露天煤矿开采产生的煤炭粉尘对矿区及周边大气、植被、土壤和居住环境造成影响，但煤炭粉尘污染程度、污染范围量化及其长时序趋势变化等相关研究滞后。提出了煤炭粉尘最大影响范围和持续影响范围概念，基于Google Earth Engine 平台和2006—2021年Landsat TM/ETM⁺/OLI影像，反演增强型煤炭粉尘指数(Enhanced Coal Dust Index, ECDI)和煤炭粉尘污染程度，综合运用线性回归和叠置分析方法，揭示了胜利矿区开采过程中煤炭粉尘的时序变化及空间差异特征。结果表明：① 煤炭粉尘污染年际变化可分为严重期(2006—2009年)、改善期(2010—2013年)和稳定期(2014—2021年)，与开采生命周期基本一致，且随开采时间的推进呈现先强后减弱的变化趋势；② 2019—2021年春季和夏季煤炭粉尘污染程度普遍偏高，季相变化在相对无污染区、轻度污染区和中度污染区相对明显；③ 煤炭粉尘持续影响范围的总体识别精度为92.67%(kappa系数为0.85)，随着原煤产量的递增，常年持续和最大影响范围不断缩小，下降幅度分别为43%和80%，年内持续影响范围空间分布保持基本稳定，年内最大影响范围下降；④ 2006—2021年相对无污染区面积净增191.69 km²(62.70%)，严重污染区面积净减4.59 km²(19.74%)，煤炭粉尘污染程度呈下降、上升趋势的面积分别占矿区面积的52.11%和4.07%。2021年相对无污染区和轻度污染区面积达矿区总面积的91.91%(约601.13 km²)。

The coal dust produced by surface coal mining has a serious impact on the atmosphere, vegetation, soil and residential environment in the coal mining area and its surrounding area, but the related research on the coverage of coal dust pollution, the quantification of pollution area and the trend of long-term series lags behind. In this study, the concepts of the maximum influence area and the continuous influence area were put forward. Based on the Google Earth Engine platform and the Landsat TM/ETM⁺/OLI images during 2006—2021, the ECDI (Enhanced Coal Dust Index) and the pollution coverage of coal dust were inversed. By using the linear regression and superposition analysis methods, the time series variation and spatial difference characteristics of coal dust pollution in the Shengli coal mining area were revealed. The results show that: ① the inter-annual variation of coal dust pollution can be categorized into three phases: a severe period (2006—2009), an improvement period (2010—2013), and a stable period (2014—2021). These phases correspond with the coal mine lifecycle, showing an initial increase followed by a decreasing trend. ② From 2019 to 2021, the coal dust pollution coverage was generally higher in spring and summer. Seasonal spatial variations were more pronounced in the relatively unpolluted area, lightly polluted area and moderately polluted area. ③ The overall accuracy of identifying the influence area of coal dust was 92.67% (kappa coefficient = 0.85). As coal production increased, both the perennial continuous influence area and the maximum influence area decreased, with the reductions of 43% and 80%, respectively. The spatial distribution of the annual continuous influence area remained relatively stable, while the maximum influence area decreased. ④ From 2006 to 2021, the relatively unpolluted area increased by 191.69 km², accounting for 62.70% of the original area. The seriously polluted area decreased by 4.59 km², representing 19.74% of the original area. The coal dust pollution area with a downward trend accounted for 52.11% of the coal mining area, while 4.07% showed an upward trend. As of 2021, the relatively unpolluted area and lightly polluted area accounted for 91.91% of total mining area (about 601.13 km²).