A Tool for Rapid-Repeat High-Resolution Coastal Vegetation Maps to Improve Forecasting of Hurricane Impacts and Coastal Resilience
We developed a Jupyter Notebook Application and a Graphical User Interface that use Planet Labs Super Dove 8-band, 3-meter multispectral imagery and a machine learning classification model to deliver high-resolution maps of coastal vegetation showing near real-time conditions. These products will help improve forecasts of hurricane impacts.
Natural coastal vegetated ecosystems provide nature-based solutions to climate-related impacts of hurricanes yet are subject to change each year. Coastal modelers seek simple and fast ways to obtain up-to-date coastal vegetation maps to produce more accurate coastal change forecasts. This project addresses this critical need by leveraging Planet Labs 8-band, 3-m daily satellite images (freely available to Federal employees) and extensive training data. These were combined with a machine learning classification model to develop a rapid repeat, user friendly, Graphical User Interface and Python Jupyter Notebook application that delivers high-resolution maps of coastal vegetation showing near real-time conditions. A User Manual walks individuals at all levels – novice to advanced – through the use of the tool, which helps to build capacity within the USGS remote-sensing community. The project addresses the CDI FY23 theme of climate-related data readiness by making data available for forecasts of hurricane impacts, and that indicate the state of our vegetated ecosystems.
RUSH: Rapid Remote Sensing Updates of landcover for Storm and Hurricane forecasts (Version 1.0.0)
We developed a Jupyter Notebook Application and a Graphical User Interface that use Planet Labs Super Dove 8-band, 3-meter multispectral imagery and a machine learning classification model to deliver high-resolution maps of coastal vegetation showing near real-time conditions. These products will help improve forecasts of hurricane impacts.
Natural coastal vegetated ecosystems provide nature-based solutions to climate-related impacts of hurricanes yet are subject to change each year. Coastal modelers seek simple and fast ways to obtain up-to-date coastal vegetation maps to produce more accurate coastal change forecasts. This project addresses this critical need by leveraging Planet Labs 8-band, 3-m daily satellite images (freely available to Federal employees) and extensive training data. These were combined with a machine learning classification model to develop a rapid repeat, user friendly, Graphical User Interface and Python Jupyter Notebook application that delivers high-resolution maps of coastal vegetation showing near real-time conditions. A User Manual walks individuals at all levels – novice to advanced – through the use of the tool, which helps to build capacity within the USGS remote-sensing community. The project addresses the CDI FY23 theme of climate-related data readiness by making data available for forecasts of hurricane impacts, and that indicate the state of our vegetated ecosystems.