Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Marine Science

Major Professor

Steve A. Murawski, Ph.D.

Committee Member

Ernst Peebles, Ph.D.

Committee Member

Claire Paris, Ph.D.

Committee Member

James Sanchirico, Ph.D.

Committee Member

David Naar, Ph.D.


environmental sensitivity indices, vulnerability assessment, marine spatial planning


This dissertation focused on oil spill risk assessment techniques historically utilized for coastal resources (e.g., environmental sensitivity indices (ESIs), vulnerability assessments, species-specific vulnerability frameworks) and their application and expansion to offshore resources found in the Gulf of Mexico (GoM). Additional included techniques (e.g., multi-sector trade-off analysis, marine spatial planning (MSP) software) provide support in decision making processes regarding the potential siting of oil production sites and/or the withdrawal of areas from oil and natural gas production. Chapter 1 included an overview and justification for this study. Chapter 2 demonstrated an initial methodology for the creation of offshore ESIs and how vulnerability to marine resources might be estimated via the inclusion of an oil fate and transport model used to simulate oil well blowouts. Chapter 3 created spatial distributions of marine resources from disparate sources and transformed the distributions into quantifiably comparable grids via the use of standardized indices. These grids were combined via a multi-attribute utility model (MAUM) to create multiple cumulative ESIs (C-ESIs) identifying resource rich “hot-spots” within the GoM. Chapter 3 also explored dissimilarity and cluster mapping methods to identify sets of resources with similar distributions and estimate the degree of influence to the overall C-ESIs. Species-specific vulnerability rankings estimated under a preliminary trait-based framework were provided for use in this study and were added to a fish species C-ESI to illustrate how C-ESIs might be combined with similar frameworks to weight one resource or suite of resources more heavily. The C-ESI was impacted as expected by the heavy weighting of one fish species with grid cells containing the weighted species being prioritized as resource “hot-spots”. The C-ESI was rather robust to the inclusion of weights for the suite of fish species owing to the conflicting spatial distributions of the individual weighted species. In Chapter 4, surface oil distributions from four oil well blowout scenarios were modeled with the Connectivity Modeling System (CMS; Paris et al. 2013). These oil distributions were created to simulate likely conditions from the Deepwater Horizon oil well blowout (DWH) and three hypothetical oil well blowouts with origin points in the western GoM (WGoM), the west Florida slope (WFS), and at the DWH origin point with a September start date. The simulated distributions of surface oil from these four scenarios were used to create bounded areas, or polygons, of minimum oil concentration thresholds (MOCT) representing spatial areas exposed to at least the specified concentration of oil for at least one day. These MOCTs are intersected with the C-ESIs developed in Chapter 3 to compare potential vulnerability of resources to the oil well blowout scenarios. The WFS scenario was found to potentially have the largest impact on the suite of included species due to the large surface area of the modeled spill and the resources found in that area. While a spill off the continental slope near Texas (WGoM) would have the smallest overall footprint, it would affect some fisheries more severely than the other simulated spills. The MOCT polygons coupled with known PAH concentration toxicity endpoints and spatial distributions of marine resources can be used to predict the extent of the spatial overlap between marine resources and oil and the likely outcome of that interaction. In Chapter 5, a spatial distribution of 2018 oil production was overlaid with the C-ESIs created in Chapter 3 to form quantifiably explicit multi-sector tradeoff curves between resource sensitivity and oil production. These tradeoff curves represent the system-wide benefit of theoretical allocations of oil production to both sectors on a cell-by-cell planning unit basis. The C-ESIs and multi-sector tradeoff functions were used to identify individual grid cells to potentially reserve from oil production. Marxan, an MSP zoning software, was used to create minimum-set “hot-spot” networks of areas to potentially reserve from oil production in both 1.) a hypothetical pristine system where future oil production siting is being planned and 2.) the current system with current levels and placement of oil production. In both scenarios, the “hot-spot” networks identified areas on the WFS as potentially valuable to reserve from oil production, with more “hot-spot” areas on the WFS when accounting for current oil production. This result indicates that in a hypothetical reserve sited around existing oil production and created to protect a significant proportion of offshore marine resources, the grid cells on the WFS would be the most valuable to include in that reserve. This study compares these “hot-spot” network solutions to the area in the eastern GoM currently withdrawn from oil production under Gulf of Mexico Energy Security Act (GOMESA) and the 2021 Congressional moratorium. The “hot-spot” networks overlap with the withdrawn area with 39.5% of the pristine “hot-spot” reserve network and 53.2% of the current oil production “hot-spot” reserve network being located within the withdrawn area. The “hot-spot” networks identified in this study can potentially be utilized in the decision-making process to continue the closure of these withdrawn areas.

The integrated collection of methods presented here were designed to add to the crucial knowledge base for planning and prioritizing oil spill response, predicting impacts from an oil spill to individual resources and groups of resources, and to assist in the decision-making processes for making new and existing sites available or unavailable to oil production. Impact estimates generated by these tools can be used in prioritizing oil spill cleanup and the acquisition or pre-positioning of oil spill response and supplies. The C-ESIs can be combined with actual or proposed oil production statistics to serve as a decision-making tool for justifying the reservation of specific lease blocks from leasing or maintaining the status of areas temporarily withdrawn from oil production under the Congressional moratorium.

The tools in this study were designed, developed, and published as open source with the hope that they may be added to, improved upon, and utilized in real world scenarios to lessen the risk of impacts on marine resources from future blowout scenarios. The Python scripts used for the analyses and figures found in this dissertation can be found at