Adaptive Sky: A Feature Correspondence Toolbox for a Multi-Instrument, Multi-Platform Distributed Cloud Monitoring Sensor Web

The current suite of spaceborne and in-situ assets, including those deployed by NASA, NOAA, and other groups, provides distributed sensing of the Earth's atmosphere, oceans, and land masses. As part of an activity supported through NASA's Earth Science Technology Office (ESTO), we have developed techniques that enable such assets to be dynamically combined to form sensor webs that can respond quickly to short-lived events and provide rich multi-modal observations of objects, such as clouds, that are evolving in space and time. A key focus of this work involves relating the observations made by one instrument to the observations made by another instrument. We have applied approaches derived from data mining, computer vision, and machine learning to automatically establish correspondence between different sets of observations. We will describe a number of Earth science scenarios that were used to direct this development and which have benefited from the approach.