地震的发生具有非线性特征,分形理论能够刻画地震时空分布特征及其变化过程。本文基于R/S分析方法确定阿富汗主要地震带的分形特征,利用ARIMA模型对兴都—库什山地震带可能发生的年度最大震级进行预测。R/S分析表明,兴都—库什山地震带Hurst指数为0.9125,地震活动记忆周期为8年; 苏莱曼山地震带Hurst指数为0.7281,地震活动记忆周期为9年。兴都—库什山和苏莱曼山地震带地震活动的变化趋势与历史变化一致,且兴都—库什山地震带的趋势延续性比苏莱曼山地震带更为显著。ARIMA模型预测结果显示,2022—2026年兴都—库什山地震带可能发生的年度最大震级分别为M_b6.2、M_b6.1、M_b5.8、M_b5.8和M_b6.1。

Earthquake occurrence exhibits nonlinear characteristics,and fractal theory effectively describes the temporal and spatial distribution and evolution of earthquakes. Using R/S analysis,this paper determines the fractal characteristics of the major seismic belts in Afghanistan and employs the ARIMA model to predict the annual maximum magnitude of the Hindu-Kush seismic belt. The Hurst index of the Hindu-Kush seismic belt is 0.9125,indicating a seismic activity memory period of 8 years. For the Sulaiman seismic belt,the Hurst index is 0.7281,with a memory period of 9 years. The variation trends of seismic activity in both the Hindu-Kush and Sulaiman seismic zones align with historical patterns,though the trend continuity is more pronounced in the Hindu-Kush zone. According to the ARIMA model,the predicted maximum annual magnitudes for the Hindu-Kush seismic belt from 2022 to 2026 are M_b6.2,M_b6.1,M_b5.8,M_b5.8,and M_b6.1,respectively.

P315

应急管理部救援协调和预案管理局项目(2023)资助