High Mountain Asia Daily 5km Landslide Hazard Indicator, Version 1

This data set projects the daily hazard of rainfall-triggered landslides in the High Mountain Asia region from 2015 through 2100, at 5 km resolution. Projections are provided for two Shared Socioeconomic Pathways (SSPs)—SSP2-4.5 and SSP5 8.5—based on downscaled temperature and precipitation projections from a 30-member ensemble climate model. Landslide hazard is represented by a landslide hazard indicator (LHI), computed with a machine learning (ML) model trained on historical temperatures and precipitation from 1990 through 2019 and a catalog of documented landslides. Two historical LHI data sets are also available: the ML model LHIs generated for 1990 through 2019; and retrodicted LHIs computed by inputting downscaled temperatures and precipitation for 1990 through 2014 to the ensemble climate model.本数据集以5公里的空间分辨率，预测2015年至2100年亚洲高山区（High Mountain Asia）内降雨诱发滑坡的逐日灾害风险。该数据集提供两种共享社会经济路径（Shared Socioeconomic Pathways, SSPs）——SSP2-4.5与SSP5-8.5的预测结果，数据基于30成员集合气候模式的降尺度气温与降水预测数据。 滑坡灾害以滑坡灾害指数（landslide hazard indicator, LHI）表征，该指数由机器学习（machine learning, ML）模型计算得到；模型训练采用1990年至2019年的历史气温、降水数据，以及已记录的滑坡目录。 另有两套历史LHI数据集可供使用：其一为1990年至2019年由上述机器学习模型生成的LHI；其二为将1990年至2014年的降尺度气温与降水数据输入集合气候模式后计算得到的回溯预测LHI。