Wetland delineation and characterization layers 2014-2017, Rwanda

Wetlands are highly productive ecosystems and provide a range of goods and services on local to global scales and have become key sites of agricultural development in sub-Saharan Africa. This is the case for East African countries such as Rwanda, where agricultural expansion into wetlands and agricultural intensification are crucial elements of the government's strategy to increase food production and decrease dependence from global food markets. However, spatially explicit information of the location and status of wetlands for informed decision making is lacking or not up to date. In the related research article, we therefore develop a framework comprising the following spatial layers: Delineation, Surface Water Occurrence (SWO), Land Use/Land Cover (LULC) classification and Wetland Use Intensity (WUI). These layers are based on satellite imagery from the European Copernicus Programme and cover the whole of Rwanda. The Delineation is derived from a static Potential Wetlands layer and Sentinel-2 imagery using object-based image analysis (OBIA). The SWO is a per-pixel count of the times a location was detected as flooded using Sentinel-1 imagery. For the LULC classification, an object-based decision tree classifier is applied to Sentinel-2 imagery in Combination with the Delineation layer and the SWO. The WUI layer is based on the Absolute Mean Spectral Dynamics indicator depicting pixel-wise changes in reflectance values in Sentinel-2 bands relevant for wetland vegetation and hydrological dynamics. All layers refer to the year 2017, except for the SWO layer, which covers surface water dynamics from 2014 to 2017. The layers can be used individually or in combination and thus adapted to the different information needs for sustainable wetland management.
This dataset contains the Potential Wetlands map based on topographic indices, the Wetland Delineation derived from it, the Surface Water Occurrence layer, the Land Use/Land Cover map, and the Wetland Use Intensity layer. For appropriate map display we provide QGIS layerstyle files associated with these layers.

湿地（Wetlands）是生产力极高的生态系统，可为局地乃至全球范围提供多样产品与服务，现已成为撒哈拉以南非洲地区农业发展的核心区域。以东非国家卢旺达为例，向湿地拓展农业空间以及推进农业集约化发展，是该国政府提升粮食产量、降低对全球粮食市场依赖的战略核心举措。然而，当前缺乏支撑科学决策的、空间明晰的湿地位置与现状信息，且相关信息往往未及时更新。在本项关联研究论文中，我们构建了一套包含以下空间图层的分析框架：湿地划界图层（Delineation）、地表水出现频次图层（Surface Water Occurrence, SWO）、土地利用/土地覆被（Land Use/Land Cover, LULC）分类图层以及湿地利用强度图层（Wetland Use Intensity, WUI）。上述图层均基于欧洲哥白尼计划（Copernicus Programme）的卫星影像数据，覆盖卢旺达全境。其中，湿地划界图层通过面向对象影像分析（object-based image analysis, OBIA），由静态潜在湿地图层与哨兵-2号（Sentinel-2）影像提取得到；地表水出现频次图层为基于哨兵-1号（Sentinel-1）影像的逐像元统计结果，记录了某一位置被检测为积水的总次数。在土地利用/土地覆被分类环节，我们结合湿地划界图层与地表水出现频次图层，采用面向对象的决策树分类器对哨兵-2号影像开展分类处理；湿地利用强度图层则基于绝对平均光谱动态指标构建，该指标可逐像元反映与湿地植被及水文动态相关的哨兵-2号波段反射率变化情况。所有图层均以2017年为基准年份，唯有地表水出现频次图层的时间跨度为2014年至2017年，用以表征地表水动态变化。上述图层可单独使用或组合应用，从而适配可持续湿地管理中多样化的信息需求。本数据集包含基于地形指数生成的潜在湿地地图、由此衍生的湿地划界图层、地表水出现频次图层、土地利用/土地覆被地图以及湿地利用强度图层。为便于图层的可视化展示，我们还提供了与上述图层配套的QGIS图层样式文件。