Flood Detection by Autonomous Satellite Operations: the ASE Experience

The use of satellite-based information for assessing floods is not new. However, the problem historically has been both the huge size of satellite data and the limited downlink capability of satellites. As much as a 2- week period, for example, is required for an image to be transmitted to scientists to determine whether a flood is transpiring, especially in remote regions of the world. Developed as part of the Autonomous Sciencecraft Experiment (ASE), ASEFLOOD is a satellite-based floodwater classification algorithm that overcomes this data size and downlink problem through autonomous satellite operations. It autonomously analyzes satellite data as they are acquired onboard the spacecraft, and when flooding is detected, it responds by autonomous rescheduling of observations to acquire additional data of high temporal, spatial, and spectral resolution of the flood region until the flooding has subsided. It is operational and is monitoring selected river locations around the world for flood conditions in near real time. ASEFLOOD is running on NASA's EO-1 spacecraft using high spatial and spectral (30m/pixel and 220+ bands) optical and infrared satellite data acquired by the HYPERION sensor. The algorithms have been tested extensively both on the ground and in space through the capture of flooding in candidate regions of greatest seasonal flood potential throughout the world (e.g., Diamantina River in Australia, Yukon River in Alaska, Brahmaputra in India, Yellow River in China). This ASE experiment has resulted in imaging major flood events such as Diamantina flooding. In this particular flood event, we were able to capture pre-, peak-, and post-flood flow conditions of the river. We will present the data acquired using ASE as a new source of data that can be readily coupled with any existing flood model to improve flood prevention. The detection of floods when they appear and tracking their manifestation provide not only early warnings of potentially hazardous conditions, but also improved understanding of flood evolution.