Analysis of typhoon intensity and extreme wind speeds considering thermodynamic mechanisms

Abstract:

To adjust existing deterministic typhoon intensity models and account for stochastic influences, a correction term containing both mean and random noise is introduced into the ordinary differential equations governing deterministic typhoon intensity. Various skewed distribution models serve as candidate probability distributions for the random noise. Using historical typhoon data from the Northwest Pacific, the geographically weighted method estimates geographic variation in mean, standard deviation, skewness, and excess kurtosis of the correction term. Furthermore, the method of moments estimates parameters for candidate probability distribution models of the random noise within the correction term, with the optimal probability distribution model determined by KS distance. By comparing simulated results of historical typhoon intensity evolution, the impact of both mean and stochastic components of the error term on model performance is examined. The results indicate that introducing the correction term significantly improves the model′s ability to simulate historical typhoon intensities and enhances its capacity to capture the stochastic nature of typhoon intensity. Additionally, this paper validates model effectiveness in extreme wind speed analysis of typhoons.