Abstract: The spontaneous combustion of coal gangue poses a serious threat to the ecological restoration of gangue dumps and the safe utilization of resources. Accurately identifying the spatial distribution of combustion hazards and quantitatively assessing spontaneous combustion risk levels are critical for effective prevention and control. To address this critical scientific issue, a self-ignition spatial identification method and risk assessment system has been established based on the integration of multi-source information and technological fusion. The spontaneously combusting gangue dump in the Xishechang area of Fushun Mining District is examined. A 3D-Kriging spatial interpolation temperature field reconstruction model was developed to analyze the spatial distribution patterns and scale of combustion hazard sources. To enable graded and zoned control of spontaneous combustion risks, a three-source risk evaluation system (spontaneous combustion development – harmful gases – spontaneous combustion disaster) was established, along with a comprehensive evaluation (CE) model for gangue fire hazards. In this model, satellite remote sensing time-series monitoring data were introduced to objectively evaluate the development risk and trend of spontaneous combustion, and a development trend function \mu was constructed, effectively overcoming the problem of missing historical monitoring data on coal gangue combustion. To assess the atmospheric pollution risk and environmental impact, a method was proposed to determine the weight of mixed harmful gases by combining the concentration of mixed harmful gases with the toxicity hazard index. To objectively assess the threat of spontaneous combustion disasters, an evaluation method was established using economic loss and casualties as evaluation indicators. The results show that the spontaneous combustion hazard sources in the gangue dump of the study area are mainly distributed near the south slope of the site, with 49 753.4 m³ of gangue exceeding 90 °C. The prediction accuracy of the 3D-Kriging spatial interpolation, with MAE, RMSE, and R² values of 2.01 ℃, 2.49 ℃, and 0.91, respectively, proves that this method can accurately locate the spatial position and scale of spontaneous combustion hazard sources. In addition, the CRITIC method was used to calculate the weights of the “three-source” sub-goals in the study area, which were 0.335, 0.069, and 0.596, respectively, indicating that the risk of spontaneous combustion disaster has the greatest impact in the CE model of this gangue dump. A reasonable land use planning deployment for the gangue dump can significantly reduce the hazard level of coal gangue spontaneous combustion. By applying the proposed methodology, a true three-dimensional spatial identification of spontaneous combustion hazards in gangue dumps is achieved, and a quantitative multi-source risk assessment framework is established, thereby providing theoretical support for ecological restoration and informed decision-making regarding the safe utilization of resources.