Year ahead prediction of US landfalling hurricane numbers: the optimal combination of long and short baselines

TL;DR

This paper proposes an optimal method combining long-term and recent hurricane data to improve the accuracy of predicting next year's US landfalling hurricane numbers.

Contribution

It introduces a new approach that optimally combines historical and recent data to minimize mean squared error in hurricane prediction.

Findings

The combined model outperforms simple averages.

Recent data significantly influences short-term predictions.

Optimal weighting improves forecast accuracy.

Abstract

Annual levels of US landfalling hurricane activity averaged over the last 11 years (1995-2005) are higher than those averaged over the previous 95 years (1900-1994). How, then, should we best predict hurricane activity rates for next year? Based on the assumption that the higher rates will continue we use an optimal combination of averages over the long and short time-periods to produce a prediction that minimises MSE.