Background Sub-alpine meadows are one of the important ecosystems in mountainous regions of northern China and beneficial for biodiversity conservation, hydrological regulation, and soil and water conservation. Dongling Mountain is in the eastern margin of China's subalpine meadow zone and presents high ecological sensitivity. The sub-alpine meadow in Dongling Mountain has been affected by both extreme rain event and anthropogenic disturbance. Increasing heavy rain events, prolonged drought, and blooming tourism have resulted in serious soil erosion and vegetation degradation. Understanding the mechanisms and spatial dynamics of meadow degradation is essential to developing adaptive strategies for ecological restoration and sustainable mountain land management.

Methods This study integrated multi-temporal high-resolution remote sensing imagery (2012, 2016, 2020, and 2023) with extensive field investigations to examine the spatiotemporal evolution of meadow degradation in Dongling Mountain. Vegetation degradation was assessed using the Visible-band Difference Vegetation Index (VDVI). Three major erosion types were classified, including patch, sheet, and linear. Field soil samples in different erosion zones were used for soil nutrient content measurement, including total nitrogen (TN), total carbon (TC), and soil organic matter (SOM) as well as vegetation.

Results 1) Over the past decades, mean annual temperature has risen, extreme rainfall events have become more intense, and dry spells have lengthened. This hydro-climatic volatility has destabilized surface hydrology, producing stronger, flashier runoff that accelerates soil detachment. 2) Erosion morphology has followed a clear succession from isolated patches type to aggregated polygon expansion type and ultimately to continuous linear gullies. The extreme storm of July 2023 alone enlarges the total eroded area by 28%, with linear types growing fastest along flow-convergence lines. This patch-sheet-linear trajectory reveals the dominant role of topographic drainage networks in guiding erosion expansion under extreme rainfall events. 3) Soil and vegetation attributes deteriorate systematically along that spatial gradient. Linear erosion belts demonstrate the lowest total nitrogen, total carbon and soil organic matter contents, reflecting severe nutrient depletion and structural collapse. Plant communities at the linear erosion belts are dominated by a few pioneer or stress-tolerant species, and diversity indices drop sharply. Zonal and patchy sectors show moderate to mild nutrient loss yet retain higher Shannon-Wiener, Simpson and Pielou values, indicating early-stage degradation with realistic prospects for natural or assisted recovery.

Conclusions The degradation of sub-alpine meadow in Dongling Mountain reflects a phased and spatially structured process driven by extreme weather patterns. Linear erosion possesses the greatest ecological threat due to its intensity and irreversibility, while patchy erosion areas present the highest restoration potential. Restoration strategies should be tailored by erosion type and degradation stage, combining preventive measures in early-stage areas with structural and biological interventions in severely degraded zones. The findings provide a scientific basis for precision management and adaptive restoration in fragile mountainous meadow ecosystems under on-going climate change.