Andrea Schumacher, CIRA/CSU Mark DeMaria and John Knaff, NOAA/NESDIS/StAR

 Generally accepted that improvements to hurricane forecasts will benefit society  Longer lead times  more time to prepare  Better track forecasts  reduce areas warned and/or evacuated unnecessarily  However, quantifying these benefits a difficult task  How much money will a better forecast save?  How many lives could be saved? 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 Two main steps 1) Develop an objective scheme that simulates official hurricane warnings based on real-time hurricane track and intensity forecasts ▪Use Monte Carlo Wind Speed Probability model 2) Make artificial “improvements” to forecasts and use warning scheme to diagnose changes in warning properties 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 Operational at NHC since 2006 (replaced Strike Probability Program)  Methodology  Samples errors from NHC track and intensity forecasts and Wind radii of realizations from radii CLIPER model over last 5 years  Generates 1,000 realizations  Calculates probabilities over domain from realizations  Versions for Atlantic, NE and NW Pacific  Current products  Cumulative and incremental probabilities  34, 50 and 64 kt winds  0, 12, …, 120 hr  Text and graphical products  Distributed via NHC web page, NDFD, AWIPS 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

1000 Track Realizations 64 kt h Cumulative Probabilities Major Hurricane Non-major Hurricane Tropical Storm Depression

63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 Rerun MC probability model  Used 64-kt (hurricane force) wind criteria  Used 36-h cumulative probabilities (best match for NHC hurricane warning criteria of 24 h)  Sample set: all U.S. mainland hurricane warnings from (20 tropical cyclones)  Used NHC official breakpoints + extras (343 breakpoints total, from Mexico to Canada)  Choose wind speed probability thresholds  p up – minimum for putting warning up  p down – maximum for taking warning down 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

NHC Hurricane Warnings Objective Scheme Hurr Warnings

First Guess (Prelim w/ Ivan) : p up = 10.0%, p down = 2.0% Best fit: p up = 8.0%, p down = 0.0% 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009 MCPNHC Average Distance Warned (mi) Average Distance Overwarned (mi) Average Distance Underwarned (mi) Average Warning Duration (hr) MCP Objective vs. NHC MAE, Distance (mi)65 MAE, Duration (hr)5 R 2, Distance0.94 R 2, Duration0.74

63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 Two steps needed  Use best tracks from ATCF to adjust tracks and intensities closer to observed values  Scale the sampled track (intensity) errors in the Monte Carlo scheme  For this study, 20% and 50% error reductions were used  Apply aforementioned hurricane warning scheme to MC wind speed probabilities based on “improved” forecasts 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

Average = mi Average = 33.6 mi

63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 An automated, objective hurricane warning scheme has been developed from NHC wind speed probabilities  Scheme issues hurricane warnings when p>8% and lowers warnings when p=0%  Scheme well correlated with official NHC hurricane warnings from  Testing suggests relationship between reduction of forecast errors and reduction of warning length/duration is not 1 to 1  20% Forecast Improvement (both track & intensity) yields ▪29 mile (5%) reduction in coastal length of warning ▪2 hour (8%) reduction in warning duration (i.e., dropped earlier)  50% Forecast Improvement (both track and intensity) yields ▪91 mile (13%) reduction in coastal length of warning ▪5 hour (24%) reduction in warning duration (i.e., dropped earlier) 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

 Quantify socioeconomic benefits resulting from  Reduction of coastal distance over-warned  Dropping warning sooner after threat passed  Integrate social science research  $600,000 - $1 million per mile estimate ▪Too generic, doesn’t account for population density differences ▪Whitehead 2003 suggests might actually be less  Some emergency management guidance products estimate costs of evacuation decisions ▪Emergency Management Decision Support System (EMDSS, Lindell and Prater 2007)  Goal: Get best, most accurate estimates possible given current knowledge 63rd Interdepartmental Hurricane Conference, 2-5 Mar 2009

20% Track and Intensity Forecast Improvement 50% Track and Intensity Forecast Improvement Reduced warning Areas (blue) Reduced warning Areas (blue)

 DeMaria, M., J. A. Knaff, R. Knabb, C. Lauer, C. R. Sampson, R. T. DeMaria, 2009: A New Method for Estimating Tropical Cyclone Wind Speed Probabilities. Wea. Forecasting, Submitted.  Jarell, J.D. and M. DeMaria, An Examination of Strategies to Reduce the Size of Hurricane Warning Areas. 23 rd Conference on Hurricanes and Tropical Meteorology, Dallas, TX, Janurary  Lindell, M.K. and C.S. Prater, 2007: A hurricane evacuation management decision support system (EMDSS). Natural Hazards, 40,  Whitehead, J.C., 2003: One million dollars per mile? The opportunity costs of Hurricane evacuation. Ocean and Coastal Management, 46, rd Interdepartmental Hurricane Conference, 2-5 Mar 2009