We developed a new approach using a cloud-based remote sensing and geospatial analysis platform, Google Earth Engine, to quantify temporal changes in river channel location and adjacent riparian vegetation extent and fraction. Our new method uses publicly available 1 m aerial images and eliminates manual processing need by incorporating an automatic image classification algorithm. Classification of riparian vegetation is enhanced by increasing the temporal resolution to monthly and by mapping vegetation at high spatial resolution. We illustrate the application of our method by characterizing temporal and spatial trends in riparian vegetation and river channel position for the mainstem of the Genesee River, New York, from 2006 to 2015. Annual change in riparian vegetation extent along the Genesee River ranged from a loss of 17% to a gain of 13%, and 64 km (28%) of the river demonstrated channel migration across consecutive aerial images. Our method also successfully distinguished differences between sections of the river above and below a dam. The enhanced capacity to map vegetation monthly allowed us to identify that seasonal active vegetation fractions along the Genesee River peaked at 75% in the summer and remained lower than 25% in winter. Our method allows stakeholders and managers to process remotely sensed imagery and investigate trends in river channel and riparian vegetation dynamics over time, while reducing the costs of data processing and storage.

中文翻译：

---

使用 Google 地球引擎评估河流地貌和河岸植被的时空变化

我们使用基于云的遥感和地理空间分析平台 Google Earth Engine 开发了一种新方法，以量化河道位置和邻近河岸植被范围和比例的时间变化。我们的新方法使用公开可用的 1 m 航拍图像，并通过结合自动图像分类算法消除了手动处理的需要。通过将时间分辨率增加到月度和以高空间分辨率绘制植被图，可以增强河岸植被的分类。我们通过描述 2006 年至 2015 年纽约杰纳西河干流河岸植被和河道位置的时空趋势来说明我们的方法的应用。沿杰纳西河沿岸植被范围的年度变化范围从损失17% 到 13% 的收益，和 64 公里（28%）的河流在连续的航拍图像中展示了渠道迁移。我们的方法还成功区分了大坝上方和下方河流部分之间的差异。每月绘制植被图的能力增强，使我们能够确定沿杰纳西河的季节性活跃植被比例在夏季达到 75% 的峰值，而在冬季保持在 25% 以下。我们的方法允许利益相关者和管理人员处理遥感图像并调查河道和河岸植被动态随时间变化的趋势，同时降低数据处理和存储的成本。每月绘制植被图的能力增强，使我们能够确定沿杰纳西河的季节性活跃植被比例在夏季达到 75% 的峰值，而在冬季保持在 25% 以下。我们的方法允许利益相关者和管理人员处理遥感图像并调查河道和河岸植被动态随时间变化的趋势，同时降低数据处理和存储的成本。每月绘制植被图的能力增强，使我们能够确定沿杰纳西河的季节性活跃植被比例在夏季达到 75% 的峰值，而在冬季保持在 25% 以下。我们的方法允许利益相关者和管理人员处理遥感图像并调查河道和河岸植被动态随时间变化的趋势，同时降低数据处理和存储的成本。