Abstract: Acid Mine Drainage (AMD) is one of the major environmental issues arising from coal mining, posing serious threats to ecosystems and human health. This study selected the Linkeng Coal Mine area in Longyan, Fujian as a case study and employed Self-Organizing Maps (SOM) technology to systematically analyze the pollution characteristics of AMD, aiming to accurately assess the current AMD pollution status and effectively identify the main pollution sources. The study comprehensively analyzed four water quality indicators including pH, Fe, Mn, and SO₄^2- of 37 water samples using the SOM method, and finally divided the samples into four clusters with varying degrees of pollution, clearly revealing the spatial distribution characteristics of water bodies with different levels of pollution. The clusters, in order from the highest to the lowest pollution levels, are Cluster IV, Cluster III, Cluster II, and Cluster I. Samples in clusters III and IV show severe pollution characteristics, mainly located near coal mine water inrush points and coal waste stone stacking areas, which are key areas for future remediation efforts. Samples in cluster II are primarily affected by the influx of polluted water; cluster I contains the most samples, indicating that most of the water bodies in the area are minimally affected by AMD. The SOM method offers an effective tool for assessing AMD pollution characteristics and has the potential to be applied in other coal mine areas. Future studies need to increase monitoring frequency to capture the impact of seasonal changes on water quality. As the monitoring data continues to increase, the application potential of the SOM method will become more evident.

Key words: acid mine drainage, Self-Organizing Map, pollution assessment, cluster analysis