College of Global Change and Earth System Science, Beijing Normal University, 100875 Beijing, China.
School of Geospatial Engineering and Science, Sun Yat-Sen University, 519082 Zhuhai, China; Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), 519082 Zhuhai, China; Joint Center for Global Change Studies, 100875 Beijing, China.
Sci Total Environ. 2020 Sep 15;735:139174. doi: 10.1016/j.scitotenv.2020.139174. Epub 2020 May 7.
Soil fugitive dust (SFD) is an important contributor to ambient particulate matter (PM), but most current SFD emission inventories are updated slowly or have low resolution. In areas where vegetation coverage and climatic conditions undergo significant seasonal changes, the classic wind erosion equation (WEQ) tends to underestimate SFD emissions, increasing the need for higher spatiotemporal data resolution. Continuous acquisition of precise bare soil maps is the key barrier to compiling monthly high-resolution SFD emission inventories. In this study, we proposed taking advantage of the massive Landsat and Sentinel-2 imagery data sets stored in the Google Earth Engine (GEE) cloud platform to enable the rapid production of bare soil maps with spatial resolutions of up to 10 m. The resulting improved spatiotemporal resolution of wind erosion parameters allowed us to estimate SFD emissions in Beijing as being ~5-7 times the level calculated by the WEQ. Spring and winter accounted for >85% of SFD emissions, while April was the dustiest month with SFD emissions of PM exceeding 11,000 t. Our results highlighted the role of SFD in air pollution during winter and spring in northern China, and suggested that GEE should be further used for image acquisition, data processing, and compilation of gridded SFD inventories. These inventories can help identify the location and intensity of SFD sources while providing supporting information for local authorities working to develop targeted mitigation measures.
土壤扬尘（SFD）是环境颗粒物（PM）的重要贡献者，但大多数当前的 SFD 排放清单更新缓慢或分辨率较低。在植被覆盖和气候条件发生显著季节性变化的地区，经典的风蚀方程（WEQ）往往会低估 SFD 排放，因此需要更高的时空数据分辨率。精确的裸土地图的连续获取是编制每月高分辨率 SFD 排放清单的关键障碍。在本研究中，我们提出利用大量存储在 Google Earth Engine（GEE）云平台中的 Landsat 和 Sentinel-2 图像数据集的优势，以高达 10m 的空间分辨率快速生成裸土地图。改进后的风蚀参数时空分辨率使我们能够估计北京的 SFD 排放量是 WEQ 计算值的 5-7 倍左右。春季和冬季占 SFD 排放量的>85%，而 4 月是扬尘最严重的月份，SFD 排放的 PM 超过 11000t。我们的研究结果强调了 SFD 在冬季和春季中国北方空气污染中的作用，并表明 GEE 应进一步用于图像采集、数据处理和网格化 SFD 清单的编制。这些清单可以帮助识别 SFD 源的位置和强度，并为地方当局制定有针对性的缓解措施提供支持信息。
