The evaluation of model simulations and analysis of error sources for typical spring dust events in China.

Dust weather is a type of hazardous weather phenomenon predominantly occurring in arid and semi-arid regions. It arises from the interaction between specific desert ecological environments and meteorological conditions, significantly affecting climate change, ecological systems, human health, and transportation. Over the past decade, the frequency of blowing dust and dust storms in China has exhibited an upward trend, particularly with multiple occurrences of severe dust storms during the spring in the last four years. This study employs the CMA_Meso/CUACE_SDS model to analyze 15 typical dust events that transpired in China from 2021, 2023, and 2024, systematically assessing the model's forecasting performance concerning key characteristics of dust, such as transmission paths, distribution ranges, and durations. Furthermore, based on the different systems that trigger dust weather (Mongolian cyclones and cold fronts or only cold fronts), we selected four representative dust cases with varying process types (blowing dust or dust storms), focusing on analyzing the model's forecast accuracy in relation to weather systems (Mongolian cyclone intensity and cold front intensity) and local meteorological factors (temperature, wind speed, etc.). The results demonstrate that the CMA_Meso/CUACE_SDS model can accurately simulate the transmission paths, distribution ranges, and durations of dust events; however, it tends to overestimate the intensity of both Mongolian cyclones and cold fronts, as well as wind speeds near dust source areas. This overestimation further exacerbates dust emissions from these areas, ultimately diminishing the model's forecasting accuracy for dust events. This study offers valuable insights for enhancing the model's capability to simulate transboundary dust events in the future.