Divergent Responses of Productivity and Its Stability to Climate Extremes in Northern China’s Grasslands

Climate forecasts predict that extreme climate events will become more frequent in the next decades, which will pose serious challenges to both the function and structure of grassland ecosystems. Understanding how grassland ecosystems respond to extreme climates is crucial for maintaining their stability. This study examined the spatiotemporal dynamics of vegetation productivity, its interannual stability index (STI), and extreme climate events in China’s northern grasslands using climate data and normalized difference vegetation index (NDVI) data from 1982 to 2021. We used the Extreme Gradient Boosting model and Shapley Additive Explanations to identify the main climatic drivers of NDVI and STI variability. The findings revealed that northern grasslands have shown a general warming and wetting trend over the past 40 years, alongside more frequent extreme high-temperature and heavy-precipitation events. Grassland vegetation NDVI exhibited a significant rising trend (slope = 0.0004 year −1 , P < 0.05), whereas STI demonstrated a significant declining tendency (slope = −0.0074 year −1 , P < 0.05). The increase in NDVI was mainly driven by the increased precipitation, while the decline in STI is jointly driven by water and thermal factors for the northern grasslands. Specifically, warming-induced evapotranspiration and intensified extreme precipitation likely exacerbated hydrothermal imbalances, thereby amplifying interannual variability in vegetation dynamics. These findings highlight the vulnerability of northern grasslands to climate extremes. They also provide valuable guidance for managing grassland ecosystems sustainably in a changing climate.