Landslide mapping using satellite imagery and machine learning algorithms

Cyclone Idai made landfall on 15th March near Beira, Mozambique, and caused heavy rainfall across Mozambique, Malawi, Madagascar, and eastern Zimbabwe. Chimanimani District of Zimbabwe received 200 to 400 mm rainfall between 15th and 19th March, which caused widespread flooding and triggered thousands of landslides. This study aims to map the landslides in Chimanimani District and differentiate concurrent flooding from the landslides using high resolution PlanetScope imagery and DEM. Three machine learning algorithms namely, Random Forest, Artificial Neural Network, and Support Vector Machine have been deployed for the supervised landslide classification.飓风伊代（Cyclone Idai）于3月15日在莫桑比克贝拉附近登陆，在莫桑比克、马拉维、马达加斯加及津巴布韦东部引发强降雨。津巴布韦奇马尼马尼地区（Chimanimani District）在3月15日至19日期间累计降雨量达200至400毫米，由此引发大范围洪涝灾害与数千起山体滑坡。本研究旨在对该地区的山体滑坡进行空间制图，并借助高分辨率PlanetScope影像与数字高程模型（Digital Elevation Model，DEM）区分同期发生的洪涝与山体滑坡。本次研究采用三种机器学习算法开展监督式山体滑坡分类任务，分别为随机森林（Random Forest）、人工神经网络（Artificial Neural Network）与支持向量机（Support Vector Machine）。