Abstract:

The analysis of the distribution characteristics of regional geological disasters and the influ⁃

encing factors of disasters is conducive to disaster prevention and mitigation and sustainable development in

mining areas.Taking Panzhihua City，a steel base，as the research area，471 historical geological disasters

in the research area were collected，189 main mines were sorted out，and elevation，slope and fault zone

were selected as analysis factors.The kernel density analysis of geological disasters and mines is carried out

to study the spatial correlation between the two.On this basis，the CatBoost machine learning method is used

to quantitatively analyze the influence of each influencing factor on geological disasters in the study area.The

results show that there is a strong spatial correlation between geological disasters and mines.The nuclear den⁃

sity and rainfall of mine have the greatest impact on geological disasters，which are 32.00% and 25.10%，re⁃

spectively，followed by the distance from the railway and active faults to the area. The research shows that

the mining activities in the study area have a significant impact on the occurrence of geological disasters.It is

recommended to rationally plan regional mining activities，do a good job in geological disaster monitoring

and early warning，and promote regional sustainable development.

Key words:

geological disasters, mines, machine learning, nuclear density analysis