Application of real time dynamic Sensor Web measurement techniques to maximize Aura/TES useful science return

The Sensor Web characterizes a future observing system concept that has the potential to significantly increase useful science data return. It accomplishes this by intelligently and dynamically reconfiguring the measurement and information processing states of its constituent sensor, computing, and storage nodes. To explore the potential benefits of Sensor Web observing strategies for the Earth science community, the Goddard Space Flight Center is developing software to demonstrate the value of real-time, collaborative, "intelligent data collection" between two formation flying spacecraft that comprise NASA\´s EOS \´\´A-Train" constellation: Aqua and Aura. After its planned Summer 2004 launch, Aura, with its Troposphere Emission Spectrometer (TES), is to be inserted into the constellation\´s orbital plane and "follow" 15 minutes behind Aqua with its Moderate Resolution Imaging Spectroradiometer (MODIS) instrument. MODIS is a nadir pointing multispectral imager, it is always "ON", and its 36 bands can be used to detect clouds. TES is a high-resolution infrared-imaging Fourier transform spectrometer designed primarily to study ozone in the lower atmosphere. Providing spectral coverage in the mid- to thermal IR bands, it is undesirable for TES to make measurements within cloud obscured fields of view. Of particular significance, TES is a pointable instrument: it can be commanded to image any target within 45deg of the local vertical. In this paper, we describe a high performance prototype software system to acquire and process direct readout MODIS data to a Level 2 cloud mask product in real time. Using a predefined list of desired targets, in conjunction with the real time MODIS cloud mask information, the software generates simulated commands and "point" TES to only those targets that are cloud free to demonstrate the value of intelligent data collection measurement techniques. To further refine where within the cloud free regions TES should point, we are also assess- ing the potential use of atmospheric chemistry models whose outputs may provide an additional set of criteria for high science value target selection. Similar "event-driven" and "model-driven" adaptive observation strategies and dynamic measurement techniques may be applied to future sensor Web Earth- and space-science missions