Heavy Rainfall Forecast of Landfalling Tropical Cyclone Over China With an Upgraded DSAEF_LTP Model.

Autor: Wang, Mingyang1,2 (AUTHOR), Ding, Chenchen3 (AUTHOR), Ren, Fumin2 (AUTHOR) fmren@163.com, Zhang, Da‐Lin2,4 (AUTHOR), Jia, Li1,2 (AUTHOR)
Předmět:
Zdroj: Journal of Geophysical Research. Atmospheres. 11/16/2023, Vol. 128 Issue 21, p1-18. 18p.
Abstrakt: Since the release of the Dynamical‐Statistical‐Analog Ensemble Forecast for Landfalling Typhoon Precipitation (DSAEF_LTP) model, a series of upgrades has been made to improve its heavy rainfall forecast performance for different cases and regions of China. However, little effort has been given to systematic evaluation of its multi‐upgraded version with a large sample size, especially for all the coastal regions of China. This study evaluates the performance of Version 1.0 (V1.0) and its improved Version 1.1 (V1.1) of the DSAEF_LTP model in simulating heavy rainfall amounts associated with 76 landfalling tropical cyclones (LTCs) over China during 2004–2016. The optimized schemes from these simulations are then applied to the heavy rainfall forecasts of 37 LTCs during 2017–2020. To further improve the forecast performance of V1.1, a subregional re‐ensemble scheme, referred to as V1.1R, is introduced after examining its performances over three coastal subregions of China. A comparison of the forecast performances of V1.0, V1.1, V1.1R, and four operational numerical weather prediction (NWP) models for the 37 LTCs shows that the performance of V1.1 is much better than that of V1.0, with a 65% higher summed threat score (TS) for over 250‐mm (TS250) and 100‐mm (TS100) rainfall; and that V1.1R's TS100 and TS250 values are further improved by 8% and 20%, respectively, as compared to V1.1. These values are superior to those of the four operational NWP models. Additionally, case studies reveal that the track distributions and intensities of LTCs may significantly impact the forecast performance of heavy rainfall. Plain Language Summary: The Dynamical‐Statistical‐Analog Ensemble Forecast for Landfalling Typhoon Precipitation (DSAEF_LTP) is used to predict heavy rainfall from landfalling tropical cyclones (LTCs). A series of model upgrades has been made to improve its heavy rainfall forecast performance for different cases and coastal regions in China. In this study, forecast experiments are conducted during the 17 years of 2004–2020 with a sample size of 113 LTCs to systematically evaluate the heavy rainfall forecast ability of the first released version (V1.0), its multi‐upgraded version (V1.1), and further upgraded version (V1.1R), in which a subregional re‐ensemble scheme is incorporated after examining its different forecast performances over three subregions of China. Results show that (a) V1.1 not only enhances the applicability of the model but also improves the performance of heavy rainfall forecasts; (b) the performance of V1.1R in predicting heavy rainfall over the coastal regions of China is superior to that of four mainstream numerical weather prediction models; and (c) case studies show that the track distribution and intensities of LTCs may significantly influence the forecast performance of heavy rainfall. Key Points: Revised track similarity regions and ensemble methods improve the forecast performance of heavy rainfall by the Dynamical‐Statistical‐Analog Ensemble Forecast for Landfalling Typhoon Precipitation (DSAEF_LTP) modelThe performance of the upgraded DSAEF_LTP model in heavy rainfall forecast excels that of four mainstream operational global modelsThe track distribution and intensity of tropical cyclones may significantly impact the distribution of their heavy rainfall after landfall [ABSTRACT FROM AUTHOR]
Databáze: GreenFILE
Externí odkaz: