The Temporal-Spatial Variations and Potential Causes of Dust Events in Xinjiang Basin During 1960–2015

Abstract

Dust events not only cause local ecosystem degradation and desertification, but also have profound impacts on regional and global climate system, as well as air quality and human health. Dust events in Xinjiang Basin, as the important dust source of Eastern Asia, have undergone a significant change under the global warming background and may be in a new active period after 2000, which is worthy of study. This study provides the temporal and spatial variations of dust events in the Xinjiang Basin based on surface meteorological station observation data during 1960–2015. The results show that Southern Xinjiang is the main dust occurrence region where dust events are significantly more than that in the Northern Xinjiang, and each year more than 73% of dust events occurred in spring and summer. The dust index (DI), which is defined to represent the large-scale variation of dust event, shows a significant downward trend during the past 56 years with a linear decreasing rate −8.2 years−1 in Southern Xinjiang. The DI is positively correlated to surface wind speed with a mean correlation coefficient of 0.79. The declining trend of surface wind speed could explain dust events variation during 1960–2000. But in the new active period after 2000, the increase of DI is not consistent with the rising wind speed with the correlation coefficient decreasing to 0.34. It is found that, compared with 1960–1999, the average annual precipitation and frequency increased by 17.4 and 13% during 2000–2015, respectively, and the NDVI also increased at the same time, which indicates that the surface condition changes induced by the increase of precipitation might suppress the occurrence of dust. Moreover, the analysis of high-altitude wind field shows that the variation of the East Asian general circulation’s intensity, dominating the upper-level wind fields in the Xinjiang basin, will change the surface wind speed and precipitation, and further affect the occurrence of dust events.

Design a regional and multistep air quality forecast model based on deep learning and domain knowledge

“…With the rapid development of industrialization and urbanization, more and more fossil fuels are burned, resulting in the deterioration of air quality and frequent haze weather (Jerrett, 2015;Mo et al, 2021;Mu et al, 2021). As a serious environmental problem air pollution has attracted worldwide attention.…”