Development of a dynamical statistical analog ensemble forecast model for landfalling typhoon disasters

Area of Science:

• Meteorology and Atmospheric Science
• Disaster Risk Reduction
• Geospatial Analysis

Background:

• Landfalling typhoon disasters (LTDs) pose significant threats to coastal communities.
• Accurate forecasting of typhoon impacts is crucial for effective disaster mitigation.
• Existing models may require enhancement for precise prediction of typhoon-induced damage.

Purpose of the Study:

• To develop and present a novel Dynamical Statistical Analog Ensemble Forecast model for LTDs.
• To apply and assess the model's performance in coastal China.
• To improve the prediction of typhoon disaster characteristics.

Main Methods:

• The model integrates four key elements: forecast track acquisition, generalized initial value (GIV) construction, analog identification, and analog disaster assembly.
• The GIV incorporates typhoon track, intensity, and landfall date.
• Ensemble forecasting techniques are employed to generate disaster predictions.

Main Results:

• Pre-assessment indicates mean threat scores of 0.48 and 0.55 for two critical LTD damage levels.
• The model effectively captures that most damage occurs near typhoon centers at landfall.
• The developed model demonstrates promising performance in predicting key typhoon disaster features.

Conclusions:

• The Dynamical Statistical Analog Ensemble Forecast model shows potential for enhancing LTD prediction.
• The model's ability to identify damage characteristics aids in disaster mitigation strategies.
• This approach offers a valuable tool for protecting coastal communities from destructive typhoons.