92nd American Meteorological Society Annual Meeting (January 22-26, 2012)

Tuesday, 24 January 2012: 12:00 AM
A Sensitivity Analysis of Hurricane Evacuation Costs in Florida
Room 243 (New Orleans Convention Center )
Michael Lowe, Pennsylvania State Univ., University Park, PA; and G. S. Young and J. L. Evans

Hurricane evacuations are an expensive endeavor; however, these expenses are compounded by significant costs owing to forecast error at the time that evacuation orders are announced. A war-game-style computer program is created to analyze evacuation costs given varying hurricane track forecast errors and evacuation order lead-times. A synthetic climatology of over 83,000 synthetic hurricanes impacting Florida is created based on the historical climatology of storms that struck Florida between 1900 and 2009. Each synthetic hurricane is created by modifying the track and intensity of a historical cyclone using a Monte Carlo method; 1,020 synthetic hurricanes are created for each observed landfall in the 110 year period. In this way, the synthetic climatology is assured of having behavior consistent with the historical record, but with much greater variability. Hurricane evacuation simulations are comprised of modules representing: (i) issuance of evacuation orders, (ii) a logistic model of cumulative population evacuation, and (iii) an “outcomes” module to calculate the evacuation costs for the total number of people who evacuated (including consideration of varying evacuation lead times), as well as casualty costs for the number of people who opted against evacuating from areas that were ultimately struck. Landfall forecast errors are simulated for Florida coastal counties given program-generated hurricane forecasts with error statistics modeled after those of the National Hurricane Center. Sensitivity analyses of evacuation cost responses to hurricane track forecast error and evacuation order lead times on a per-county basis will be presented and potential societal responses will be discussed.

Supplementary URL: