CCU Researchers Create Their First Hurricane Outlook for 2013. The Only Major Model to Include Landfall Outlooks.

Friday, June 7, 2013

In April 2013, Dr. Tingzhuang Yan, Dr. Len Pietrafesa, and Dr. Paul T. Gayes completed efforts to develop a new modeling technique that would allow the Burroughs & Chapin Center for Marine and Wetland Studies at Coastal Carolina University to begin forecasting storm probabilities during hurricane seasons.  What makes this forecast unique is that in addition to predicting the number of storms in any given season the outlook also includes the probability of hurricanes making landfall along the US Atlantic and Gulf Coasts. 

The outlook includes common predictions, similar to those created annually by well-known atmospheric research organizations and institutions (e.g. NOAA, Colorado State University), but is the only model to attempt predicting landfalls.  Many of the model predictions will be familiar to most people who live along the coasts and includes the number of named storms (TS), number of hurricanes (NH), and number of major hurricanes (MH, category 3 or higher).  It also includes the number of landfall hurricanes along the Atlantic seaboard (ECLF) and the number of landfall hurricanes along the US Gulf Coast (GMLF).  The number of landfalls is given as a probability in order of decreasing likelihood.  For both the East Coast and Gulf Coast of the US, the most likely scenario for 2013 is 1 hurricane landfall, with 0 landfalls being less likely and 2 landfalls or greater being very unlikely.  The model, once it calculated these probabilities, determined the number of landfalls to be a “fraction of a storm” which simply isn’t practical (obviously you can’t have 0.96 storms making landfall) so the researchers have chosen to display the data as the most likely to least likely number of hurricane landfalls 1, 0, 2. 

The science of creating these outlooks is based heavily on statistics from data that is available in the historic storm records.  Reliable information on the number of storms, hurricanes, landfalls and physical conditions that were present during storm formation only dates back into the 50s.  Sixty years of data, while not ideal, is the best the researchers have to work with.  Using this information, what is produced are tables that indicate the probability of having 1 named storm, 2 named storms, 3 named storms etc… and that information is then summarized for official reporting which is why the outlooks are typically displayed as ranges or probabilities.

A comparison of the new Burroughs & Chapin Center for Marine and Wetland Studies outlook versus other organizations that create similar results can be seen below.

Table 1: Comparison of predictions for the 2013 hurricane season in the Atlantic and Gulf of Mexico
Agency Affiliation Named Storms Hurricanes Major Hurricanes East Coast Landfall Gulf Coast Landfall
CCU-BCCMWS Coastal Carolina University 13 to 15 6 to 8 3 to 4 1,0,2 1,0,2
TSR University College London 15 8 3 n/a n/a
WSI/TWC Weather Services International 16 9 5 n/a n/a
CSU Colorado State University 18 9 4 n/a n/a
NCSU North Carolina State University 13 to 17 7 to 10 3 to 6 n/a n/a
UKMO British Meteorological Office 14 8 n/a n/a n/a
NOAA US Federal Agency 13 to 20 7 to 11 3 to 6 n/a n/a
FSU COAPS Florida State University 18 9 4 n/a n/a