Implementation of a Surface Water Extent Model using Cloud-Based Remote Sensing - Code and Maps

This data release comprises the raster data files and code necessary to perform all analyses presented in the associated publication.
The 16 TIF raster data files are classified surface water maps created using the Dynamic Surface Water Extent (DSWE) model implemented in Google Earth Engine using published technical documents. The 16 tiles cover the country of Cambodia, a flood-prone country in Southeast Asia lacking a comprehensive stream gauging network. Each file includes 372 bands. Bands represent surface water for each month from 1988 to 2018, and are stacked from oldest (Band 1 - January 1988) to newest (Band 372 - December 2018). DSWE classifies pixels unobscured by cloud, cloud shadow, or snow into five categories of ground surface inundation; in addition to not-water (class 0) and water (class 1), the DSWE algorithm distinguishes pixels that are less distinctly inundated (class 2: “moderate confidence”), comprise a mixture of vegetation and water (class 3: “potential wetland”), or are of marginal validity (class 4: “water or wetland - low confidence”). Class 9 is applied to classify clouds, shadows and hill shade.
Two additional documents accompany the raster image files and XML metadata. The first provides a key representing the general location of each raster file. The second file includes all Google Earth Engine Javascript code, which can be used online (https://code.earthengine.google.com/) to replicate the monthly DSWE map time series for Cambodia, or for any other location on Earth. The code block includes comments to explain how each step works.
These data support the following publication:
These data support the following publication: Soulard, C.E., Walker, J.J., and Petrakis, R.E., 2020, Implementation of a Surface Water Extent Model in Cambodia using Cloud-Based Remote Sensing: Remote Sensing, v. 12, no. 6, p. 984, https://doi.org/10.3390/rs12060984.

本数据集发布包含支撑关联论文中全部分析工作所需的栅格数据文件与代码。本次发布的16份TIF（Tagged Image File Format）栅格数据文件，为采用谷歌地球引擎（Google Earth Engine）结合已发表技术文档实现的动态地表水体提取（Dynamic Surface Water Extent, DSWE）模型生成的分类地表水体图。16幅瓦片覆盖东南亚洪涝频发且缺乏完整水文测站网络的柬埔寨全境。每份文件包含372个波段，各波段对应1988年至2018年各月的地表水体数据，按时间顺序堆叠（第1波段为1988年1月，为最早数据；第372波段为2018年12月，为最新数据）。DSWE模型将未被云、云阴影或积雪遮蔽的像素划分为5类地表淹没状态：除非水体（类别0）与水体（类别1）外，该算法还可区分出淹没可信度中等的像素（类别2：“中等置信度”）、植被与水体混合的像素（类别3：“潜在湿地”）、有效性存疑的像素（类别4：“水体或湿地——低置信度”）。类别9用于标记云、阴影与山体阴影区域。本数据集除栅格影像文件与XML（Extensible Markup Language）元数据外，还附带两份辅助文档：第一份为各栅格文件的位置索引图；第二份包含全部谷歌地球引擎JavaScript代码，可通过在线平台（https://code.earthengine.google.com/）运行，用于复现柬埔寨月度DSWE水体图时间序列，或适配全球任意其他区域。代码块内附有注释，用于说明各步骤的运行逻辑。本数据集支撑以下发表论文：Soulard, C.E.、Walker, J.J.与Petrakis, R.E.，2020年，《基于云遥感的柬埔寨地表水体范围模型实现》，刊载于《Remote Sensing》，第12卷第6期，第984页，https://doi.org/10.3390/rs12060984。