Abstract: The ecological environment in Xinjiang Gobi open-pit mining area is fragile and easily affected by natural and man-made factors. Therefore, the ecological environment in Xinjiang mining area has been paid more attention. Taking Wucaiwan open-pit mining area in Xinjiang Uygur Autonomous Region as study area, coal dust index, salinity index and land degradation index were introduced to construct arid coal mine ecological index(I_ACME) based on the remote sensing ecological index(I_RSE) and Landsat images from 1990 to 2020. M-K test and Sen trend analysis were used to reveal the spatial and temporal evolution of I_ACME in the study area, which evaluated the eco-environmental quality of arid mining area. Multiple regression, neural network, random forest and support vector machine methods were employed to establish the relationship between climate factors and I_ACME since mining and the prediction of I′_ACME based on only climatic conditions was obtained. The influence of mining activity was quantitatively assessed by the residual analysis and difference-in-difference method. The results showed that:(1) The average I_ACME of the study area was 0.28 during the past 30 years, indicating the quality of ecological environment was poor. The I_ACME of study area showed a decreasing trend from 1990(0.30) to 2020(0.22). The percentage of degraded area was 62.63%,which was much higher than that of improved area(28.3%). Due to the influence of mining activities, the percentage of significantly degraded area increased by 12.96%,mainly located in the outward radiation area around the mine. The proportion of significantly improved area decreased by 17.28%.(2) The relationship model between climate factors and I_ACME was established using multiple regression, neural network, random forest and support vector machine methods in Wucaiwan mining area. The results indicated that the random forest of sparrow search algorithm(SSA-RF) had the highest precision in modeling with R² being 0.82 and E_RMS being 0.109 in verification set. SSA-RF had higher reliability and stronger applicability among modeling methods used.(3) SSA-RF was adopted to predict Quantitative analysis I′_ACME based on climatic conditions. The ecological environment of the study area was mainly in poor(0.2-0.4) and medium(0.4-0.6) levels. Assuming no mining activity, the average value of I′_ACME in the study area ranged from 0.26 to 0.59 during 2006-2020,and the average value of I′_ACME was about 0.40,which was significantly higher than I_ACME in the same year, mainly concentrated in the middle of the study area where the mines was located.(4) The residual analysis demonstrated that mining activities in the study area had a negative impact on the ecological environment with an average residual(δ) of-0.116. The results of difference-in-difference method showed that the negative value was-0.112,which further indicated that coal mining had a negative impact on the ecological environment, leading to a decline in the quality of the ecological environment. The results of these two methods were similar, indicating that SSA-RF and difference-in-difference method all can be used to quantitatively analyze the impact of coal mining.(5) The proportion of ecological environment degradation and improvement caused by mining activities was 57.52% and 34.67%,respectively. While the proportion of ecological environment degradation and improvement caused by climatic conditions was 42.28% and 65.33%,respectively. The proportion of ecological environmental degradation caused by mining was more than 50%,indicating that mining activity was one of the important factors affecting ecological degradation. The ecological environment of Wucaiwan mining area needs to be restored urgently. Manual intervention should be supplemented to promote the environmental protection, ecosystem restoration and comprehensive management in the study area.