A County‐Level Risk Assessment of the Deep Water Horizon Oil Spill in the Gulf of Mexico

Abstract

This paper undertakes a risk assessment of coastal counties in the Gulf of Mexico impacted by the 2010 Deep Water Horizon oil spill. The study evaluates hazard risk from the perspective of community resilience, social capital, and access to resources. The proposed hazard risk location model re‐specifies risk as a function of hazard, exposure, and coping ability. The model employs an autoregressive function and a threshold analysis to develop a place‐based risk assessment. The results indicate that spatial variation in risk levels coincides with locational differences in social capital across the study area. Geographical proximity to the spill, population density, and unemployment rate are also key factors in determining overall risk. Furthermore, temporal variation in risk levels is determined by exposure to previous hazard events and changes in the business cycle.

Key words:

• coping ability
• Gulf of Mexico
• oil spill
• social capital
• vulnerability

Notes

1. Risk is articulated by Wisner and others (2004) in the pseudo‐equation, R = H x V, where H represents the event (hazard) and V represents vulnerability.

2. Location Quotient:Numerator: $\frac{No.ofWorkersinaSectorintheCounty}{TotalLaborForceintheCounty}$Denominator: $\frac{No.ofWorkersinaSectorintheStudyArea}{TotalLaborForceintheStudyArea}$

3. The regression equation with the lag operator is articulated as follows: $Y_t=B_0+{(B_{1}L{X}_1+B_{2}L{X}_2+.\dots \dots \dots B_{m}L{X}_m)}_j+e_t$where, Y_t is the value of the dependent variable (coping ability) in a given time period, t; B₀, the constant; and B₁, B₂…. representative of parameter estimates of the respective independent variables denoted by X₁ and X₂ in a set of m number of variables, k = 1…..m. The lag operator, L, represents the value of the independent variable in the previous time period (t – 1) in a set of j number of time periods, t = 1…..j. e_t is the error associated with estimating the dependent variable in time period, t.

4. The regression equation applicable to the HRLM builds on the preceding formula (Note 3) to include control variables and interaction terms: $Y_t=B_0+{(B_{1}L{X}_1+B_{2}L{X}_2+.\dots \dots \dots B_{m}L{X}_m)}_j+B_{T}T+B_{D}D+B_{Tx}Tx+e_t$The control variable for time is T, and its value is set at 1 for the event year and increments of one for subsequent years. T will be zero for years before the event. B_T, therefore, is the parameter estimate of time after the event. D is the control variable for geographical proximity from the event. It is calculated by dividing the actual distance of each county from the spill by 1 + T. B_D is the parameter estimate of the distance‐decay variable. T_x is an interaction term that evaluates the parameter for each independent variable (x) before and after the event year, T. B_Tx is the parameter estimate of the interaction term, Tx. e_t is the error associated with estimating the dependent variable in time period, t.

5. The threshold formula: $\frac{x_{\left(j\right)}}{TH}-1$where, x_(j) is the observed or predicted value of the variable in question for county j; and TH, the threshold value of the variable in question.

6. Weighted Average (Ratick and Osleeb 2011): $I_j=\sum _{i\in A}{W}_i{M}_{ij}\forall j\in J$where, Ij is the composite weighted average for the risk index for spatial unit, j; Wi, is the weight associated with attribute i; and Mij is the attribute value i applicable to spatial unit j. A is the total number of attributes that contribute to risk and J is the set of spatial units in the study area.

Additional information

Notes on contributors

Naomi W. Lazarus

Dr. N. W. Lazarus Department of Geography, Binghamton University, New York; [[email protected]].