A new approach to atmospheric sounding in the microwave spectral region is under development at the Jet Propulsion Laboratory. Based on 2-D aperture synthesis, it is particularly well suited for deployment on geostationary satellites, such as the GOES weather satellites operated by NOAA. Microwave sounders, designed to measure vertical temperature and humidity profiles, have long been in use on polar-orbiting LEO weather satellites. The data from these instruments, the latest of which operate near 50 GHz for temperature soundings at 50 km spatial resolution and near 183 GHz for humidity soundings at 15 km resolution, have contributed measurably to the greatly improved accuracy of numerical weather prediction that has been achieved in recent years. However, although highly desired, it has not been feasible to build similar instruments for the GEO satellites, due to the very large antenna that would be required to achieve the same spatial resolution as from LEO. While the GOES satellites are equipped with infrared sounders, these instruments cannot penetrate clouds, and since a large fraction of the Earth is covered by clouds at any given time, the GOES IR sounders have had negligible impact on forecast accuracy.

The aperture synthesis approach employed by GeoSTAR solves this problem and offers a number of other advantages as well. It uses a sparse 2-D array of wide-beam microwave receivers to synthesize the required large aperture and produces complete 2-D radiometric images of the entire Earth disc in a number of spectral channels, with a “pixel” size equivalent to the required spatial resolution, and without any mechanical scanning. Temperature and humidity profiles can then be obtained from these brightness temperature images, just as is done with the LEO sounders, but continuously and without any coverage gaps. The high-frequency water vapor channels can also be used to estimate rain rates, and the same physical effect – scattering from ice particles above rain cells – can be used to detect and track hurricanes and other high intensity weather phenomena in near real time. The GeoSTAR system will provide continuous full-disk coverage, with an effective refresh time of less than 1 hour. Hurricane tracking and high intensity precipitation detection can be done with a refresh time of less than 10 minutes, with full-disk coverage.

Sponsored by the NASA Instrument Incubator Program, a ground based prototype is currently under development at the Jet Propulsion Laboratory, with contributions from investigators at the NASA Goddard Space Flight Center and the University of Michigan. The objective is to develop the required technology, test calibration subsystems, and demonstrate a working end-to-end system. This effort is being closely monitored by NOAA, and GeoSTAR is being considered for a possible deployment – as a so-called pre-planned product improvement sensor – on the next generation of GOES satellite systems, GOES-R. Current estimates of power and mass of a space based GeoSTAR are well within platform capabilities.

The present paper focuses on the GeoSTAR science requirements, the basis of measurements, the system concept and a high level description of GeoSTAR and the prototyping effort. The purpose of the GeoSTAR prototyping effort is to retire much of the technology risk, especially at the system level. Although others have been working on 2-D aperture synthesis systems for some years – notably European investigators, who are planning to launch the space based Soil Moisture and Ocean Salinity (SMOS) mission in 2007, none has yet been successfully demonstrated. It is expected that the GeoSTAR prototyping effort will culminate in such a demonstration on the ground in 2005. A space based GeoSTAR effort can then be initiated soon afterwards and would be ready for launch in the GOES-R time frame or earlier.

GeoSTAR, while a radical departure from conventional sensors, represents the most promising and realistic approach to providing a microwave sounding capability in GEO and thus fill a notable gap in the nation’s remote sensing capabilities. While alternative solutions have been proposed, they are based on using large real-aperture dish antennas to achieve the required spatial resolution at the 50-GHz temperature sounding frequencies – with inherent obstacles related to size, mass, spatial coverage and refresh time, or they are based on using higher frequency bands that require smaller antennas to achieve high spatial resolution – but with inherent obstacles related to high atmospheric opacity at those high frequencies that prevent soundings through moist tropical atmospheres.