Seasonal Prediction Models for North Atlantic Basin Hurricane Location

Abstract

Using multivariate discriminant analysis techniques, statistically significant and skillful models are developed for making extended-range forecasts of hurricane activity within specific locations of the North Atlantic basin. These forecasts predict the presence or absence of hurricane activity and not the actual number of storms that will occur within a region. Successful models are developed for predicting intense hurricane activity in both the Gulf of Mexico and the Caribbean subbasins separately. Extended-range forecasts of all hurricane activity are also possible within the Caribbean Sea. More significantly, lead-time forecasts of landfalling hurricanes on the southeastern Atlantic coast of the United States are possible and show a substantial improvement over climatology. Extended-range forecasts of hurricane activity for the northeastern United States and for the Gulf of Mexico are not feasible due, respectively, to the relative lack and abundance of hurricane activity. Crossvalidated forecast accuracies range from 78% to 81% for the regions in which successful models can be developed. An all-possible subsets selection algorithm is used to identify the predictor models, while bootstrap techniques are used to assess model significance. Statistical tests using normal approximations are employed to compare cross-validated (hindcast) forecast accuracy to climatology.

Current approaches to seasonal to interannual climate predictions

“…Most models for the Indian monsoon use multiple linear regression, and this method, together with CCA, has been adopted extensively in Africa also. Auto-regressive approaches have received attention recently (Elfandy et al, 1994;Chu et al, 1995;Mentz et al, 2000), as have probabilistic methods, most notably discriminant analysis Casey, 1995;Carter and Elsner, 1997;Lehmiller et al, 1997;Mason, 1998;Mattes and Mason, 1998;Mutai et al, 1998;Mason and Mimmack, 2001).…”

“…Notable exceptions include attempts to forecast frequencies and tracks of tropical cyclones. The main focus has been on North Atlantic hurricanes (Gray et al, 1992(Gray et al, , 1994Hess et al, 1995;Lehmiller et al, 1997;Mielke and Berry, 2000), but some attention has been given to the northwest Pacific (Chan et al, 1998), the southwest Indian Ocean (Jury et al, 1999b), and the southwest Pacific (Basher and Zheng, 1995). Looking beyond forecasts of seasonal conditions, very few efforts have been made to forecast climate at interannual time scales (White, 2000).…”

“…More sophisticated statistical methods of defining forecast probabilities have been implemented Casey, 1995;Carter and Elsner, 1997;Lehmiller et al, 1997;Mason, 1998;Mattes and Mason, 1998;Mutai et al, 1998;Mason and Mimmack, 2001), but generally, maintain the assumption of climate stationarity. Regression-based forecasting approaches can provide confidence intervals (Wilks, 1995), but these, while being indications of average model uncertainty, are poorly related to the inherent unpredictability of a particular forecast period.…”

Forecasting US insured hurricane losses

“…Studies focusing on climate factors that influence hurricane frequency regionally (Lehmiller et al, 1997;Bove et al, 1998;Maloney and Hartmann, 2000;Elsner et al, 2000a;Murnane et al, 2000;Saunders et al, 2000;Jagger et al, 2001;Larson et al, 2005) are more recent. Insights into climate conditions that affect regional hurricane activity are used to help predict landfall activity (Lehmiller et al, 1997;Elsner andJagger, 2004, 2006;Saunders and Lea, 2005). Preseason forecasts of the number of hurricanes expected to affect the coast are useful especially if they are issued with significant lead time.…”

ENSO influence on tropical cyclone regional landfall counts

“…Over the past few decades, people have increasingly relocated to, and have built property in, coastal regions (known as "coastal migration"), which has increased the potential impacts of these storms (Pielke and Landsea 1998). Because of the extensive losses associated with these storms, considerable research has been performed over the past few decades to develop methods by which to accurately predict these storms on a broad range of time scales (e.g., Gray et al 1993Gray et al , 1994Lehmiller et al 1997;Klotzbach 2014). Of interest here is the potential to predict the severity of a hurricane season-often quantified using the number of named storms-months in advance.…”