Adaptive Sky: From Instrument Pixels to a Sensor Web Gestalt

A variety of sensors have been developed and deployed to monitor the Earth, ranging from in situ seismographic networks to hyperspectral imaging instruments carried onboard NASA satellites. Despite an impressive collection of sensing assets, there is still much untapped potential, as evidenced by the limited number of studies that successfully employ high-resolution data from multiple instruments. Sensor webs offer the potential to go beyond simple data fusion by dynamically combining sensing assets into coordinated, multi-instrument observers of specific geophysical objects, phenomena, and processes. In this paper, we describe Adaptive Sky, an algorithm package for sensor webs developed through funding from the NASA Earth Science Technology Office under the Advanced Information Systems Technology program. Fundamentally, Adaptive Sky aims to relate the observations from one sensor at time t to the observations from another sensor at time t´, providing a “gestalt,” or unified, perspective that is more than the sum of its parts. A scenario involving the eruption of Bezymianny Volcano on the remote Kamchatka Peninsula on 14 October 2007 demonstrates conceptually how Adaptive Sky can be leveraged to create unprecedented spatio-temporal and phenomenological coverage of a complex geophysical event of interest, despite limitations inherent in the individual sensors.