Asia is one of the largest dust source regions in the world. However, the temporal variations and drivers of different types of dust events in this region remain unclear. Based on surface observation data, we explored spatiotemporal changes in three types of dust events and their driving factors in Asia by using machine learning methods. Results indicated that the frequency of moderate dust events (MDE) and severe dust events (SDE) decreased significantly from 2000 to 2022, which could be primarily attributed to a decrease in strong wind days (contribution >50%), and to a lesser extent to increases in soil moisture, precipitation, and leaf area index (LAI). When the daily maximum wind speed exceeds 13.0 m/s, the probability of MDE tends to decrease, while the probability of SDE tends to increase. These findings enhance our understanding of the variation in frequency and intensity of dust events in response to climate change.

亚洲是世界上最大的尘埃源地区之一。然而，该地区不同类型沙尘事件的时间变化和驱动因素仍不清楚。基于地表观测数据，我们利用机器学习方法探讨了亚洲 3 种沙尘事件的时空变化及其驱动因素。结果表明，从 2000 年到 2022 年，中度沙尘事件 （MDE） 和重度沙尘事件 （SDE） 的频率显著下降，这可能主要归因于强风天数的减少 （贡献 >50%），较小程度上归因于土壤水分、降水和叶面积指数 （LAI） 的增加。当日最大风速超过 13.0 m/s 时，发生 MDE 的概率趋于降低，而发生 SDE 的概率趋于增加。这些发现增强了我们对响应气候变化的沙尘事件频率和强度变化的理解。