is a good agreement between the hindcast and independent estimates of Loop Current behavior inferred from altimetry and ocean color both in terms of temporal and spatial characteristics. Furthermore, the assimilation of a limited number of in-situ and climatological profiles of temperature and salinity is effective in controlling bias and reduces errors in the vertical density structure in the model. The upper ocean temperature and salinity match observations to within 1.5°C and 0.3 psu both well below the climatological variance. Comparisons with independent drifter derived velocity data indicate significant correlation (>0.5) between the drifter data and model velocities. Simplified oil spill simulations of the Deep Water Horizon event captures the main features of surface plume indirectly verifying the underlying model circulation fields.

Overall, the hindcast dataset is consistent with what is observed during the 2003-2012 time frame and is can be considered as appropriate for a range of applications including oil spill risk analysis.