Map data showing concentration of landslides caused by Hurricane Maria in Puerto Rico

On September 20, 2017, Hurricane Maria hit the U.S. territory of Puerto Rico as a category 4 storm. Heavy rainfall caused landslides in mountainous regions throughout the territory. This data release presents geospatial data describing the concentration of landslides generated by Hurricane Maria in Puerto Rico. We used post-hurricane satellite and aerial imagery collected between September 26, 2017 and October 8, 2017 to visually estimate the concentration of landslides over nearly the whole territory. This was done by dividing the territory into a grid with 4 square km cells (2 km x 2 km). Each 4 square km grid cell was classified as either containing no landslides, fewer than 25 landslides/ square km or more than 25 landslides/ square km. We used 12 WorldView satellite images (~0.5 m-resolution) available from Digital Globe, Inc. and ~0.15 m-resolution aerial imagery collected by Sanborn and QuantumSpatial. We identified landslides by locating areas with exposed soil or rock and morphology typical of landslides. Because leaves were stripped from much of the vegetation, landslide scars were readily visible in both sources of imagery. We assume that the majority of landslides were triggered by rainfall from Hurricane Maria, but rainfall from Hurricane Irma during the first week of September and rainfall from thunderstorms after Hurricane Maria may have also initiated landslides.

During this investigation, we visually examined a total area of 8103 square km, which encompassed most of the territory and nearly all the mountainous areas. Approximately 846 square km of the land area of the territory was not examined because either 1) imagery was unavailable or 2) the area was obscured by cloud cover. Approximately 61% of the examined area was unaffected by landslides. Landslides were observed in the remaining 39% of the examined area, but, for the most part, the landslide concentration was less than 25 landslides/ square km (37% of the examined area). Landslide concentration was greater than 25 landslides/ square km in about 2% of the examined area (156 square km), which includes parts of the Aasco, Mayagez, Las Maras, Maricao, Lares, Utuado, Adjuntas, and Jayuya municipalities. Based on visual examination of imagery, the municipality of Utuado appears to have been the most severely impacted, with about 40% of the municipality having a concentration of landslides greater than 25 landslides/ square km. This preliminary assessment serves to inform response and recovery efforts and as a basis for more detailed studies on the impacts of landslides in Puerto Rico caused by Hurricane Maria.

The data included in this release are: 1) a polygon shapefile (.shp) that classifies each 2 km x 2 km grid cell based on landslide concentration, and 2) a layer file (.lyr) that can be used to display the landslide concentration data. We also included a map showing the landslide concentration grid and the legal municipalities of Puerto Rico. A list of the municipality abbreviations used on the map and the corresponding municipality names are included in the attached Excel file. This data release only includes landslide concentrations mapped from satellite and aerial imagery, not the imagery itself.

2017年9月20日，四级飓风玛丽亚（Hurricane Maria）以四级风暴强度登陆美国海外领地波多黎各，强降雨引发全领地范围内山区滑坡灾害。本数据集发布了用于描述波多黎各境内飓风玛丽亚诱发山体滑坡浓度的地理空间数据。
我们采用2017年9月26日至10月8日期间收集的飓风后卫星与航空影像，通过目视解译估算了几乎覆盖全领地的山体滑坡空间浓度。具体处理方式为：将研究区划分为4平方公里（2km×2km）的网格单元，对每个网格单元按滑坡浓度划分为三类：无滑坡、每平方公里滑坡数少于25处，以及每平方公里滑坡数多于25处。
本次研究使用了Digital Globe, Inc.提供的12幅WorldView卫星影像（分辨率约0.5米），以及Sanborn与QuantumSpatial采集的分辨率约0.15米的航空影像。通过识别裸露土壤或岩石区域以及典型滑坡地貌特征来定位滑坡点；由于多数植被的叶片被飓风刮落，两种影像中的滑坡痕迹均清晰可辨。我们假设绝大多数滑坡由飓风玛丽亚的降雨触发，但9月初飓风艾尔玛（Hurricane Irma）带来的降雨以及飓风玛丽亚过境后的雷暴降雨也可能诱发部分滑坡。
本次调查共目视检查总面积达8103平方公里，覆盖了该领地的大部分区域与几乎所有山区。约846平方公里的陆地未被纳入检查范围，原因包括：1）无可用影像；2）区域被云层遮挡。经检查的区域中，约61%未受滑坡影响；剩余39%的区域观测到滑坡，其中绝大多数区域的滑坡浓度低于25处/平方公里（占检查区域的37%）。仅约2%的检查区域（156平方公里）滑坡浓度高于25处/平方公里，涉及阿斯科（Aasco）、马亚圭斯（Mayagez）、拉斯马拉斯（Las Maras）、马里考（Maricao）、拉雷斯（Lares）、乌图阿多（Utuado）、阿洪达斯（Adjuntas）以及哈尤亚（Jayuya）等自治市。通过目视影像解译，乌图阿多市似乎受滑坡影响最为严重，该市约40%的区域滑坡浓度高于25处/平方公里。本次初步评估可为灾后响应与恢复工作提供决策参考，同时可为后续针对飓风玛丽亚引发的波多黎各山体滑坡影响的深入研究奠定基础。
本次发布的数据包括：1）按滑坡浓度对每个2km×2km网格单元进行分类的多边形形状文件（.shp）；2）用于可视化展示滑坡浓度数据的图层文件（.lyr）。此外还附带了一张展示滑坡浓度网格与波多黎各法定自治市边界的地图，以及包含地图所用自治市缩写与对应全称的Excel附件。本数据集仅包含通过卫星与航空影像解译得到的滑坡浓度数据，不包含原始影像本身。