Complex and diverse topography and frequent lightning activities in Inner Mongolia pose a serious threat to agricultural production and infrastructure safety. Based on geographic information system and spatial interpolation techniques, this study explores the correlation between lightning activity and terrain features in Inner Mongolia and establishes an accurate prediction model. The study adopts the improved DBSCAN algorithm to cluster lightning activities, combines the kernel density estimation to adaptively determine the clustering parameters, utilizes kriging interpolation to determine the lightning fall area, and fits the thunderstorm movement trajectory through the least squares method. The experimental results show that the improved DBSCAN algorithm has an average offset error of 1.16 km on thundercloud center of mass prediction, which is significantly better than the linear extrapolation method (1.92 km) and the least squares method (3.09 km). The prediction accuracy is more than 80% and the false alarm rate is controlled below 40%. The topographic analysis found that the frequency of intensity lightning is highest on the northeast slope; the influence of slope on lightning density decreases gradually with the increase of lightning intensity, and Class I to III lightning is prone to occur on steep slopes. The study can provide scientific basis for the layout of lightning protection facilities, lightning disaster risk assessment and early warning in Inner Mongolia, and has practical application value for enhancing regional disaster prevention and mitigation capability.
