1

10.4 GEOSTAR: A NEW PAYLOAD FOR GOES-R

Bjorn Lambrigtsen*, Shannon Brown, Todd Gaier, Pekka Kangaslahti, Alan Tanner, William WilsonJet Propulsion Laboratory/California Institute of Technology, Pasadena, California

Jeff Piepmeier, NASA Goddard Space Flight Center, Greenbelt, MarylandChris Ruf, University of Michigan, Ann Arbor, Michigan

* Corresponding author address: Bjorn Lambrigtsen, Jet Propulsion Laboratory, California Insttitute ofTechnology, 4800 Oak Grove Drive, Pasadena, CA 91109; e-mail: lambrigtsen@jpl.nasa.gov

ABSTRACT

The Geostationary Synthetic Thinned ApertureRadiometer (GeoSTAR) is a new concept for amicrowave sounder, intended to be deployed onNOAA’s next generation of geostationary weathersatellites, GOES-R – due to first launch in 2012. Thiswill fill a serious gap in our remote sensing capabilitiesof long standing – a key capability that NOAA is bookkeeping at the top of its list of “pre-planned productimprovements” for GOES-R – i.e. the most urgentlyneeded additional payload, which will be added assoon as funding has been allocated and programmaticissues resolved. Although real-aperture GEOmicrowave sounders have been proposed over theyears, only GeoSTAR is capable of meeting themeasurement requirements and is therefore now theleading candidate. A ground based prototype hasbeen developed at the Jet Propulsion Laboratory,under NASA Instrument Incubator Programsponsorship, and is currently undergoing tests andperformance characterization. Initial tests have beenvery successful, and images of the sun transitingthrough the field of view – the first successful imagingusing a 2D aperture synthesis system and in effectconstituting proof of concept – demonstrate that thesystem is very stable and that aperture synthesis is afeasible approach. The initial space version ofGeoSTAR will have performance characteristics similarto those of microwave sounders currently operating onpolar orbiting environmental satellites, but subsequentversions will significantly outperform those systems.In addition to all-weather temperature and humiditysoundings, GeoSTAR will provide continuous rainmapping, tropospheric wind profiling and real timestorm tracking. With the aperture synthesis approachused by GeoSTAR it is possible to achieve very highspatial resolutions without having to deploy theimpractically large parabolic reflector antenna that isrequired with the conventional approach. GeoSTARtherefore offers both a feasible way of getting amicrowave sounder with adequate spatial resolution inGEO as well as a clear upgrade path to meet futurerequirements. GeoSTAR offers a number of otheradvantages relative to real-aperture systems as well,such as 2D spatial coverage without mechanicalscanning, system robustness and fault tolerance,

operational flexibility, high quality beam formation, andopen ended performance expandability. Thetechnology and system design required for GeoSTARare rapidly maturing, and it is expected that a spacedemonstration mission can be developed before thefirst GOES-R launch. GeoSTAR will be ready foroperational deployment 2-3 years after that. Althoughthe GeoSTAR team has been closely collaboratingwith the NOAA Office of System Development, whichis responsible for overseeing the development ofGOES-R, there are programmatic barriers that make itdifficult for NOAA to develop new-technologypayloads. Traditionally this has been the role ofNASA, and both organizations are working on findingways to implement this “research to operations” modelwithout negatively impacting their other objectives.GeoSTAR is a good candidate for this model, and it isexpected to go forward as a space mission within thenext decade.

1. INTRODUCTION

The National Oceanic and Atmospheric Administration(NOAA) has for many years operated two weathersatellite systems, the Polar-orbiting OperationalEnvironmental Satellite system (POES), using low-earth orbiting (LEO) satellites, and the GeostationaryOperational Environmental Satellite system (GOES),using geostationary earth orbiting (GEO) satellites.Similar systems are also operated by other nations.The POES satellites have been equipped with bothinfrared (IR) and microwave (MW) atmosphericsounders, which together make it possible todetermine the vertical distribution of temperature andhumidity in the troposphere even under cloudyconditions. Such satellite observations have had asignificant impact on weather forecasting accuracy,especially in regions where in situ observations aresparse, such as in the southern oceans. In contrast,the GOES satellites have only been equipped with IRsounders, since it has not been feasible to build thelarge aperture system required to achieve sufficientspatial resolution for a MW sounder in GEO. As aresult, and since clouds are almost completely opaqueat infrared wavelengths, GOES soundings can only beobtained in cloud free areas and in the less importantupper atmosphere, above the cloud tops (i.e. lessimportant in a weather context). This has hindered the

2

effective use of GOES data in numerical weatherprediction. Full sounding capabilities with the GOESsystem is highly desirable because of the advanta-geous spatial and temporal coverage that is possiblefrom GEO. While POES satellites provide coverage inrelatively narrow swaths, and with a revisit time of 12-24 hours or more, GOES satellites can providecontinuous hemispheric or regional coverage, makingit possible to monitor highly dynamic phenomena suchas hurricanes. Such observations are also importantfor climate and atmospheric process studies.

In response to a 2002 NASA Research Announcementcalling for proposals to develop technology to enablenew observational capabilities from geostationaryorbits, the Geostationary Synthetic Thinned ApertureRadiometer (GeoSTAR) was proposed as a solution tothe GOES MW sounder problem. Based on a conceptfirst developed at the Jet Propulsion Laboratory in1998 and intended for the NASA New Millennium EO-3mission, GeoSTAR synthesizes a large aperture tomeasure the atmospheric parameters at microwavefrequencies with high spatial resolution from GEOwithout requiring the very large and massive dishantenna of a real-aperture system. With sponsorshipby the NASA Instrument Incubator Program (IIP), aneffort is currently under way at the Jet PropulsionLaboratory to develop the required technology anddemonstrate the feasibility of the synthetic apertureapproach – in the form of a small ground basedprototype. This is being done jointly with collaboratorsat the NASA Goddard Space Flight Center and theUniversity of Michigan and in consultation withpersonnel from the NOAA/NESDIS Office of SystemDevelopment. The objectives are to reduce technologyrisk for future space implementations as well as todemonstrate the measurement concept, testperformance, evaluate the calibration approach, andassess measurement accuracy. When this riskreduction effort is completed, a space basedGeoSTAR program can be initiated, which will for thefirst time provide MW temperature and water vaporsoundings as well as rain mapping from GEO, with thesame measurement accuracy and spatial resolution asis now available from LEO – i.e. 50 km or better fortemperature and 25 km or better for water vapor andrain. Furthermore, the GeoSTAR concept makes itfeasible to expand those capabilities without limit, tomeet future measurement needs.

The GeoSTAR prototype has now been completed,and tests are under way to assess its performance.Results so far are excellent, and this developmentcan now be characterized as proof of the aperturesynthesis concept. This constitutes a majorbreakthrough in remote sensing capabilities. Furthertechnology development is under way, both as riskreduction and to enhance the measurementcapabilities of the GeoSTAR system. At the sametime, efforts are also under way to identifysponsorship and secure funding for a spacedemonstration mission in the 2012-2015 time frame. I tis likely that a GeoSTAR mission will be a joint NASA-

NOAA undertaking, which increases the programmaticcomplexity, and many issues remain to be resolved.

2. FUNCTIONALITY AND REQUIREMENTS

In developing the GeoSTAR technology and prototypea notional space system performing at the same levelas the Advanced Microwave Sounding Unit (AMSU)system now operating on NASA and NOAA polar-orbiting LEO satellites was used for design and sizingpurposes. The notional operational GeoSTAR willprovide temperature soundings in the 50-60 GHz bandwith a horizontal spatial resolution of 50 km and watervapor soundings and rain mapping in the 183-GHzband with a spatial resolution of 25 km. A possiblethird band would operate in the 90-GHz window region(and would possibly also cover the 118-GHz oxygenline for additional information about clouds).Radiometric sensitivity will be better than 1 K in allchannels. These are considered to be the minimumperformance requirements, but the first spaceimplementation could be built to exceed this minimumperformance. It should be emphasized that it isnecessary to operate in these bands in order toprovide soundings to the surface. Others haveproposed building sounding systems operating athigher frequencies, where an adequate spatialresolution can be attained with a smaller aperture (andthus could be implemented as a real-aperture system),but since the opacity of a moist atmosphere increasessharply with frequency, as illustrated in Fig. 1, it is notpossible to reach the surface under all conditions atthose higher frequencies. In particular, moist andcloudy tropical conditions, such as encountered in thetropical cyclones that are of paramount interest, canonly be fully sounded in the 50- and 183-GHz bands.

With the notional GeoSTAR system, it will be possibleto produce temperature soundings within thetroposphere for most of the visible Earth disc (out toan incidence angle of 60° or more) with a 2-4 kmvertical resolution every 30 minutes and humiditysoundings with a vertical resolution of 3-4 km every 5-

Figure 1. Atmospheric transmittance in MW bands(From Grody in [1])

3

10 minutes. GeoSTAR is a non-scanning 2D imagingsystem. These soundings are obtained everywhere atthe same time – i.e. there is no time lag betweendifferent portions of the scene as there is in amechanically scanned system. That also makes thissystem ideal for derivation of wind profiles throughtracking of water vapor features – although thevertical resolution is limited. GeoSTAR producesseveral 2D “snapshots” every seconds. These imagesare combined over longer time periods to produce low-noise radiometric images that are then used forgeophysical “retrievals” or directly assimilated intoforecast models. It is also possible to recovertemporal information at a much higher resolution thanthe “averaging window” of 30 minutes (in the case oftemperature soundings), and this can be exploitedwhen rapidly evolving processes need to be resolvedmore precisely. The retrieval of vertical profiles oftemperature, water vapor density and liquid densityfrom spectrally sampled brightness temperatures iswell established (see, e.g., [1]), and such profiles areroutinely derived from AMSU observations. Thesemethods will also be used with GeoSTAR.

In addition to atmospheric profiles GeoSTAR will alsobe used to measure precipitation. There are twoapproaches for this, both depending on scattering.The first method is one developed by Ferraro andGrody [2], which uses window channels at 89 and 150GHz to measure the scattering caused by iceparticles formed in and above rain cells. This methodmay be adapted to use a 50-GHz window channel inlieu of 89 GHz. A second approach has beendeveloped by Chen and Staelin [3] and uses a numberof channels in the 50-GHz band and the 183-GHzband to derive precipitation estimates – this is alsobased on the ice scattering signature. Although thereare limitations with these methods (some stratiform

and warm rain conditions are problematic), GeoSTARoffers the advantage of continuous full-disc coverageand can therefore be used to fill in the gaps betweenthe narrow swaths of LEO-based systems. In addition,frozen precipitation (snow), which is difficult to detectby conventional means, can also be observed withthese methods. These capabilities will be used tocomplement the observations obtained from theplanned Global Precipitation Mission and itssuccessors.

3. INSTRUMENT CONCEPT

As illustrated schematically in Fig. 2, GeoSTARconsists of a Y-array of microwave receivers, wherethree densely packed linear arrays are offset 120°from each other. Each receiver is operated in I/Qheterodyne mode (i.e. each receiver generates both areal and an imaginary IF signal). All of the antennasare pointed in the same direction. A digital subsystemcomputes cross-correlations between the IF signals ofall receivers simultaneously, and complex cross-correlations are formed between all possible pairs ofantennas in the array. In the small-scale example ofFig. 2 there are 24 antennas and 276 complexcorrelations (=24*23/2). Accounting for conjugatesymmetry and redundant spacings, there are 384unique so-called uv-samples in this case. Eachcorrelator and antenna pair forms an interferometer,which measures a particular spatial harmonic of thebrightness temperature image across the field of view(FOV). The spatial harmonic depends on the spacingbetween the antennas and the wavelength of theradiation being measured. As a function of antennaspacing, the complex cross-correlation measured byan interferometer is called the visibility function. This2-dimansional function is essentially the Fouriertransform of the function of brightness temperature

15 12 9 6 3 0 3 6 9 12 15

18

15

12

9

6

3

3

6

9

12WindSTAR Array Configuration

X Offset (cm)

Y Offset (cm)

60 50 40 30 20 10 0 10 20 30 40 50 6060

50

40

30

20

10

0

10

20

30

40

50

60WindSTAR U,V Coverage

U Offset (wavelengths)

V Offset (wavelengths)

Arm #3 Arm #2

Arm #1

Figure 2. Antenna array (left) and uv sampling pattern (right), as implemented in the GeoSTAR prototype

4

versus the incidence and azimuth angles (or, directioncosines). By sampling the visibility over a range ofspacings and azimuth directions one can reconstruct,or “synthesize,” an image in a computer by discreteFourier transform. These techniques are well known inradio astronomy, but are relatively new to earthremote sensing problems. Thus, in Fig. 2, the leftpanel shows the distribution of receivers in theinstrument’s aperture plane, and the right panel showsthe resulting sampling points in spatial Fourier spacei.e. in terms of spatial harmonics).

The “Y” configuration of the GeoSTAR array ismotivated by the need to measure a complete set ofvisibility samples with a minimum number of antennas.In principle, one can measure the visibility functionwith just two antennas by mechanically varying theirspacing and orientation. But this is not practical forthe present application, and would require too muchobservation time for the sequential measurements.Instead, GeoSTAR uses a thinned (or “sparse”) arrayto simultaneously measure all the required spacingsfrom a fixed antenna geometry. There are many kindsof sparse arrays, and the “Y” array of Fig. 2 is one ofthe best in terms of efficient use of antennas and interms of the simplicity of the structure - which lendsitself well to a spaceborne deployment. As illustratedin Fig. 2, the spacings between the various antennapairs yield a uniform hexagonal grid of visibilitysamples. By radio astronomy convention, thespacings are called the “baselines,” with thedimensions “u” and “v.” The primary advantage of thesparse array is that it uses far less physical antennaaperture space than the comparable real aperture.

The smallest spacing of the sample grid in Fig. 2determines the unambiguous field of view, which forGeoSTAR must be larger than the earth disk diameterof 17.5° when viewed from GEO. This sets both theantenna spacing and diameter at about 3.5wavelengths, or 2.1 cm at 50 GHz, for example – asillustrated in Fig. 2. The longest baseline determinesthe smallest spatial scale that can be resolved, whichfor the array in Fig. 2 is about 0.9° (i.e. 17.5°/√384).To achieve a 50 km spatial resolution at 50 GHz, abaseline of about 4 meters is required. Thiscorresponds to approximately 100 receiving elementsper array arm, or a total of about 300 elements. Thisin turn results in about 30,000 unique baselines,60,000 uv sampling points (given conjugatesymmetry), and therefore 60,000 independent pixels inthe reconstructed brightness temperature image, eachwith an effective diameter of about 0.07° - about 45km from GEO.

4. PROTOTYPE

A small-scale prototype has been built to address themajor technical challenges facing GeoSTAR. Thesechallenges are centered around the issues of systemdesign and calibration. (Power consumption has alsobeen a major concern, but recent and continuingminiaturization of integrated circuit technology hasdemonstrated that this should no longer be seen as a

major issue.) Synthesis arrays are new and untestedin atmospheric remote sensing applications, and thecalibration poses many new problems, including thoseof stabilizing and/or characterizing the phase andamplitude response of the antenna patterns and of thereceivers and correlators. System requirements needto be better understood - and related to real hardware.To these ends the prototype was built with the samereceiver technology, antenna design, calibrationcircuitry, and signal processing schemes as areenvisioned for the spaceborne system. Only thenumber of antenna elements differ. Progress on thissystem has been rapid in recent months, and thefollowing discussion will attempt to emphasize themost recent achievements at the time of writing.

The prototype consists of a small array of 24elements operating with 4 AMSU channels between 50and 54 GHz. Fig. 3 shows a photo of the prototype.As illustrated in both Fig. 2 and Fig. 3, there is noreceiver at the exact center of the array as therewould be in a fully symmetric system. A centerfeedhorn poses a number of unnecessary complica-tions to the system, related to the physical package(there is not enough room) and the electrical design(to be discussed below). The solution is to remove theone horn from the center of the array, stagger thethree arms counter clockwise, and then bring themtogether so that the three innermost horns form anequilateral triangle. This staggered-Y configurationeliminates the need for an odd receiver at the center,which simplifies both mechanical and electronicdesign. The only penalty is a slight and negligible lossof visibility coverage.

A simplified diagram of the GeoSTAR prototype isgiven in Fig. 4. From left to right in Fig. 4 - or fromfront to back in Fig. 3 - the signal starts at the hornaperture with a vertical polarization (say), and thenpasses through a waveguide twist which aligns thewaveguide to the orientation of the 8-element arrayarm. Each of the three arms require different twists:the top two arms of Fig. 3 twist 60° in oppositedirections, and the bottom arm doesn’t twist at all.This results in all receivers detecting the same linear

Figure 3. GeoSTAR prototype

5

polarization, as is commonly required for sounderswith channels sensitive to surface radiation (which ispolarized). GeoSTAR is very sensitive to antennapattern differences among antennas, and a waveguidetwist proved to be the easiest solution to guarantee aprecise polarization match.

The signal in Fig. 4 then passes through a calibrationfeed which periodically injects a noise signal into allreceivers from a common noise diode source. Thissignal will be used as a reference to stabilize thesystem against gain, phase, and system noise drifts.The antenna signal passes into a MMIC receivermodule, where it is amplified and double-sidebanddownconverted in phase quadrature to two IF signalsof 100 MHz bandwidth.

The in-phase (I) and quadrature (Q) IF signals fromeach receiver are then digitized. For reasons ofproduct availability, the analog to digital converter ispresently an 8-bit device, but this will be replaced witha two-bit or possibly just a one-bit converter to savepower in a space system. The correlations onlyrequire 1-bit resolution (i.e. the sign bit), and the extrabits are only used to monitor changes in system noisetemperature. There is a multiplexer for each arm of thearray - the term “multiplexer” here refers to the factthat eight receivers are combined on a single digitalbus for transmission to the central correlator. In alarge space based system there would be amultiplexer for each grouping of 8 or 16 receivingelements. An FPGA is presently performing most ofthe functions of the multiplexer, which also includes“totalizers” that are used to count the occurrences ofeach ADC output state so that the threshold levelscan be compared with the known Gaussian statisticsof the IF voltage.

Perhaps the most important subsystem is thecorrelator, which must perform multiplications of all100-MHz signal pairs in real time. For a spaceborneoperational system with 100 elements per armdiscussed earlier, that requires on the order of 20trillion multiplications per second. To achieve such ahigh processing rate with a reasonable powerconsumption, the correlators are implemented as 1-bitdigital multiply-and-add circuits using a design

developed by the University of Michigan. 1-bitcorrelators are commonly used in radio astronomy.The correlator for the GeoSTAR prototype, where lowcost was more important than low power consumption,is implemented in FPGAs. An operational system willuse low-power application specific integrated circuits(ASICs). Current state of the art would then result in apower consumption of less than 20 W for the 300-element system discussed above, and per Moore’sLaw this will decline rapidly in future years.

5. EARLY TEST RESULTS

A number of tests have already been done in alaboratory setting, and the results are veryencouraging – no serious problems have beenidentified so far, and the system is working exactly asexpected, which is a remarkable achievement.Following the laboratory tests, the system was movedoutdoors to observe the sky and the sun. The arraywas pointed into the path of the sun, and the sun wasallowed to pass through the center of the field of viewat an elevation of 45°. Pictures of the basic setup areshown in Fig. 5. During these tests, the antennafixture was disturbed several times as different sunshields (made of pieces of cardboard) were arrangedon the structure to keep the receivers from

Figure 5. GeoSTAR prototype outdoors to observe thesun’s transit: a) Initial configuration (top panel); b) finalconfiguration (bottom panel)

Figure 4. Prototype configuration

6

overheating in the sun, as shown. This activityresulted in several interruptions in the observations,and it is likely that the uneven temperatures of theantennas resulted in uneven receiver noisetemperatures. Although none of the calibrationsubsystems were operated during this test (exceptthe LO phase switching), the results were spectacularand give us confidence that both system design andperformance will exceed expectations.

We have done some further analysis of the solar data,and have created a series of reconstructed (butuncalibrated) brightness temperature images. (Ananimation of the entire sequence also exists.) Theimages were reconstructed using the so-called G-matrix approach and accounts for the elementalantenna patterns. The most notable feature is ahexagonal-symmetric sidelobe pattern. Fig. 6 shows acolor coded image of this pattern (top panel), derivedfrom the observations when the sun was near thecenter of the field of view as well as a line plot along a

particular azimuth direction (bottom panel). The mostnotable feature here is that the pattern is indistin-guishable from the theoretical “sinc” function. Inparticular, note that the sidelobes are both positiveand negative. That makes it possible to apply simpleimage processing techniques to achieve an optimalbalance of image sharpness (i.e. spatial resolution)and beam efficiency (i.e. effective sidelobe level).Typically, this consists of applying a “window”, suchas a “box-car” or a Hamming window – i.e. techniquesthat are well known in image processing. This is oneof the most powerful features of an aperture synthesissystem such as GeoSTAR.

The imaging of the solar transit is the first successfuluse of a 2D aperture synthesis technique to generatemeaningful radiometric fields and can be taken as aproof of concept. This is a major breakthrough inremote sensing capabilities that will enable newmeasurements from space.

6. PROGRAMMATIC CONSIDERATIONS

The GeoSTAR team has worked closely withrepresentatives from the NOAA National Environ-mental Satellite Data and Information Service(NESDIS) Office of System Development (OSD) sincethe inception of the IIP prototyping project. NASAHeadquarters has also provided programmatic andscientific oversight. This has resulted in the GeoSTARdesign being closely aligned with “customer” needs.The measurements that GeoSTAR will provide areneeded by NOAA for operational use (i.e forassimilation into numerical weather predictionsystems) and are needed as well by NASA forresearch use (i.e. for investigations related to thehydrologic cycle). The requirements of NOAA NESDISare collected in the GOES-R Performance Require-ments Document (GPRD), where the need for amicrowave sounder is expressed in the form of one ofseveral Pre-Planned Product Improvements (P3I). TheMW sounder P3I is currently on top of the list, i.e.NOAA considers this to be their highest-priority unmetneed. Normally, as soon as funding for such apayload has been identified and the technologicalmaturity is sufficient for an operational mission, thepayload is elevated to baseline status. This has beendifficult to effect in the case of GeoSTAR, however,both because of the relatively low technologic maturitylevel and because of the difficulty of identifying thefunds required in the current budgetary environment.On the other hand, several independent assessmentshave concluded that the conventional approach (i.e.using a real-aperture system) will not satisfy therequirements, and GeoSTAR is now recognized as themost likely candidate to provide the missingfunctionality. The prototype is intended to retire someof the tall-pole technology risk, and results obtainedto date show that this has largely been accomplished.However, it still difficult for NOAA to fund the firstGeoSTAR space mission, and it generally looks toNASA to do that. On the NASA side, science researchmissions tend to have higher priority than pre-operational missions (as a GeoSTAR space

Figure 6. GeoSTAR raw spatial resolution element

7

demonstration might be), and it has therefore beendifficult also for NASA to identify funding forGeoSTAR. A promising mechanism is through the“Research-to-operations” path, which has beenidentified by both organizations as worthwhile. Here,NASA first develops the necessary technology,followed by a space “research” mission thatdemonstrates the capabilities and elevates thematurity to operational status – at which point themission is handed over to NOAA for operational use.NOAA would subsequently procure additional copiesof the payload to populate future platforms. AGeoSTAR space demonstration mission is now beingdiscussed between NASA and NOAA, and it is likelythat such a mission will be recommended. Anotherpossible path is as a NASA research mission, such asthrough the Earth System Science Pathfinder (ESSP)series. Those missions generally require fairly hightechnologic maturity, but the ongoing GeoSTARrelated technology development efforts may makesuch a mission plausible. A science driven GeoSTARmission has been discussed in some detail in a whitepaper provided in response to the RFI issued for theNRC Decadal Survey of the NASA Earth scienceprogram that is currently under way.

The GeoSTAR IIP project will come to an end in early2006, but as was alluded to above, several additionaltechnology development efforts are under way, fundedthrough the NASA Earth-Sun System TechnologyOffice (ESTO), which also sponsors the IIP program,as well as through internal JPL development funds. I tis expected that those efforts will continue, and theresult will be to retire additional technology risk. Inaddition, efforts are under way to assess the potentialbenefits of a GEO MW sounder, both in terms of theprojected forecast impact of operational use of theobservations and in terms of the scientific impact ofresearch enabled by the GeoSTAR measurements.

7. SUMMARY

The GeoSTAR concept and the related technologyhave been maturing rapidly. The recent test resultsamount in effect to proof of concept, and thisrepresents a major breakthrough in remote sensingcapabilities. The continuing efforts to develop thetechnology further will enhance the system’sperformance as well as retire technology risk, and it isanticipated that the concept will be mature enoughthat a space mission can be implemented in the 2012-2015 time frame. The only major obstacles remainingwill then be of a programmatic and budgetary nature.It is likely that those will be overcome, and so aGeoSTAR mission is likely within the next 10 years.This will add significantly to the nation’s remotesensing capabilities, and the GeoSTAR observationsare expected to have a significant forecast impactand will greatly benefit research related to thehydrologic cycle as well. In particular, the GeoSTARobservations will add much to our ability to observe,understand and predict severe storms, such ashurricanes.

The advantages of an synthetic aperture system overa real aperture system are significant. For example,error budget calculations based on simulationsindicate that a synthetic aperture system can beexpanded in size without unduly stressing the phasestability requirements. It is therefore well suited tomeet future needs as the spatial resolution ofnumerical weather prediction models increase. Anotheradvantage is that the GeoSTAR system does notrequire platform-disturbing mechanical scanning, andthere is no time lag between different portions of theimages, as there is in mechanically scanned real-aperture systems – where there can be a time lag ofas much as an hour between the start of scan at thenorthern limit of the Earth disk and the end of scan atthe southern limit. GeoSTAR thus produces truesynoptic soundings; no other sounder has thatcapability. An additional advantage is fault tolerance.It is easy to add redundancy in the correlator system.Also, if one receiver should fail, the result is simply aslight degradation in image detail – there are no gapsin the image. (The reader can easily verify that byconsidering the effect of removing one receiver in Fig.2.) This “graceful degradation” is in sharp contrastwith the catastrophic failure modes of a conventionalsystem, where the loss of one receiver will cause theloss of an entire sounding band.

REFERENCES

[1] Michael Janssen (ed.), Atmospheric RemoteSensing by Microwave Radiometry, Wiley, 1993

[2] R.R. Ferraro, F. Weng, N.C. Grody and L. Zhao,“Precipitation characteristics over land from the NOAA-15 AMSU Sensor”, Geophys. Res. Lett., 27, 2669-2672, 2000

[3] D. Staelin & F. Chen, “Precipitation observationsnear 54 and 183 GHz using the NOAA-15 satellite”,IEEE Trans. Geosci. Rem. Sens., 38, 2322-32 (2000)

[3] A. Camps et al., “The processing of hexagonallysampled signals with standard rectangular techniques:Application to 2-D large aperture synthesisinterferometric radiometers”, IEEE Trans. Geosci.Remote Sens., 35, 183-190 (1997)

ACKNOWLEDGMENTS

This work was carried out at the Jet PropulsionLaboratory, California Institute of Technology under acontract with the National Aeronautics and SpaceAdministration. The authors wish to acknowledge thesupport of Ramesh Kakar of NASA HQ, George Komarof NASA ESTO, and Shyam Bajpai, Gerry Dittbernerand Mike Crison of NOAA NESDIS OSD. Thanks alsogo to numerous technical contributors at JPL, GSFCand U. Michigan.