An Image Recognition Framework for Detecting Large-Scale Persistent Extreme Precipitation Events (LPEPEs) and Its Revelation of Southward Displacement and Dynamic Mechanisms.

Saved in:

Bibliographic Details
Title:	An Image Recognition Framework for Detecting Large-Scale Persistent Extreme Precipitation Events (LPEPEs) and Its Revelation of Southward Displacement and Dynamic Mechanisms.
Authors:	MA, JIAO¹, WEI, KE² weike@mail.iap.ac.cn, ROUQI, LONG³, ZHANG, SHAOPENG⁴
Source:	Journal of Atmospheric & Oceanic Technology. Jun2026, Vol. 43 Issue 6, p657-671. 15p.
Subjects:	Extreme weather, Image recognition (Computer vision), Rainfall, Global warming, Monsoons, Atmospheric models
Geographic Terms:	China
Abstract:	An image recognition algorithm (IRA) is proposed for the rapid detection of large-scale persistent extreme precipitation events (LPEPEs). Compared with the traditional iterative algorithm (TIA), the IRA reduces computational cost by two orders of magnitude while maintaining the same hit rate and false alarm ratio. Three independent precipitation products}the China Meteorological Administration’s Daily Precipitation Dataset of China (DPDC), fifth generation ECMWF atmospheric reanalysis (ERA5) reanalysis, and GPM IMERG, with resolutions of 0.58, 0.258, and 0.18}were jointly employed for cross validation from 2000 to 2021. All three datasets indicate that LPEPEs are predominantly concentrated in June–August, consistent with China’s primary rainy season, and the northernmost initiation latitudes of LPEPEs expand northward from approximately 308N in April to around 408N in September, mirroring the seasonal northward shift of China’s monsoon rainband; this consistency, together with coherent results from the three independent datasets, confirms the robustness of the IRA in LPEPE detection. Besides, all datasets consistently reveal a pronounced southward displacement preference of nontyphoon LPEPEs during the 22 years, with the largest displacement occurring in boreal summer and mei-yu season. Composite analyses indicate that this preference is governed by synergistic large-scale synoptic forcing and internal thermodynamic feedback, enhanced by positive local feedback, and supported by the three independent datasets. The southward displacement preference persists after 2013 despite accelerated global warming, underscoring the robustness of its underlying dynamic–thermodynamic mechanism. [ABSTRACT FROM AUTHOR]
	Copyright of Journal of Atmospheric & Oceanic Technology is the property of American Meteorological Society and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	An image recognition algorithm (IRA) is proposed for the rapid detection of large-scale persistent extreme precipitation events (LPEPEs). Compared with the traditional iterative algorithm (TIA), the IRA reduces computational cost by two orders of magnitude while maintaining the same hit rate and false alarm ratio. Three independent precipitation products}the China Meteorological Administration’s Daily Precipitation Dataset of China (DPDC), fifth generation ECMWF atmospheric reanalysis (ERA5) reanalysis, and GPM IMERG, with resolutions of 0.58, 0.258, and 0.18}were jointly employed for cross validation from 2000 to 2021. All three datasets indicate that LPEPEs are predominantly concentrated in June–August, consistent with China’s primary rainy season, and the northernmost initiation latitudes of LPEPEs expand northward from approximately 308N in April to around 408N in September, mirroring the seasonal northward shift of China’s monsoon rainband; this consistency, together with coherent results from the three independent datasets, confirms the robustness of the IRA in LPEPE detection. Besides, all datasets consistently reveal a pronounced southward displacement preference of nontyphoon LPEPEs during the 22 years, with the largest displacement occurring in boreal summer and mei-yu season. Composite analyses indicate that this preference is governed by synergistic large-scale synoptic forcing and internal thermodynamic feedback, enhanced by positive local feedback, and supported by the three independent datasets. The southward displacement preference persists after 2013 despite accelerated global warming, underscoring the robustness of its underlying dynamic–thermodynamic mechanism. [ABSTRACT FROM AUTHOR]
ISSN:	07390572
DOI:	10.1175/JTECH-D-25-0108.1