Keywords
flood detection, near real-time, flood mapping, flood monitoring, spacecraft autonomy, flood sensorweb
Start Date
1-7-2006 12:00 AM
Abstract
Extreme floods have been reported to be more frequent partly due to global warming. As such, the necessity for timely detection and mapping of floods is increasingly important in order to protect lives and livelihoods. Floods affect large regions of the Earth and cannot be reliably predicted. Hydrological data from in-situ sensors are sparse and cannot map the full extent of flooding. The use of satellite-based information for assessing floods is not new. However, the problem with satellite remote sensing historically has been both the large areas affected and obtaining timely ground-based reception of satellite data. The Autonomous Sciencecraft Experiment (ASE) experiment overcomes the data size and downlink problems. For flood processes, the ASE includes a satellite-based floodwater classification algorithm (ASE_FLOOD), which reliably detects flooding as it occurs and autonomously triggers further image acquisition to map and track flood changes through time. In addition, the ASE enables more effective and timely monitoring for other dynamic transient events on Earth, which include volcanic eruptions and sea ice breakups. The Flood Sensorweb is an extension of ASE and serves to link different remote sensing assets obtained at different spatial and temporal resolutions for flood detection and monitoring. It is a demonstration in which Dartmouth Flood Observatory’s Water Surface Watch (a satellite-based global runoff monitoring system) alerts ASE operations of sites where there is potential flooding. Based on these alerts, ASE autonomously retargets NASA’s EO-1 spacecraft to verify flooding conditions at these sites, thereafter acquiring local high-resolution images of these flooded areas. The Flood Sensorweb offers an important asset for the study of transient hydrological phenomena globally, especially at remote locations. The use of autonomous change detection, triggering the needed local high-resolution imaging by automatic systems, provides the critical near real-time data needed for early detection and modeling of seasonal and extreme floods.
Autonomous Flood Sensorweb: Multi-Sensor Rapid Response and Early Flood Detection
Extreme floods have been reported to be more frequent partly due to global warming. As such, the necessity for timely detection and mapping of floods is increasingly important in order to protect lives and livelihoods. Floods affect large regions of the Earth and cannot be reliably predicted. Hydrological data from in-situ sensors are sparse and cannot map the full extent of flooding. The use of satellite-based information for assessing floods is not new. However, the problem with satellite remote sensing historically has been both the large areas affected and obtaining timely ground-based reception of satellite data. The Autonomous Sciencecraft Experiment (ASE) experiment overcomes the data size and downlink problems. For flood processes, the ASE includes a satellite-based floodwater classification algorithm (ASE_FLOOD), which reliably detects flooding as it occurs and autonomously triggers further image acquisition to map and track flood changes through time. In addition, the ASE enables more effective and timely monitoring for other dynamic transient events on Earth, which include volcanic eruptions and sea ice breakups. The Flood Sensorweb is an extension of ASE and serves to link different remote sensing assets obtained at different spatial and temporal resolutions for flood detection and monitoring. It is a demonstration in which Dartmouth Flood Observatory’s Water Surface Watch (a satellite-based global runoff monitoring system) alerts ASE operations of sites where there is potential flooding. Based on these alerts, ASE autonomously retargets NASA’s EO-1 spacecraft to verify flooding conditions at these sites, thereafter acquiring local high-resolution images of these flooded areas. The Flood Sensorweb offers an important asset for the study of transient hydrological phenomena globally, especially at remote locations. The use of autonomous change detection, triggering the needed local high-resolution imaging by automatic systems, provides the critical near real-time data needed for early detection and modeling of seasonal and extreme floods.
