Special Issue on RISAT-1, Current Science, 25 March 2013 - Space ...

Radar Imaging Satellite-1With the launch of Radar ImagingSatellite-1 (RISAT-1), a new chapterhas been opened in Indian SpaceResearch Organisation’s (ISRO), EarthObservation (EO) programme – indigenouscapability of developing,launching and operation of microwaveimaging sensor. A set of fivearticles, in this special issue, tracesall aspects of RISAT-1 from developingthe SAR sensor to illustratingdifferent possibilities of utilizing itsdata in day to day practical applications.In the Foreword, A. S. Kirankumar(page 444), one of the primearchitects of ISRO’s EO programme,enunciates the significance of RISAT-1from the overall perspective ofISRO’s EO mission, especially howit complements existing array of sensorsin optical bands and its additionalstrengths which are unique tomicrowave bands and its ability tomeasure all the components of electromagneticwaves, i.e. polarization,amplitude and phase.RISAT-1 Synthetic Aperture Radar(SAR) payload is a complex payloadwith a very large array of imagingcapabilities in-built. It is not onlyISRO’s first SAR payload in space,but also India’s first state-of-the-artactive antenna in space. Further, forthe first time, it is carrying the capabilityof all traditional SAR imagingmodes in optional hybrid polarimetryconfiguration. Design philosophybehind RISAT-1 SAR, its realization,illustration of initial results and itscalibration have been brought out inthe article by Tapan Misra et al.(page 446).In this issueRISAT-1 SAR is the heaviest payloadbuilt in ISRO, weighing close to950 kg. The challenge was to accommodateit in a satellite bus which canbe flown in PSLV. The large SARantenna provides obstruction to viewinggeometry of satellite sensors anddata transmission systems. Further,designers wanted a simplified antennadeployment system which guaranteessuccess. Moreover, spacecraft resourceshave to cater to large powerdemand (~5 kW) and high data rate(~1.5 Gbps) for SAR operation. Allthese requirements resulted into anew and unique bus configuration,distinct from ISRO’s traditionalworkhorses like IRS and INSAT.The complete gamut of RISAT-1 satelliteconfiguration is presented byN. Valarmathi et al. (page 462).RISAT-1 is also the heaviest satellite(dry mass wise) built in ISRO,bordering on the outer limit of thelaunch capability of highest versionof PSLV. Specific PSLV XLlauncher configuration and its resultantperformances, very close to theintended ones, are presented in thearticle by P. Kunhikrishnan et al.(page 472).RISAT-1 has a very large numberof flexible imaging capabilites, operatedtransparently by on-board computer.For its seamless operation,complex mission planning and operationwere built in ground control ofspacecraft from ISTRAC, Bangalore.Further, ISRO’s data reception systemat NRSC, Hyderabad had to beupgraded by six folds from 110 to640 Mbps for RISAT-1 operation.The data processing, archiving anddissemination system for RISAT-1had to be integrated in new IMGEOSfacility of NRSC. The heart of theground segment is a complex SARsignal processor, implemented inboth off line and near real time processingsystems at NRSC, to convertnoise like SAR signals to meaningfuldigital images with all the correctionsand map projections for disseminationto users in user friendlyand universally acceptable data formats.All the facets of Ground Segment,as these activities are referredto in ISRO parlance, are presented inthe article by V. Mahadevan et al.(page 477).Ultimate aim of RISAT-1 programmeis to convert digital imagesto meaningful, user-specific geophysical,ecological and myriad otherapplication information so thatstrength of RISAT-1 data is harnessedby governmental agencies,commercial entities, global users, resourcescientists and general public.During the course of spacecraft’slifetime and beyond, there is a possibilityof getting the data used formany different applications, limitedby basic physics behind the radarsignal, its interaction with earth elementsand ingenuity of resourcescientists. Very initial results, in thisendeavour, are presented in the articleby Manab Chakraborty et al.(page 490).402CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1ForewordSignificance of RISAT-1 in ISRO’s EarthObservation ProgrammeA. S. Kiran Kumar*Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, IndiaThe Indian Space Research Organisation (ISRO) beganits space-based Earth Observation Programme in an experimentalmode with Bhaskara-1 in 1979, from a low earthorbiting platform. Operational Earth observation activitiesfrom a geostationary platform began with the IndianNational Satellite (INSAT) series (since 1982) with VeryHigh Resolution Radiometer (VHRR) payload. TheIndian Remote Sensing (IRS) satellite series, since 1988,has been providing Earth observation data in a wide rangeof spatial, spectral, radiometric and temporal resolutionsusing push-broom technology optical imaging sensors.The data from these optical sensors have contributed tovarious application activities encompassing resourcemonitoring, agricultural crop-yield forecasting, prospectivefishing-zone identification, cyclone monitoring andtracking, weather monitoring and forecasting, cartographyand disaster management, etc.However, space-based Earth observation in the opticaldomain is constrained by cloud cover and during monsoonperiod over the Indian region it becomes a severelimitation. This significantly restricts the capability toprovide useful inputs for resource and disaster monitoringfrom space. Microwave imaging sensors – by virtue oftheir ability to see through the clouds – can overcome thisconstraint and enable establishing an effective Earthobservation capability from space.The launch of Radar Imaging Satellite-1 (RISAT-1) on26 April 2012 with a C-band Synthetic Aperture Radar(SAR) on-board, marks the initiation of a new class ofEarth observation imaging products and services. RISAT-1, the indigenous space-based radar imaging mission ofISRO, is capable of observing the Earth at any time of theday as it carries its own source of illumination.Imaging products of SAR are complementary to opticalimaging, avoiding information duplication. In other words,while optical sensors give information regarding frequencyselective absorption and reflection properties, SAR signalsare sensitive to structural shapes and dielectric propertiesof the objects being imaged. SAR hardware can be configuredfor exploring object sensitivity to polarizationdiversity and phase angle. Coupled with this sensitivity*e-mail: kiran@sac.isro.gov.in444and being transparent to cloud cover, SAR images findapplications in crop assessment during monsoon, imagingin perennially cloud-covered regions, disaster management,soil moisture, forestry, geology, oceanography,bathymetry, etc.Another notable advantage of SAR technology is thatthe resolution of the system is independent of height orrange, limited only by available transmitted power andconsequent signal-to-noise ratio. Because of the natureof processing, geometric accuracy of the images is notaffected by angular accuracy of the satellite and is limitedonly by the knowledge of orbital positional accuracy ofthe satellite and Digital Elevation Model resulting inrepeatable location accuracy of the images.RISAT-1 has been built with state-of-the-art technologyand is endowed with many SAR imaging modes likeconventional stripmap, high-resolution sliding spotlight,wide swath scanSAR, etc. It can be operated in single ordual polarization or in quad polarization modes, providingimaging ability from 1 m to 50 m resolution over 10 kmswath to 225 km swath.Its configuration of dedicating separate set of poweramplifiers for V and H polarization transmission, has madeit a unique spaceborne Hybrid Polarimetric Sensor. Theother operational spaceborne SARs like on Radarsat-2,TerraSAR-X or CosmoSkymed, are equipped with specificlinear polarimetric mode which is usually operatedwithin the restricted coverage of 20° to 30° incidenceangle, because of doubling of pulse repetition frequency(PRF) and usually a specific imaging mode is dedicatedfor linear polarimetric operation. However, in the hybridpolarimetric operation of RISAT-1, signal is transmittedin circular polarization and the received signal is digitizedin two orthogonal polarization chains. This ensuresconventional PRF of operation without any increase indata rate. Hybrid polarization in RISAT-1 can be activatedfor any imaging mode (spotlight/stripmap/scan-SAR) and can be operated over any incidence angleranging from 12° to 55°. Its dual polarized antennaelements are fed by an array of transmit–receive (TR)modules, controlled by more than 300 processors andpowered by miniaturized pulsed electrical power conditioners.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1RISAT-1 comprises around 1400 subsystems, including300 processors. The active array subsystems are largein number and less on design variety. Each of the subsystemsrequires rigorous space grade fabrication and qualification.Fabrication and characterization of each of thesesubsystems are typically spread over 5–6 weeks. Industrialproduction and space qualification of the subsystemswere carried out by the Indian industry based on in-housedesigns of ISRO. These industries had limited exposureto space-grade electronics and therefore in the spirit ofpartnership, they had to undergo a rigorous regime oftraining in space-grade fabrication processes, qualificationmethods and documentation processes. This alsohelped in the development of indigenous source of RFMonolithic Microwave Integrated Circuits (MMICs), TRmodules, Application Specific Integrated Circuits(ASICs), miniaturized power supply and printed antennaarray. RISAT-1 effectively acted as a catalyst in expandingthe indigenous industrial base for production ofspace-grade SAR subsystems.ISRO used its in-house pool of ingenuity in conceptualizing,engineering and realizing the SAR systemof RISAT-1, which is a vastly complex payload withsignificant level of flexibility in reconfiguration to meetdifferent imaging requirements and ease of operability.This was possible because of large on-board softwarespread over 300 processors. The characterizationof the system itself was unique, where all the 126beams have been characterized with precision. Thisresulted in calibration and quick operationalization of thesystem.Realization of state-of-the-art radar imaging satelliteRISAT-1 needed significant developments in the spacecraftcapabilities to accommodate large weight, powerand transmission data rates. For example, the data rate oftransmission was increased six fold from 110 to 640 Mbits.Though weighing 1858 kg, RISAT-1 is heaviest amongISRO’s remote sensing satellites, it is the lightest satellitecompared to those belonging to the same class.Data products from RISAT-1 have already been releasedto users from 19 October 2012. RISAT-1 imagingproducts are expected to enhance the application potentialof SAR data not only in India, but also globally in manyimportant resource applications and disaster managementsituations. Radio Detection and Ranging (RADAR) datafrom space platforms have already made a significantmark world over because of the ability of the radars tomake observations during the day or night, look throughcloud cover and achieve resolution and observe detailsthat are difficult for optical and infrared sensors. Manyoperational modes and hybrid polarimetric capability ofRISAT-1 are expected to open up newer avenues, as itprovides many more observable parameters like amplitude,phase and state of polarization enabling many newscientific studies leading to diverse and novel applicationsusing microwave data.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 445

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Synthetic Aperture Radar payload on-boardRISAT-1: configuration, technology andperformanceTapan Misra 1, *, S. S. Rana 2 , N. M. Desai 1 , D. B. Dave 1 , Rajeevjyoti 1 ,R. K. Arora 1 , C. V. N. Rao 1 , B. V. Bakori 1 , R. Neelakantan 2 and J. G. Vachchani 11 Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, India2 Formerly with Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, IndiaThe launch of Radar Imaging Satellite (RISAT-1)marked a new chapter in the remote sensing programmeof Indian Space Research Organisation(ISRO). RISAT-1, carrying a multi-mode SyntheticAperture Radar (SAR) system, will provide complementaryimaging capability in microwave along withoptical images, being obtained from the well-establishedIRS class of satellites. RISAT-1 supports a variety ofresolution and swath requirements. Both conventionalstripmap and scanSAR modes are supported, withdual polarization mode of operation. Additionally aquad polarization stripmap mode is provided foravailing additional resource classification. In all thesemodes, resolutions from 3 to 50 m can be achievedwith swath ranging from 25 to 223 km. On experimentalbasis, a sliding spotlight mode is also available. In allthe imaging modes, a novel polarimetry mode calledcircular or hybrid polarimetry can be exercised seamlessly.The system is capable of imaging on either sideof the flight track depending upon prior programmingof the satellite. The satellite is placed in a Sun-synchronousorbit with 6 am–6 pm equatorial crossing. Thisorbit configuration is chosen to maximize solar poweravailability. The satellite has an on-board-solid staterecorder for supporting data acquisition beyondground station visibility. The payload is based on activeantenna array technology. Crucial technologyelements like C-band MMICs, TR module and miniaturizedpower supplies have already been developed inIndia. A pulsed mode near-field test facility has alsobeen developed in-house in order to characterize thepayload in the integration laboratory itself.Keywords: C-band, near field antenna facility, radarimaging satellite, stripmap mode, scanSAR mode, SyntheticAperture Radar, transmit–receive module.Introduction*For correspondence. (e-mail: misratapan@sac.isro.gov.in)446SINCE the launch of SeaSAT mission in 1978, there was aspurt of operational and experimental spaceborne SyntheticAperture Radar (SAR) missions by major spaceagencies. The major SAR missions carried out over thelast three decades 1–13 have covered significant grounds inSAR applications from L to X bands. SAR system configurationalso graduated from single-beam fixed resolutionsystem to multiple resolution and swath systems. Consequently,SAR technology also witnessed transformationfrom passive antenna-based system to active antenna/phased array-based systems (e.g. PALSAR, Radarsat-1and -2, TerraSAR-X1, etc.). Growing application needshave brought about significant changes in SAR systems,over the period, from single polarization configuration tomulti-polarization/polarimetric configuration.Along with the presently orbiting spaceborne SAR systems(Table 1), Radar Imaging Satellite-1 (RISAT-1),which has been launched on 26 April 2012, is available tothe international SAR community and researchers as asource of multi-resolution/multi-swath/multi-polarizationSAR data. RISAT-1 is a new class of satellite, dedicatedto imaging in the microwave band. Its primary aim is tocomplement and supplement operationally optical imagingsystems, being flown in its established Indian RemoteSensing (IRS) class of satellites. RISAT-1 carries a multimodeSAR satellite in C-band as the sole payload.Mission objectivesThe primary applications envisaged for RISAT-1 are: (a)Agriculture monitoring, mainly during monsoon seasonand (b) flood mapping, as part of the National DisasterManagement Programme. Agricultural monitoring, basedon Radarsat-1 and -2 data, has been made operational inIndia. RISAT-1 is designed to maintain continuity of serviceand improve temporal sampling. Hence, C-band wasselected as the operating frequency. For monitoringgrowth cycle of paddy, the principal crop cultivated duringmonsoon season, a revisit period of 25 days, with thesame incidence angle reference, was decided upon. Consequently,two scanSAR swath modes of 223 km swath/50 m resolution and 115 km swath/25 m resolution wereselected. Flood mapping and monitoring require bothCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 1.Major international spaceborne SAR missions (currently operational)Frequency band,SAR system Operational from resolution (azimuth × range) SwathRadarsat-1 November 1995 C-band, 9–100 m 45–510 kmRadarsat-2 December 2007 C-band, 3–100 m 10–500 kmTerraSAR-X1 June 2007 X-band, Spotlight 1: 5 km × 10 km,Spotlight 1: 1 m × 1.2 mSpotlight 2: 10 km × 10 km,Spotlight 2: 2 m × 1.2 mStripmap: 40 km,Stripmap: 3.2 m ×

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 2.RISAT-1 image quality parametersSwath coverageSelectable within 107–659 km off-nadir distance on either sideIncidence angle coverage 12°–55°Image qualityModePolarizationSingle pol Dual pol Circular (hybrid) Quad polHH/HV/VV/VH HH + HV/VV + VH polarimetry HH + HV + VV + VHTX: CPRx: V and H(experimental)HRSFRS-11 m (azimuth) × 0.67 m (range) resolution,10 × 10 km (10 × 100 km experimental) spot,Min σ 0 = –16 dB3 m (azimuth) × 2 m (range) resolution,25 km swath,Min σ 0 = –17 dBFRS-2 3 m (azimuth) × 4 m 9 m (azimuth) × 4 m(range) resolution, (range) resolution,25 km swath, 25 km swath,Min σ 0 = –19 dB σ 0 = –20 dBMRSCRS21–23 m (azimuth) × 8 m (range) resolution,115 km swath,Min σ o = –17 dB41–55 m (azimuth) × 8 m (range) resolution,223 km swath,Min σ 0 = –17 dBrestricted over 550 km distance starting at a stand-off distanceof 107 km (Figures 1 and 2).RISAT-1 satellite448Figure 2.Imaging geometry of RISAT-1.The hybrid polarimetry mode does not require doublingof pulse repetition frequency (PRF) as in the case of linearpolarimetry. So unlike linear polarimetry mode whereincidence angle coverage is very narrow to accommodatedata collection with doubled PRF, hybrid polarimetrymode can be operated seamlessly over all incidenceangles and all resolution-swath modes as PRF remainsunchanged.The SAR can image on either side of the track by rolltilting of the spacecraft. However, imaging option ofeither on left or right side is operationally kept fixed inany particular orbit. On either side, imaging area isThe RISAT-1 spacecraft has been built around the SARpayload in order to optimize the spacecraft weight andstructure. RISAT-1 satellite, in fully deployed configuration,is shown in Figure 3. The prism shape of the satelliteallows stowing of the active antenna, in three foldsaround the prism structure. The prism structure is builtaround a central cylinder. Most of the spacecraft subsystemsand the complete payload are integrated in the prismstructure and the central cylinder. The solar panel and therest of the spacecraft subsystems are mounted on thecuboid portion of the RISAT-1 satellite, so that the deployedantenna does not cast its shadow on the solar panels.In-orbit mass of the satellite is around 1858 kg, of whichthe SAR payload contributes around 950 kg. Two solarpanels with high efficiency multi-junction solar cells,charge Ni–H2 batteries with 70 Ah capacity. The batterysizing is done such that it is possible to operate maximumof 10 min duration in each orbit. The satellite has thecapability of storing up to 240 Gbits of data in the solidstate recorder (SSR) on-board. The on-board data transmittercan transmit with a maximum data rate of640 Mbits/sec in X-band, on two polarizations (RHC andLHC) on the same X-band carrier.In the non-operating condition, the active antennapoints towards nadir. Prior to operation, the spacecraftCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 3. Artistic and actual view of RISAT-1 with antenna in deployed and stowed condition.Table 3.Salient features of RISAT-1 spacecraftOrbitCircular polar sun synchronousOrbit altitude536 kmOrbit inclination 97.552°Orbit period95.49 minNo. of orbits per day 14Equator crossing6.00 a.m./6.00 p.m.Spacecraft height3.85 mMass1858 kgPowerSolar array generating 2200 W andone 70 AH Ni-H2 batteryMax power handling capacity 4.3 kWData rate2 × 160 Mbps (total 640 Mbps intwo chains)SSR240 Gbits (End of Life)TT&CS-bandPayload down linkX-bandPower70 V bus/42 V busPointing accuracy 0.05°Drift rate5.0 × 10 –5 °/secAttitude knowledge 0.02°will be roll-tilted by ±36° to enable viewing either theright or left side of the flight track. Additionally, thesatellite has a capability of pitch steering up to ± 13° foroperation in HRS mode. The satellite also has yaw steeringcapability in order to minimize the earth rotationeffects on the SAR signal. In fact, yaw steering implementedon-board ensures Doppler centroid within± 100 Hz. RISAT-1 operates in a sun-synchronous orbitat an altitude of 536 km with revisit period of 25 days forMRS mode. Equator crossing time is kept at 6 a.m.(descending)–6 p.m (ascending). in order to minimizeeclipse period to the extent of maximum of 20 min duringsummer only, that too in the southern hemisphere, and toensure maximization of battery charging period. Salientfeatures of RISAT-1 satellite are presented in Table 3.SAR system design considerationsThe basic goal of SAR system design was to provide 3 mresolution capability as well as 115 km swath with coarseresolution of 25 m. Also fine resolution requirement of3 m, in stripmap mode, restricted the antenna length to6 m. Further imaging capability from 12° to 55° incidenceangle, with the above modes with basic beam illuminationrequirement over 25 km swath per beam,restricted the elevation width to 2 m. The limitation inwidth of the antenna was also dictated by accommodationenvelop of PSLV. The 1 : 3 ratio of antenna dimensionsalso resulted in basic prism-shaped architecture ofRISAT-1 spacecraft with simplified deployment mechanism.These two basic imaging modes were compatiblewith the existing capability of data transmission rate of640 Mbps with 4 BAQ (Block Adaptive Quantization)operation. ScanSAR mode of operation dictated a choiceof either phased array or active array architecture. Thelatter though complex, was selected mainly from considerationof availability of existing technology in India.Power sizing of transmit–receive (TR) modules in theactive antenna was carried out from the prime considerationof meeting the user requirement of minimum noiseequivalent σ 0 ~ –18 dB.The SAR system was conceived to provide simultaneouslyco- and cross-polarization response and systemarchitecture was drawn accordingly. A quad polarizationmode (FRS-2) was added after taking advantage of thedual polarized configuration of the SAR system.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 449

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 4.Hardware specification of RISAT-1–SARFrequencyAntenna typeAntenna sizeSide lobe levelNo. of TR modulesPulse widthAverage DC input power5.350 GHzPrinted antenna6 m (along flight) × 2 m (cross flight)–13 dB (azimuth), –13 dB (elevation)288, each with 10 W peak power20/10 μs1.8–3.9 kWHRS FRS-1 FRS-2 MRS/CRSChirp bandwidth (MHz) 225 75 37.5 18.75Sampling rate (MHz) 250 83.3 41.67 20.83PRF (Hz) 3500 ± 200 3000 ± 200 3000 ± 200 3000 ± 200Quantization 2/3 bit BAQ 2/3/4/5/6 bit BAQMaximum data rate 739 Mbits/sec 556 Mbits/sec 564 Mbits/sec 142 Mbits/sec@ 3 bit BAQ for HRS (single polarization) (single polarization) (single polarization)@ 6 bit BAQ for rest of the modes 1478 Mbits/sec 1112 Mbits/sec 284 Mbits/sec(dual polarization) (dual polarization) (dual polarization)Figure 4.Simplified schematic of RISAT-1 SAR system configuration.RISAT-1 SAR architecture was unique in the sense forH and V polarization transmission, separate TR modules,with independent transmit and receive chains, were provided.In other SAR systems of similar active array architecture(e.g. TerraSAR-X, Radarsat-2), one Solid StatePower Amplifier (SSPA) is switched between V and Htransmission. This advantage of architecture was harnessedto add hybrid polarimetric measurement as a separatefeature.Though the main aim of RISAT-1 is to provide SARdata to users operationally, purely from technology experimentationconsideration, sliding spotlight mode (HRS)was introduced in the configuration. Introduction of HRSmode called for substantial increase in system bandwidthand data rate.SAR system on-board RISAT-1450The SAR system, on-board RISAT-1, is configured on adual receiver concept (Figure 4) providing identical resolutionand swath in both simultaneous operation inco- and cross polarization. Single feeder SSPA is sharedbetween the two linear polarization channels for maintainingcommon amplitude and phase reference. Majorspecifications of the SAR system are summarized inTable 4. The SAR system consists of two broad segments,namely (i) deployable active antenna and (ii) RFand baseband systems housed on the satellite deck.Deployable active antennaThe Earth-facing side of the active antenna is a broadbanddual polarized microstrip radiating aperture. The activeantenna system consists of three deployable panels,each of 2 m × 2 m size. Each of the panels is subdividedinto four tiles of 1 m × 1 m size (Figure 5). Each tile consistsof 24 dual polarized linear arrays, aligned along theazimuth direction. Each of the linear arrays, of length1 m, is basically composed of 20 equispaced microstrippatches, EM coupled by two orthogonal strip lineCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 5.Organization of RISAT-1 antenna with detailed view of a tile.Figure 6. Typical configuration of microstrip patch used in RISAT-1.networks (Figure 6). Each of these linear arrays is fed byfunctionally two separate TR modules feeding two separatedistribution networks for V and H operation with thesame radiating patches. The outer duroid layer also doublesup as a radome and the patches are printed on theinner side of the outer duroid layer. A glass-wool blanketon the antenna isolates it from heating by the Earth aswell as solar radiation or from cooling in the absence ofsolar radiation, when the antenna points away from solarillumination.The printed antenna is grown on one side of a CarbonFibre Reinforced Plastic (CFRP) honeycomb plate. Therest of the active antenna electronics is mounted on theother side of this plate. Fast beam switching and beamwidthcontrol is achieved by electronic elevation beamcontrol in the active antenna. Sixty-one beam-pointingpositions have been identified to enable imaging anywhereover 550 km region on one side of the sub-satellitetrack, with the best possible performance. Each beam iscentred at off-nadir intervals of ~ 9 km. Two additionalbeams with no pointing (0° with respect to antenna orientationangle, i.e. ± 36°) are defined for two halves of theantenna, 6 m × 1 m each. Therefore, there are 63 beampositions defined for imaging on each side of the subsatellitetrack. As a result, a total of 126 beams are usedfor imaging on either side of the track. Option of yaw rotationfor left–right imaging would have reduced the requirementof the number of beams by half. But operationally,this option would have an implication on the time forswitching to imaging on either side of the track.The active beam-width in elevation direction is controlledsuch that for each beam a 25 km swath with nearidentical σ 0 performance is achieved irrespective of theelevation pointing. Typical σ 0 performance over differentoff-nadir distances is shown in Figure 7. TR-modules areswitched off in the width direction, equally from theouter edges of the two adjacent tiles to control elevationbeam-width between 2.48° and 1.67°. Such a strategy hasbeen adopted for elevation beam control for easing outthermal management.The peak RF power, fed by each TR module, is 10 Wat a duty cycle of ~ 7%. These two functionally separateTR modules are housed in two different physical enclosures,sharing the same power supply and TR controlCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 451

SPECIAL SECTION: RADAR IMAGING SATELLITE-1electronics (TRC). Basic architecture of TR module isshown in Figure 8. Phase and amplitude control of the TRmodule is achieved by 6 bit phase shifter and 6 bit attenuator,which in turn are shared by both transmit andreceive paths. Each of the TR modules is extensivelycharacterized over ambient temperature from –10°C to60°C. Typical temperature-dependent characteristics of aTR module are shown in Figure 9. The low noise amplifier(LNA) of the TR module is protected by a PIN diodeswitch. At the circulator output a coupler provides the requiredcalibration stimuli. On the tile, two rows of TRmodules, each consisting of 12 modules, feed alternateantenna arrays.Both the TR modules (H and V) and TRC are poweredby a miniaturized pulsed Electronic Power Conditioner(EPC) called Power Conditioning and Processing Unit(PCPU). Application Specific Integrated Circuit (ASIC)-based TRC controls both H and V TR modules andPCPU. Power sequencing is such that both transmit andreceive paths are switched on by power pulsing only, forthe required duration in every PRI (pulse repetition interval),in order to conserve power. It not only sequences theFigure 7. Minimum σ 0 performance over the swath for FRS-1 modeoperation.smooth operation of TR modules, but also provides requisitetemperature compensation of phase and amplitudevariation from stored characterization table. A thermistorvoltage from the TR modules provides the requisite inputfor appropriate reading of LUT (look-up table).All the 24 TR modules on the tile are controlled by oneTile Control Unit (TCU). It provides synchronization ofTR modules with a master reference. It also providesrequisite amplitude and phase correction required on eachTR module for appropriate collimation for a particularbeam pointing and pattern weighting. No weighting ispossible to be provided during transmission as all the TRmodules operate in saturation condition. Only on reception,is the weighting applied.The RF pulse from the feeder SSPA is distributed tothree panels where in turn they are again distributed tofour tiles. In each tile, the signal is fed into two sets ofTR modules via two 1 : 12 RF distribution networks(DNs). The two sets of TR modules, grouped on each sideof the tile, feed alternate rows of linear arrays. Considerableemphasis has been laid in matching group delaythrough different signal distribution paths, to the tune of10 ps, in order to get the optimal frequency response overthe signal bandwidth. A glimpse of the active antennasubsystems developed and qualified at ISRO is providedin Figure 10.An extensive on-board calibration facility is providedwith the help of a set of CAL switches and dedicateddistribution networks, for calibrating transmit and receivepaths of each of the TR modules separately. This switchmatrix also provides calibration of the feeder SSPA andreceiver path, bypassing the active antenna.The peak power from the active antenna is 2880 kWand it consumes up to 3.1 kW DC power during SARoperation. Thermal management of the tiles under such alarge dissipation during SAR operation is of prime importance.A set of four heat pipes, embedded in the CFRPhoneycomb substrate of the tile, carries away heat fromthe base of the TR modules and PCPU to the OpticalSolar Reflector (OSR). The OSR is pasted on the edge ofthe printed patch antenna. The back side of all the tiles iscovered with a Multi Layer Insulation (MLI) blanket forthermal isolation from free space or from the Sun dependingupon the antenna position. During SAR operation, thetemperature of the active antenna will be maintainedbetween 0°C and 50°C.Figure 8.Block schematic of a TR module for RISAT-1.RF and baseband subsystemsThe RF and baseband subsystems are housed on the satellitedeck just behind the static antenna panel. The blockschematic of the same is shown in Figure 11. Two separatechains of receiver (Rx) and Data Acquisition andCompression System (DACS) cater to simultaneousoperation in two polarizations. However, feeder SSPA,frequency generator (FG) and digital chirp generator452CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 9.Typical temperature-dependent characteristics of TR module.Figure 10. Active antenna components developed for RISAT-1 indigenously.a, Dual Polarised Antenna; b, MMICs; c, TR module; d,Miniaturized PCPU and e, TR control.(DCG) are common to both the polarization chains. Allthe subsystems are configured with 100% redundancy.Feeder SSPA transmits a chirped pulse of 20 μsec to theactive antenna during transmission duration. However, incircular polarimetry mode, the pulse width is reduced to10 μsec to maintain power balance as in this case both theSSPAs in V and H TR modules are active. Flexible DCGgenerates expanded pulses of four different bandwidths of225/9, 75/9, 37.5/9 and 18.75/9 MHz for operation invarious imaging modes. They are I–Q modulated atintermediate frequency (IF) of 500 MHz and frequencymultiplied nine times in FG to final RF frequency of4.5 GHz with chirp bandwidths 225, 75, 37.5 and18.75 MHz respectively. Finally, the 4.5 GHz chirpedcarrier is up-converted with an IF of 850 MHz to a finalcarrier of 5.35 GHz.The combined receive signal from the active antenna isdown converted with an LO of 4.5 GHz to IF of 850 MHz.The IF signal is subsequently I–Q detected prior to digitization.No provision for on-board range compressionexists. Hence, range compression has to be carried out onground. The baseband I–Q detected received signal issuitably band limited to maximize signal-to-noise ratio(SNR) by a set of four selectable I–Q filters.The first stage of the data acquisition unit is an 8 bitdigitizer which digitizes the received signal at 250 MHz.Subsequently, the data are decimated depending upon themode and fed to the BAQ subsection. RISAT providesthe unique feature of user selection of seamless BAQoption, from 2 to 6 bits, depending upon the applicationrequirements. However, for HRS mode, BAQ is limitedCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 453

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 11.Configuration of RF and baseband system.Figure 12.454SQNR performance of DACS measured in the laboratory.to 2–3 bits to maintain maximum data rate per channel upto 750 Mbits. Typical measured signal to quantizationnoise ratio (SQNR) performance of DACS is shown inFigure 12.The choice of antenna dimension puts a constraint inthe selection of PRF within a range of 2800–3700 Hzfrom both Doppler sampling requirement and rangeambiguity consideration. Total data rate from the SARpayload to Spacecraft Data Handling subsystem rangesfrom 140 to 1,500 Mbps.Payload control and managementRISAT–SAR payload is controlled by an array of controllersorganized in three-tier hierarchy as depicted in Figure13. At the top level of hierarchy, complete payload iscontrolled by a central computer called Payload Controller(PLC), which interfaces with all the RF and basebandsubsystems: DCG, V and H Receivers, FG, Feeder SSPA,calibration Switch Matrix and DACS. PLC is an autonomouscontroller with the only spacecraft interface beingDC power and 1553 interface with On Board Computer(OBC) of the spacecraft. OBC is the central computer ofthe spacecraft. PLC generates all the timing signalsrequired for synchronized operation of the SAR payload.The payload operation is controlled and managedautonomously by PLC. OBC translates the ground commandto suitable instructions to PLC for macro-levelpayload operation sequence.Six serial data channels, each operating at 240 Mbps, atthe DACS output are directly interfaced with the BasebandData Handling Unit (BDH) of the spacecraft for furtherformatting, recording, encryption and transmission toground reception system. The sensor operates with variablePRF and variable data window depending upon theantenna position. In scanSAR mode, different bursts areoperated at different PRFs and different data windowlengths. In order to maintain constant data rate at the datatransmission end, two levels of formatting are implemented.DACS formats every record with separate auxiliarydata and pseudo-random noise (PN) sequence header.In the BDH subsystem, fixed number of bytes from thepayload data is collected and reformatted. In case, no dataare available from DACS during formatting, null data areposted. In the on-board recorder, null frames are notstored. However, for maintaining data rate constant forground transmission, null frames are sent along withvalid data in real time mode.The payload controller in turn controls the active antennavia the TCUs residing in each tile. PLC essentially transmitsbeam definition command and switching sequencedefinitions to the active antenna. It also provides synchronizationsignals for beam switching and pulse operationto TCU, which in turn relays them back to TRC. PLCalso sends the beam weighting function definitions andresidual amplitude and phase correction parameters toTCU so that collimation of the beam is ensured.Default values of these parameters have been stored onboardin PLC, after integrated testing of the antenna. Provisionis made to impart these corrections either fromhard coded parameters residing in LUT of PLC or fromCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 13.Three-tier control of RISAT-1.modules, each dedicated for one polarization and oneEPC (PCPU) powering the TRC and two TR modules.TRC has in its memory all the temperature-related phaseand amplitude calibration data for each TR module andimparts the corresponding corrections from instantaneousmeasurement of ambient temperature.Redundancy approachFigure 14. Degradation in gain of RISAT-1 antenna because of randomfailure of TR modules.OBC, with parameters uploaded from the ground. However,till today, there has been no need to upload the collimationparameters from the ground.TCU controls beam pointing and beam setting in a tilevia TR controller. It also sequences TRM power on/offcommand. TCU transmits TR module-specific beamshifting, beam weighting, and phase and amplitude coefficients,after due correction of phase and amplitudemismatch, to specific TR modules.Each of the TR modules is controlled by the correspondingTRC. Each TRC controls two independent TRSeparate main and redundant chains are provided for theRF and baseband subsystems. However, for tile electronics,no redundancy is provided except for TCU. TCU isarranged in main-redundant configuration, printed on thesame Printed Circuit Board (PCB), fed by two separatepower supplies. Additionally, cross patching is providedbetween PLC-main and PLC-redundant and TCU-mainand TCU-redundant, to ensure better reliability. Randomfailure in TR modules will lead to graceful degradation inantenna gain as predicted in Figure 14.Performance of integrated payloadRISAT-1 SAR has a total of 126 antenna beams. Eachbeam definition comprises two transmit beams and tworeceive beams in H and V polarizations. These 126 beampointings are divided into 63 pointings for each side ofthe satellite track. Since left and right orientations areachieved in the shortest time by roll tilting, the 63 pointingsCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 455

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 15. Integrated RISAT Active Antenna in near-field antenna test set up in the clean room. a, In stowed condition, prior todeployment; b, After deployment.Figure 16. Time domain isolation of unwanted returns in the pulsecompression-based near-field measurement system for antenna patternmeasurement of RISAT–SAR.456on either side of the flight track are distinct. These pointingsalso include two patterns at zero pointing withantenna illumination of two longitudinal halves ofantenna of 6 × 1 m dimension. Antenna pattern wasmeasured in the integration laboratory itself. The measurementphilosophy is based on a novel concept of planarnear field (PNF) using time gating (Figure 15). It hasfacility to measure both transmit and receive patternmeasurements in both polarizations using the SAR pulse.The measurement was carried out under zero-g condition.The proposed measurement scheme ensured that themechanical references are kept the same for the fourdifferent pattern (Tx-V, Tx-H, Rx-V and Rx-H) measurements.The scanner is basically capable of scanning aprobe in the X–Y plane of the clean scan area of8 m × 4 m. However, it has a limited Z-axis scan capabilityof 20 cm. The scan plane is made parallel to the actualantenna plane. A laser tracking instrument providesinformation about the antenna plane.The PNF set-up is integrated with the payload such thatpayload itself is the source of PNF signal and payload receiverand digitizer are used as PNF receiver. The probemovement and CNC control are synchronized with thepayload operation such that the radar chirp pulse itself isused as a stimuli for near field (NF) measurement. Phaseand amplitude information for NF measurement at differentscan positions is derived from the compressed pulseresponse. In fact, at each scan position, the beam pointingis rotated through 12 pointing positions so that in effectany 12 antenna patterns can be measured simultaneously.Pulse compression aids in the suppression of returns fromunwanted targets from the vicinity, by separating them intime domain and thereby enabling NF measurement in theintegration laboratory (Figure 16). Use of chirped pulsefor antenna measurement ensured that the antenna patternobtained is essentially the one integrated over the fullbandwidth.Amplitude and phase deviations of each of the TRmodules are estimated in PNF through unique calibrationmethod by a single scan across the antenna and the reversecorrections are introduced in the TR module. The effectivenessof the correction procedure is shown in Figure17 for a single tile. In fact, the final correction factors arecoded in the PLC prior to payload shipment. However,on-board calibration can track relative drift from thedefined correction coefficients and provision is made toupload the residual correction coefficients to collimatethe beams. The effectiveness of the collimation strategycan be gauged from integrated antenna illuminationmeasurement carried out in PNF as shown in Figure 18.Though the integrated antenna can carry out 12 patternmeasurements simultaneously for a particular polarizationCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 17.Measured illumination profile for a tile prior to (a) and after (b) collimation.and transmit/receive configuration, each cycle of measurementwas time-consuming of the order of 10 days on24 × 7 basis. So 12 antenna pattern measurements, welldistributed in pointing angles, in both Tx/Rx configurationand H and V polarization were carried out over 45days. Subsequently, the rest of the antenna patterns werepredicted from the above measurements. In order toassess the prediction algorithm, 12 sets of antenna patternswere selected randomly and measured in the testset-up. Typical comparison of the predicted and measuredantenna patterns is shown in Figure 19. The figure showsclose agreement of predicted and measured antenna patternsup to a number of sidelobes.Post-launch performanceFigure 18. Typical antenna illumination profile for two differentpointings of RISAT-1 antenna as measured during integrated testing.Figure 19. Comparison of a typical predicted and measured antennapattern.Imaging operations of RISAT-1 SAR commenced from1 May 2012. After calibration and validation of the imageproducts, the RISAT-1 image products were released forglobal users from 19 October 2012. They are availablefrom National Remote Sensing Centre (NRSC), Hyderabad.Typical images, obtained by RISAT-1, are shown inFigure 20. They demonstrate the quality of the RISAT-1SAR images in a nutshell.During In-Orbit Test (IOT), azimuth antenna pattern ofthe RISAT active antenna was measured through a groundreceiver. Figure 21 shows close agreement of the measuredpattern with the predicted antenna pattern. It is to benoted that during integrated testing, antenna patterns werepredicted based on the limited NF measurements. DuringIOT only, for the first time far-field antenna patternswere measured. Further, radiometric correction usingpredicted elevation pattern could result in excellent radiometricbalance over required swath within ± 0.5 dB. Boththe above observations led to confidence in achievingcalibration of RISAT–SAR system using a single cornerreflector. The performance of calibration is shown inFigure 22 for FRS-1 mode image over Amazon rainforest.Reported average sigma naught (σ 0 ) over Amazonrainforest is –7.5 dB. The calibrated average estimated σ 0from RISAT-1 is close to the reported number.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 457

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 20.Typical images obtained by RISAT-1.Industrial partnership458RISAT-1, when conceived way back in 2002, was anambitious project where many new technologies were requiredto be mastered (Table 5 lists those which weremastered). The access to these technologies was restrictedbecause of international technology embargo, imposed onISRO. The number of subsystems needed for this systemwas prohibitively large (precisely 1391 subsystems, ofwhich 312 are 8 bit computers). Of these subsystems,those which were part of RF and baseband subsystemsand define the basic pulse Doppler radar blocks wereelectronics and software-intensive. They required complexdesigns and were required in fewer numbers. Theywere completely designed, fabricated and qualified inhouse.Whereas 95% electronics for active antenna realizationwere of few types (6 in number of types), withcomplex design, required in large numbers. Further, itwas decided to source all MMICs needed for TR modulefrom GAETEC, which was not even qualified for spaceuse at that point of time. All the electronics packageswere designed in-house, but production replications werecarried out in the Indian industry. However, MMICs weredesigned jointly with GAETEC.To qualify each of these subsystems to space grade, itrequired almost 3 weeks of testing for each of theseCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 21.Close match of measured active antenna patterns during in-orbit test with predicted ones.Figure 22. Estimated average σ 0 over Amazon rainforest (FRS-1 mode).subsystems. ISRO’s experience with satellites never exceededdealing with more than 100–200 subsystems.ISRO’s resources were inadequate to carry out fabricationand testing of such large number of elements. Indian industrywas not equipped to handle space-quality fabrication,let alone handle such large numbers with zero defectproduction and testing approach. A number of industrieswere imparted training to carry out mass productionbased on final blueprints which were designed and perfectedin-house.The highlight of RISAT-1 programme was developmentof hardware elements through industry partnership.Various new technology elements like different types ofMMICs, miniaturized C-band TR module and pulsedpower supplies, dual polarized printed antenna, integrationblock and power distribution network, high-speeddigitizers, micro-controller and FPGA-based central distributedcontrol systems have been realized with the activeparticipation and collaboration of public and private sectorindustries like GAETEC, ASTRA, CENTUM, CMC, ShahjanandLaser Technologies (SLT), etc. Indigenous MMICfabrication line has been qualified at the GAETEC foundry.Design, development and qualification of an indigenouson-board controller ASIC for tile electronics haveCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 459

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 5.New technologies/concepts implemented in RISAT-1Imaging modesExtended spotlightCircular polarimetryHardware technologySpace qualification of GAETEC MMIC foundryDesign, production and qualifications of MMICs by three processesDesign, development and characterization of TR modulesMiniaturized pulsed EPCs with 3 HMCs and planar transformersLaser guided printed antenna fabricationDual channel digitized at 250 MHz sampling with 22 W dissipationIn-house designed ASIC with 8 bit processor core3-Tier synchronized software control spread over 312 ASICSMeasurement technologyNear field processorNovel time gated near-field antenna measurement systemNovel active antenna beam collimation methodTR module characterization methodPrecise measurement of group delay of RF cables with accuracy better than 1 psFabrication technologyGroup delay equalized RF cable fabrication with S.D. better than 4–5 ps withmore than 90% yieldAlgorithm developmentExtended spotlight SAR processing algorithmDecomposition algorithm for hybrid polarimetric SAR dataFirst time in RISATFirst time in RISATCopyrightedPatent appliedPatent appliedPatent appliedCopyrightedKnowhow transferredPatentedPatent appliedTable 6.Industrial partnership in RISAT-1 developmentIndustryGAETEC, HyderabadASTRA Microwaves, HyderabadKomolin, Ahmedabad as part of ASTRACentum Electronics, BangaloreBEL, GhaziabadAgilent IndiaCG-COREL/AEROFLEXSLT, GandhinagarCMC, Hyderabad and AhmedabadBombay Machines, BangaloreK.V. Microwaves, GhaziabadMaharshi Electronics and Komolin, AhmedabadHardware contribution7 types of MMICS involving three processesTR module, TRC, 1 : 12 RF power dividers, power amplifier modules,etching of antennaTR module, TRC, 1 : 12 RF power dividers, Power amplifier modulesPCPU, TRC, harness, EPC for TCU, TCUPower amplifier moduleTR module characterization systemASIC for TRCAntenna fabricationAll digital subsystems for design verification model of RISAT9 m × 6 m scanner for near-field measurementIndigenous microwave absorbers for testingHarness and RF cable fabrication and integrationalso been accomplished in collaboration with privateindustries like CG COREL/AEROFLEX. The participatingindustries had to go through a learning curve with issueslike space-grade circuit fabrication, quality control,test facility development and testing methodology. Detailedlisting of these partner industries and their contributionshas been made in Table 6. This programme hasalso added to industrial capacity building within thecountry. Thus, the challenge of microwave SAR payloadrealization has been addressed and with industry participationin these activities, a new beginning has been made.ConclusionInitial performance assessment indicates satisfactory performanceof the SAR system of RISAT-1 in terms of460image quality and active antenna pattern synthesis. Thisperformance is the result of development of complex SARhardware, active antenna system, accurate ground measurementand all the necessary innovations which havegone into making the technology development completelyindigenous. Such complex development would not havebeen possible without the partnership forged throughindustry after due training was imparted to them on fabricationand qualification of space-grade electronics.1. Kramer, H. J., Observation of the Earth and its Environment:Survey of Missions and Sensors, Springer, 2002.2. Henderson, F. M. and Lewis, A. J. (eds), Principles and Applicationsof Imaging Radar, Manual of Remote Sensing, John Wiley,1998, 3rd edn, vol. 2.3. Jordan, R. L., The Seasat-A synthetic aperture radar system. IEEEJ. Ocean. Eng., 1980, OE-5, 154–164.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-14. Cimino, J. B., Elachi, C. and Settle, M., SIR-B – the second shuttleimaging radar experiment. IEEE Trans. Geosci. Remote Sensing,1986, GE-24, 445–452.5. Almaz to add dimension to earth study. Space News, 18–24 March1991, p. 1.6. Attena, E. P. W., The active microwave instrument, on-board theERS-1 satellite. Proc. IEEE, 79, 791–799.7. Nemoto, Y., Nishino, H., Ono, M., Mizutamari, H., Nishikawa, K.and Tanaka, K., Japanese earth resources satellite-1 syntheticaperture radar. Proc. IEEE, 79, 800–809.8. Jordan, R. L., Huneycutt, B. L. and Werner, M., The SIR-C/X-SAR synthetic aperture radar system. Proc. IEEE, 1991, 79, 827–838.9. Raney, R. K., Luscombe, A. P., Langham, E. J. and Ahmed, S.,RADARSAT, Proc. IEEE, 1991, 79, 839–849.10. Jordan, R. L. et al., Shuttle Radar topography mapper. In Proceedingsof the EUROPTO Conference: Symposium on Remote Sensing,Taormina, Italy, 1996.11. ENVISAT <strong>Special</strong> <strong>Issue</strong>, ESA Bulletin, June 2001, no. 106.12. Fox, P., The RADARSAT-II Mission. In Proceedings of theIGARSS’99, Hamburg, Germany, 1999, vol. III, pp. 1500–1502.13. Wakabayashi, H. et al., PALSAR system on the ALOS. Proc.SPIE, 1998, vol. 3498, p. 181.14. Keith Raney, R., Hybrid-polarity SAR architecture. IEEE Trans.Geosci. Remote Sensing, 2007, 45, 3397–3404.ACKNOWLEDGEMENTS. Development of RISAT SAR payloadwould not have been possible without active involvement and dedicatedefforts of a large number of engineers and scientists from ISRO andalso our industrial partners. We thank all of them for their efforts tobring this advanced technology to maturity. We also thank Dr K.Radhakrishnan (Chairman, ISRO), Dr K. Kasturirangan, Dr G. MadhavanNair (former Chairmen, ISRO), A. S. Kirankumar (Director, SAC),Dr A. K. S. Gopalan, Dr K. N. Shankara and Dr R. N. Navalgund (formerDirectors, SAC) and Dr P. S. Goel and Dr T. K. Alex (formerDirectors, ISAC) for providing sustained support, encouragement andguidance.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 461

SPECIAL SECTION: RADAR IMAGING SATELLITE-1RISAT-1 spacecraft configuration:architecture, technology and performanceN. Valarmathi 1, *, R. N. Tyagi 2 , S. M. Kamath 1 , B. Trinatha Reddy 1 ,M. VenkataRamana 1 , V. V. Srinivasan 1 , Chayan Dutta 1 , N. Veena 1 ,K. Venketesh 1 , G. N. Raveendranath 1 , G. Ravi Chandra Babu 1 ,K. Sreenivasa Prasad 1 , Rajeev R. Badagandi 1 , P. Natarajan 1 , S. Sudhakar 1 ,J. Subhalakshmi 3 , Sreenivasa Rao 4 and M. Krishna Reddy 51 ISRO Satellite Centre, Indian Space Research Organisation, Bangalore 560 017, India2 Formerly with ISRO Satellite Centre, Indian Space Research Organisation, Bangalore 560 017, India3 Laboratory for Electro-Optics Systems, and 4 Liquid Propulsion Systems Centre, Indian Space Research Organisation, Bangalore 560 058, India5 ISRO Inertial Systems Unit, Indian Space Research Organisation, Thiruvananthapuram 695 013, IndiaRISAT-1 is the first indigenous active Radar ImagingSatellite launched by Polar Satellite Launch Vehicle(PSLV) in April 2012 from Sriharikota. It carries C-band Synthetic Aperture Radar (SAR) payload whichcan operate at various values of resolution and swathfor various applications. RISAT-1 is the heaviest andhigh-power satellite with many new technologies tosupport the SAR payload and the associated elements.RISAT-1 operates at polar sun-synchronous orbit of536 km altitude with the inclination of 97.554° and itis designed for 5 years lifetime. Performance of thespacecraft system and the SAR payload is satisfactory.This article outlines the architecture, design and onorbitperformance of RISAT-1.Keywords: Architecture and design, radar imaging satellite,synthetic aperture radar payload.IntroductionTHE space-based remote sensing programme is applications-drivenand covers observations on land, ocean andatmosphere. To serve these applications effectively, satellitesneed to fly various types of sensors which may operatein optical or microwave region of the electromagneticspectrum. So far ISRO has operated satellites using bothregions. While Indian Remote Sensing Satellite (IRS)series of satellites were mainly based on electro-opticalsensors, there are other sensors like Multi-frequencyScanning Microwave Radiometer (MSMR), scatterometerand X-band Synthetic Aperture Radar (SAR) onOCEANSAT-1, OCEANSAT-2 and RISAT-2 respectively.Sensors operating in microwave region of electromagneticspectrum have the capability to provide dataduring the day, night and all weather conditions. SAR*For correspondence. (e-mail: valar@isac.gov.in)462technology provides data on structural information togeologists for mineral exploration, oil spill boundaries onwater to environmentalists, state of the sea and ice hazardmaps to navigators and reconnaissance data to strategicapplications. SAR data is well suited for the quantitativeobservation of critical national and global resources suchas tropical forests and will be a primary source of informationfor resource-monitoring and analysis.The SAR programme began with C-band SAR for airborneapplications and the user community is providedwith imagery from airborne SAR and C-band imageryfrom RADARSAT-1/RADARSAT-2. The attributes ofSAR imagery depend upon the sensor parameters such asfrequency, polarization, look angle and the attributes ofobjects such as surface roughness, moisture content, orientation,dielectric constant, etc. Thus C-band was chosenfor RISAT-1 to cater to a wide variety of applications.The mission elements of RISAT-1 are:• Space segment comprising three-axis stabilized satellite,flying SAR payload and mainframe systems.• Data reception, recording, processing and disseminationfacilities having the required hardware and softwarecomponents on ground.• Spacecraft control centre for tracking, commandingand receiving telemetry data for satellite health monitoringand analysis, orbit maintenance and payloadprogramming functions.• Development of user-friendly data products and dataarchival.Orbit choice for RISAT-1The main guiding parameter for choosing the orbit forRISAT-1 is achieving global coverage in a systematicway for a given swath. Other considerations such as thepresence of atomic oxygen and atmospheric drag havealso been kept in view.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 1.Block diagram of RISAT-1 spacecraft.A polar sun-synchronous orbit at 536.38 km altitudeand inclination of 97.554° with repetivity cycle of 377orbits in 25 days is chosen. Global coverage is achievedtwice in the repetivity cycle, once by a set of descendingpasses and next by a set of ascending passes, as SAR is amicrowave payload with no illumination constraints.RISAT-1 configuration and architectureRISAT-1 is a high power, heavy-weight active microwavesatellite. It is a 3-axis stabilized satellite operatingin sun-synchronous 6 a.m.–6 p.m. orbit of 536 km. It carriesthe SAR payload supported by other mainframe elements.It can take images with ± 36° roll bias on orbit.The architecture is similar to that of earlier missions suchthat it supports the synthetic aperture microwave payload,electrically and mechanically. RISAT-1 adapted a technologyfor the mainframe system in order to meet itsspecification and requirements. The basic block diagramof RISAT-1 is given in Figure 1. It carries many new elements(with new designs) like power, battery, X-bandmodulator, phased array antenna (PAA), on-board computer,reaction wheels, solar array drive electronics andassembly, payload data handling system, structure andSAR antenna deployment mechanism. Single-point failureand redundancy for the subsystems have been takencare in the configuration.The structure is built around a central cylinder withequipment decks for accommodation of subsystems. Ithas a unique structure carrying elements of bus systemand payload system which is different from earlier remotesensing missions. The spacecraft bus subsystems providebasic housekeeping functions like orbit correction, attitudedetermination, thermal control, electrical power generationand distribution, ground communications, SARimage data handling and storage and data transmission tothe ground. The bus module contains all the necessarysystems to operate and maintain the spacecraft in orbitand support the SAR payload. The overall mechanicalconfiguration of RISAT-1 is simple and efficient. It alsosatisfies the envelope and Centre of Gravity (CG) constraints/requirementsof the launch vehicle. PSLV-XLwas the launch vehicle for RISAT-1 spacecraft.Stowed configuration of RISAT-1 at the launch padinside Polar Satellite Launch Vehicle (PSLV) is shown inFigure 2 and the deployed configuration of RISAT-1 isshown in Figure 3.RISAT-1 designStructureThe structure is designed to meet the stiffness, strengthand pointing requirements of the payload, sensors andalso confining the overall bus volume within the launchvehicle envelope. It is based on a single bus concept builtaround a central cylinder. A truncated triangular structureis built around the cylinder to hold the SAR antenna andCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 463

SPECIAL SECTION: RADAR IMAGING SATELLITE-1major bus service elements. A cuboid structure is built ontop of the cylinder to accommodate the solar arrays, majorityof the sensors and antennae. The primary structureconsists of a central cylinder, interface rings and shearwebs. The central cylinder is of sandwich constructionwith aluminium core and carbon-fibre-reinforced polymer(CFRP) face skin. It has an aluminium alloy interfacering at the bottom to interface with the launch vehicle.The cylinder also provides interface for the propellanttank and reaction wheel deck. Secondary structures consistof equipment panels/decks of the payload module andthe cuboid module.The payload module structure consists of three equipmentpanels, three corner panels and top and bottom deck.All the equipment panels and corner panels of the payloadmodule are made of sandwich construction with aluminiumcore and aluminium face skin, whereas the shear webs aremade of sandwich construction with CFRP face skin. Thetriangular decks carry the hold-down brackets to hold theSAR antenna in launch configuration.The SAR antenna is comprised of three panels, ofwhich one is fixed and the other two are stowed ontoeither sides of the triangular structure during launch andare deployed in the orbit. Tile Substrate and Panel Frameare two basic structures over which the SAR payloadis built. The radiation patch antennae are bonded on oneside of the Tile substrate and the Tile electronics mountedon the other side of the substrate. Four Tiles form a panelfor the SAR antenna. To support these four Tiles, aframed structure is evolved. Most of the sensors, antennae,solar arrays and their associated electronics aremounted in the cuboid module. RISAT-1 main structureis shown in Figure 4.The subsystem layout has been evolved consideringvarious factors like electrical requirements, interfacesamong various subsystems, physical size and locationfeasibility, look angle and field-of-view (FOV) requirementsof various elements (payloads, sensors, antennae),thermal requirements, mechanical loads, transmissibilityfactors, physical parameters and balancing, ease ofassembly/dis-assembly and accessibility during assembly,integration and testing (AIT) and pre-launch operations.All the subsystem electronics packages are accommodatedon the equipment decks/panels.The payload module (triangular structure) accommodatesmost of the mainframe systems and the payloadelectronics. The cuboid module accommodates solararrays, most of the sensors and antennae, viz. Digital SunSensor (DSS), Solar Panel Sun Sensor (SPSS), EarthSensor (ES), 4π Sun Sensors, PAA, TTC Antennae andSatellite Positioning System (SPS). All the Reaction ControlSystem (RCS) components are accommodated on oneof the shear webs and the exterior surface of the triangularbottom deck. The propellant tank is mounted insidethe main cylinder. The reaction wheels are mounted on acircular deck in a tetrahedral configuration. The circularFigure 2. Stowed configuration of RISAT-1.Figure 3.Deployed configuration of RISAT-1.Figure 4. RISAT-1 main structure.464CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1deck is accommodated inside the main cylinder below thetank and is connected to the cylinder through a ring.Thermal systemThermal design should maintain the temperatures of thesubsystems within design limits during all conditions ofoperation, all seasons and throughout the lifetime of thespacecraft.The configuration and equatorial crossing time ofRISAT-1 are different from other satellites in the IRSseries. Though it is an earth-oriented satellite, during payloadoperation the satellite will be rotated by ± 36° aboutroll axis. This new configuration, orientation and equatorialcrossing time result in new external load patterns andextreme load conditions which are different from otherIRS satellites. Moreover, a number of heat dissipatingpackages are accommodated inside the structure.Thermal control is provided using space-proven thermalcontrol elements such as optical solar reflector(OSR), multilayer insulation (MLI), paints, thermal controltapes, quartz wool blanket, sink plates and heat pipes.In addition, heaters will be provided to maintain temperaturesduring cold conditions.MechanismRISAT-1 spacecraft employs SAR antenna deploymentmechanism and solar array deployment mechanism. SARantenna and solar array are stowed during the launch andare deployed in the orbit in order to meet the constraintsimposed by the launch vehicle. In order to performdeployment in the orbit, a hold-down and release mechanismis employed. The solar array deployment mechanismis identical to earlier IRS missions.The deployed SAR antenna has dimensions of6.29 m × 2.09 m × 0.220 m. It consists of three panels outof which one is rigidly attached to the triangular structure.In the launch configuration, the deployable panelsare folded over the triangular structure and are held byusing a hold-down mechanism. In the orbit both thedeployable panels are released sequentially and deployed.The mass of each panel is about 290 kg.Electrical power system: The power system consists ofsolar array for power generation, chemical battery forpower storage and power electronics for power conditioningand distribution. It is designed to meet the 6 a.m./6 p.m. orbit illumination conditions, large power requirementof SAR payload and solar eclipse conditionsduring summer solstice.The solar array consists of six panels arranged in twowings with three panels in each wing in positive roll andnegative roll axes. The array consists of multi-junctioncells connected in series and parallel for optimum configuration.The solar array drive assembly helps in compensatingthe roll bias (± 36°) given during payloadoperation and also aids in obtaining more generation nearpole transit.The energy storage system for RISAT-1 employs asingle NiH 2 battery of 70AH capacity to meet the peakload requirement and also the eclipse requirement.It is a single-bus system operating at 70 V and the configurationis arrived at to meet all the requirements ofusers and interfaces. During the sunlit period, the array isregulated to 70 V and the battery gets charged. A BatteryDischarge Regulator (BDR) supports power to the buswhen the load demand exceeds the array generation duringpayload operation and eclipse conditions by regulatingthe bus to 70 V. Bus voltage selection is mainly driven bypayload requirement. The single bus of 70 V is fully protectedagainst over voltage, over current and is singlepointfailure proof. The bus is distributed to all usersthrough fuses, centrally located in fuse-distribution packages.Software Logics (software resident in the on-boardcontroller) enhances the safety of the power system.On-board computerIn order to minimize power, weight and volume, thespacecraft functions like commanding, housekeeping (telemetry),attitude and orbit control, thermal management,sensor data processing, etc. have been integrated into asingle package called On-board Computer (OBC), whichalso implements the MIL STD 1553B protocol for interfacingwith other subsystems of the spacecraft (Figure 5).The use of MIL STD 1553B interfaces between OBCand other subsystems greatly decreases the volume andmass of cabling, and the associated connectors. The OBCsystem is realized with the functions of sensor electronics,command processing, telemetry and house-keeping,attitude and orbit control and thermal management.Besides, the OBC interfaces with power, telemetry–telecommand (TM–TC; RF) for command and telemetry,sensors, heaters, thrusters and reaction wheels throughspecial logics.Figure 5. Block diagram of on-board computer.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 465

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Integrated control subsystem: AOCS specifications duringimaging are as follows: pointing: ± 0.05° (3σ ); driftrate: ± 3.0e-04°/s (3σ ).The AOCS configuration is as follows: The attitudeorbit control system for RISAT-1 is configured with 4πSun Sensor, Magnetometer, Inertial Reference Unit(IRU), Star Sensor, Earth Sensor, Digital Sun Sensor andSolar Panel Sun Sensor. Acutators are eight numbers ofcanted 11 N thrusters (mono propellant hydrazine systemoperating in blow-down mode) with two-axis cantingfrom + pitch axis for acquisition and OM operation, 1number of central 11 N thruster for OM operation, 4numbers of reaction wheels (of capacity 0.3 Nm torqueand 50.0 NMS) mounted in tetrahedral configurationabout – pitch axis and magnetic torquers of 60.0 A m 2capacity for momentum dumping. Sun sensors, star sensorsand magnetometer provide attitude data in the formof absolute attitude errors. The magnetometer, 4π sunsensor and temperature sensor data are processed in OBC.All AOCS software modules are implemented in OBC.Reaction Control SystemThe reaction control system comprises propellant tank,thrusters (9 numbers of 11 N), latch valves, fill and drain/vent valves, pressure transducers, system filters, thermocouples,flow control valves and titanium tubes to connectall the reaction control elements. Block schematic ofreaction control system is given in Figure 6. One central11 N thruster is meant for orbit control and the remainingeight 11 N thrusters for attitude control.Telemetry Tracking and Command SubsystemThe Telemetry, Tracking and Command (RF) system forRISAT-1 consists of two chains of Phase Locked Loop(PLL) coherent S-band transponder connected to a commonantenna system. The basic configuration is identicalto the ones employed in earlier IRS missions. The TCdemodulation scheme is phase shift keying (PSK)/pulsecode modulation (PCM) with a date rate of 4 kbps. Thetransponder consists of a receiving and transmitting systemand can operate in either coherent or non-coherentmode. Range and two-way Doppler data from the transponderare useful for orbit determination.Payload Data Handling SubsystemRISAT-1 payload data need to be transmitted either inreal time or in playback mode depending upon the datarates at different modes. The data-handling system ofRISAT-1 is configured with two formatters for each ofthe SAR payload receivers respectively (Figure 7).They are high data rate formatters for different datarates of payload with memories for burst data formatting.Systems have been realized by field-programmable gatearrays (FPGAs) and the design is optimized for weight,power and volume. When the data rate of SAR payloadand BDH overhead together is greater than 640 Mbps,real-time transmission is not possible and the data isrecorded in SSR. Recorded data can be played back later.Data handling system can operate in real time, real-timestretch mode, record and playback modes.Solid State RecorderThe RISAT-1 Solid State Recorder (SSR) has a capacityof 300 Gbits, realized with six memory boards of50 Gbits capacity each. The memory boards, by defaultare configured into two partitions each of 150 Gbits withthree memory boards per partition. The SSR has two controlunits for configuring and controlling the internal operations.The controller has two separate 32-bit parallelinterface with memory boards. The default configurationFigure 6.Block schematic of reaction control system.Figure 7. Payload data formatters.466CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1is for two partitions; however, the system can be configuredfor single partition with allocation of all the memoryboards to the selected partition. The SSR is able to manageup to 32 different files for each input port.The memory management guarantees the usage of allgood devices by automatic configuration after the diagnosticscommand is issued.The allocation of address during recording is managedby the SSR and ground control will not be possible. Eraseoperation of files is carried out by a command with thefile identifier. Playback with segment erase is also one ofthe options wherein the segments played back shall be released,with a gap of one segment. The record operationdepends on the available free space in the SSR capacity.It is possible to read the data files in any order and anynumber of times. The bit error rate (BER) at SSR level isbetter than 1 × 10 –12 .The memory plan is constituted by four physical levels,namely memory device level, memory module level,memory bank level and board level and partition level.There are a total of six memory boards which can beswitched ON/OFF by command.X-band RFThe X-band RF is required to accept the payload datafrom the baseband data handling system, modulate theabove data on two X-band carriers and transmit the sameto the ground after suitable amplification and filtering.The SAR payload of RISAT-1 when operated in dualpolarization imaging mode generates data at the rate of640 Mbps and this needs to be transmitted to the groundstations. Data rates up to 170 Mbps have been transmittedat X-band using shaped beam antenna in earlier missionslike IRS-1C/1D and PAA in Technology experimentsatellite. In order to meet the high data rate transmissionrequirement in X-band quadrature phase shift keying(QPSK) modulation with frequency reuse by polarizationdiscrimination is implemented.In the data transmission for RISAT-1, half the data, i.e.320 Mbps will be transmitted in right-hand circularpolarization (RHCP) and the remaining 320 Mbps in theleft-hand circular polarization (LHCP); two identicalchains operating at X-band are used to transmit 640 Mbpsof payload data. The carrier generation section, QPSKmodulator section, filter units, selection of main andredundant chain units are identical in all the chains as thefrequency of operation and modulation schemes is identical.Both the chains have end-to-end redundancy.Phased Array AntennaThe spherical PAA has radiating elements distributedalmost uniformly on a hemispherical surface. It generatesa beam in the required direction by switching ‘ON’ onlythose elements which can contribute significantly towardsthe beam direction. It is proposed to use the 64 elementarray.Operationally it consists of two identical phased arrays,one operating in RHCP and the other operating in LHCP,and located in the same hardware. On the spherical domean element is located at a defined location. A waveguideradiating element fed by a septum polarizer is plannedand this has two ports, one for RHCP and the other forLHCP. The radiating element is optimized to provide therequired isolation (better than –25 dB) between the twopolarizations to minimize the interference.The RHCP and LHCP ports of the phased array areconnected to two separate sets of power dividers andmonolithic microwave integrated circuit (MMIC) amplifiers.A common beam steering electronics controls theswitch position and phase setting for all the MMIC amplifiers.Data transmission chain is given in Figure 8.Satellite Positioning SystemSatellite Positioning System (SPS) for RISAT-1 comprises10-channel C/A code GPS receiver at L1(1575.42 MHz) frequency. SPS is designed for computingthe state vector of the high-dynamic platform.SPS for RISAT-1 will have full-chain (end-to-end)redundancy. Each chain consists of a receiving antenna,low-noise amplifier, RF amplifier and power divider inL-band followed by a 10-channel and 8-channel GPSreceiver with MIL 1553B interface. Each GPS receiverconsists of two high dynamics GPS receiver core engine(RCE) modules to compute state vectors and one receiverchain will be active at a time.SPS is placed in RISAT-1 to track GPS signals continuously.It requires an antenna system with hemisphericalradiation coverage to receive the circularly polarizedGPS signal from the navigational satellites. Micro-strippatch antenna is used for this application.RISAT-1 specifications, new elements andchallengesFor all the new elements, the concept was proved withdevelopmental model followed by qualification modeland flight model. Baseline design review (only for newelements), preliminary design review, detailed design reviewhave been conducted for all the subsystems and therecommendations have been implemented successfully.Preliminary design review and critical design reviewwere conducted for the space segment and the groundsegment. Thermal analysis, derating analysis and failuremode, effects and criticality analysis (FMECA) were carriedout for all the subsystems and suitable measures havebeen adopted. Fabrication procedure/new processinghas been followed after due approval by the MaterialCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 467

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 8. Data flow from SAR payload to Phased Array Antenna.468Table 1. RSIAT-1 Major specificationsOrbit536.38 km, 6 a.m.–6 p.m.Inclination 97.554°SAR payloadC-band (5.35 GHz)SAR payload operating modes CRS, MRS, FRS1, FRS2, HRS (withlinear and circular polarization)Resolution1–50 mData rate for transmission 2 × 320 MbpsSSR240 GbitsTT&CS-bandPayload down linkX-band (frequency reuse)PowerRegulated bus 70 V/42 V/U-busBattery70 AHTelecommand4 Kbps PSKTelemetry4 Kbps PSKSatellite mass1858 KgReview Board. Test and evaluation results of all the subsystemswere submitted for review.Total number of RISAT-1 subsystems were 1300(approx.) with many controllers, DC–DC converters, highfrequency and high dissipating packages, a variety ofinterfaces (MIL STD 1553 B, RS 422, RS 488, LVDS),RF cables, power lines and control signal lines whichdemanded proficiency in carrying out the assembly integrationactivities. Disassembled configuration to assembledconfiguration was executed faultlessly. Testing themicrowave payload and the related mainframe elementswas quite taxing at all test phases and SAR testing calledfor antennae panel in deployed condition.Handling RISAT-1 satellite which is the heaviest satelliteof mass 1858 kg in the remote sensing satellite classand high power of 4200 W at all test phases, was the biggesttask. Thermovac test posed a big challenge for theteam. The spacecraft was positioned inside the thermovacchamber in deployed condition with thermal instrumentation(Figure 9). The thermovac test was conducted successfully.<strong>Special</strong> tests, namely radiation, EMI/EMC, RF compatibility,wheel interaction test with main structure (byhanging the satellite in on condition with wheels operation)and polarity test were conducted. The satelliteunderwent vibration test and acoustic test in the newacoustic test facility at ISITE. Pre-launch test at SHARwas executed smoothly. Fuel filling was the final activityCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1before PSLV mating and was executed accurately as themass of RISAT-1 had to meet the requirement of PSLV.RCS performance was critical for raising the orbit to536 km from 436 km. Data reception, data processing anddata product generation (first of their kind) are being executedearnestly.Performance of RISAT-1 systemAfter mating with PSLV, launch base configurationof RISAT-1 was selected according to plan. PSLV had injectedthe spacecraft into 436 km orbit. Solar arrays andSAR antenna panels were deployed by snap commandTable 2. RISAT-1 new elementsSubsystem (no heritage) HighlightSAR payloadBased on TR module architectureStructureTriangular prism, cuboid module,tile frame, tiles, etc.Thermal systems Designed for 6 a.m.–6 p.m. orbitMechanismSAR antenna deployment for 300 kg panelReaction wheels 50 NMS angular momentum and 0.3 NMtorque for mission manoeuvrePower70 V bus high power (4.2 KW)Battery 70 AH NiH 2BDH320 Mbps (variable data rate)X-band systemHigh data rate modulator 320 MbpsPhased array antenna Wave guide antenna with RHCP and LHCPSSR300 Gb with high data rate handlingFigure 9. RISAT-1 spacecraft in thermovac.according to mission requirements soon after the spacecraftseparation from PSLV. The performance of themechanism systems is listed in Table 3.Orbit manoeuvres were carried out to raise RISAT-1from the present orbit to 536 km and also get it ready foroperational services. Reaction control system completedits task according to design. The fuel available on-boardis estimated to be 53.485 kg. Considering all aspects offuel consumption, the RISAT-1 can be expected to have auseful life of more than design value. SPS data is the primarymode of orbit determination for this mission and theperformance of the same is satisfactory.On-board computer performance is normal. Launchphase sequencer operations for deployment of solar arrayand SAR panels were executed successfully. Initialacquisition with inertial acquisition (IAC) mode and laterearth acquisition operations were normal. Safety logicsare enabled. Interfaces with sensors, actuators, power andthermal systems are functioning well. On-board performanceof inertial reference unit (IRU) is satisfactory and allperformance parameters are within specified limits.The telemetry, tracking and command system has beenproviding health information, access to configuring thespacecraft for various routine operational requirements.The link margins on S-band uplink and downlink areabove 15 dB as observed from ISTRAC ground stations.The telemetry data storage and SPS data storage system isoperated routinely in every orbit to assess the health ofthe spacecraft even outside the network visibility.Solar array power generation is 2100 W and power generationstarted right after the heat shield separation. Thepower system has been generating, conditioning and supportingthe various load requirements of all the systems.Battery supported the launch, all the initial operations likeorbit manoeuvres and payload operations successfully, andsupport continues subsequently for payload operation andeclipse conditions. Battery charging operation is automaticallytaking place according to the set rate and charge terminationvoltage. Power system performance is normal andall modes of payload operations are well supported.Attitude and orbit control system (AOCS) has beenperforming well in the three-axis stabilized mode. Variousmodes of AOCS have been exercised and the performanceis normal. Four reaction wheels, two gyros anda star sensor with corresponding control electronics havebeen maintaining the attitude of the spacecraft within thespecified limits. Earth sensors are used for safe modedetection. Spacecraft rates during three-axis acquisitionare given in Figure 10. Attitude performance during firstpayload operation is given in Figure 11.A summary of AOCS performance is as follows:• Pointing out errors during imaging within 0.0034°.• Residual S/C body rates during imaging:yaw: 3 × 10 –5 deg/sec, roll: 5 × 10 –5 deg/sec,pitch: 2 × 10 –5 deg/sec.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 469

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 10.Spacecraft rates during three-axis acquisition.Figure 11.Attitude performance during first payload operation.• S/C pointing error as measured by star sensor:yaw: 0.03°, roll: 0.01–0.04°, pitch: 0.05–0.08°.On-board performance of all four reaction wheels is satisfactory.Wheel current and bearing temperature are withinspecified limits. Wheel speed during payload operationsis according to expectation (Figure 12).Thermal performance of RISAT-1 spacecraft is normaland as expected. Temperatures are stable. Functioning ofall thermal control elements (253 temperature sensors,133 heaters, 54 heat pipes, 13 m 2 of quartz wool blanket,470MLI, OSR, thermal control tapes, heat sink plates) isnormal and as expected. Temperatures of spacecraft andon-board systems are matching closely with the predictedvalues for on-orbit cold season. ATC heaters for payload,battery, RCS elements, wheels, SSR, BDR, PAA andcuboid are enabled and operating with predicted dutycycles.RISAT-1 has been functioning satisfactorily and the inorbitoperations are being executed according to plan.The mainframe has been supporting the payload operationsand the SAR payload is sending quality picturesCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 12.Speed of wheels during payload operations.Figure 13.Sample of RISAT-1 images.Table 3.On-orbit performance of the mechanism systemsOn-orbit predicted Actual on-orbitSubsystem deployment time (sec) deployment time (sec)P2-SAR antenna 87 87.17P3-SAR antenna 81 80.89+ Roll solar array 6.8 6.4– Roll solar array 6.6 6.34during morning and evening passes over India. Variousmodes of payload are carried out in sequencer mode(auto) through timers which are set prior to the intendedoperations by the spacecraft control centre personnel. Allspecial requests and events of importance in and aroundIndia are being covered. Sample data products of FRS 1and MRS are given in Figure 13.The X-band data transmission link has been performingsatisfactorily and the margin established during initialphase is still valid and satisfactory.ConclusionsWith the launching and operationalization of RISAT-1,India has emerged as one of the few counties in the worldwith a capability of using satellite-based microwaveremotely-sensed data for various resource applications onan operational basis. The capabilities of RISAT-1 arecomparable with other contemporary satellite missions.RISAT-1 has thus laid a strong foundation for the futureof microwave remote sensing activities in the country.Advanced versions of SAR missions will provide the continuityof RISAT-1 services to the user community inIndia in the coming years.ACKNOWLEDGEMENTS. We thank Chairman, ISRO, Director,ISAC and Programme Director (IRS&SSS) for their support and encouragementthroughout the project. We thank Centre Director of variousISRO Centres and Units for their guidance and support. We also thankthe various Groups and the Facilities of ISRO Satellite Centre, whichhave been involved in the development and realization of this mission.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 471

SPECIAL SECTION: RADAR IMAGING SATELLITE-1PSLV-C19/RISAT-1 mission: the launcheraspectsP. Kunhikrishnan*, L. Sowmianarayanan and M. Vishnu NampoothiriPSLV Project, Vikram Sarabhai Space Centre, Indian Space Research Organisation, Thiruvananthapuram 695 022, IndiaIndia’s Polar Satellite Launch Vehicle (PSLV) preciselyplaced the indigenous Radar Imaging Satellite(RISAT-1) in the intended orbit successfully on26 April 2012. This marked the 20th successive successof the launcher. This article gives the specialaspects of the ‘PSLV-C19/RISAT-1 mission’ from thetechno-managerial perspective of launch vehicle project.Also given is a summary of the performance ofthe launch vehicle in this flight.Keywords: Launch vehicle, orbit planning, satellite,techno-managerial challenges.IntroductionTHE Polar Satellite Launch Vehicle (PSLV) is a fourstagelaunch vehicle designed and developed with theprimary objective of placing spacecraft in the Sun-Synchronous Polar Orbits (SSPO). After three valuabledevelopmental flights, PSLV was operationalized in 1997with the launch of PSLV-C1 carrying the 1205 kg IndianRemote Sensing satellite, IRS-1D.During the operational phase, the payload capacity ofPSLV was enhanced by increasing the propellant loadingof the first stage from 125 tonnes to 139 tonnes, inductionof augmented second stage (liquid PL40), high performancethird stage solid motor (HPS3) and carbon fibrereinforced polymer (CFRP)-based structures. Provisionswere added for accommodating multiple spacecraft eitherusing Dual Launch Adaptor (DLA) or auxiliary payloaddecks on the Equipment Bay (EB) of the vehicle.PSLV has three variants, namely, PSLV – the genericversion with six regular strap-on motors (S9), PSLV–CA – the core alone version without strap-on motors andthe more powerful PSLV-XL with S12 strap-on motors(S12 is the extended version of the regular S9 strap-onsin terms of length and propellant loading). The upperstage has three versions based on propellant loading requiredfor a particular mission. The vehicle configurationfor a mission is selected based on spacecraft characteristicsand the mission requirements. The current payloadcapability of the PSLV-XL vehicle is 1750 kg in 600 kmSSPO and 1425 kg for the Sub Geosynchronous TransferOrbit (Sub-GTO) of 284 × 21,000 km.*For correspondence. (e-mail: p_kunhikrishnan@vssc.gov.in)472All the three variants of PSLV have been successfullyemployed to place spacecraft in different destinations likeSSPOs, planar orbits with specific inclination and alsothe Sub-GTOs. Today, PSLV is a universal and versatilelaunch vehicle. Each of its missions is unique with respectto the orbit, spacecraft, trajectory parameters and otherrequirements, posing fresh techno-managerial challengesto the planning and execution teams. The recent PSLV-C19/RISAT-1 mission is an apt example in this regard.New aspects of PSLV-C19The PSLV-C19/RISAT-1 mission employed PSLV-XLconfiguration of the launch vehicle (Figure 1) with itsupper stage (PS4) loaded with 2.5 tonnes of liquid propellantto carry the heaviest satellite (1858 kg) ever entrustedto PSLV. There are several unique features andrequirements specific to PSLV-C19 vehicle, the followingbeing the salient ones.Figure 1. PSLV-C19 lifting off with RISAT-1 on 26 April 2012 at05.47 h IST.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Orbit planningThe initial mass budget for RISAT-1 was 1725 kg aiminga SSPO, 627 km above the Earth. Later, the satellite masswas respecified to 1858 kg. Corresponding capability ofPSLV-XL was assessed for various feasible orbits and itwas decided to inject the satellite in 480 km circular orbitwith an inclination corresponding to 536 km SSPO mission,so that the orbit could be raised to 536 km using spacecraftpropulsion system.Accommodation of RISAT-1 within the payloadfairing of the launcherCentre, Sriharikota. At FLP, the rocket is built up on astationary launch pedestal near the umbilical tower usinga Mobile Service Tower (MST) which houses the entirelaunch vehicle during preparation and moves away on railsat the time of launch. Both the previous launches ofPSLV-XL variant took place from the Second LaunchPad (SLP), where the launch vehicle gets integrated on amobile launch pedestal inside a conveniently large stationarybuilding. Once the integration is complete, themobile pedestal carries the launch vehicle to the umbilicaltower of SLP. It was important to verify that MST atFLP could handle the S12 strap-on motors with adequateRISAT-1 was the biggest satellite to occupy the envelopeof payload fairings of PSLV. Initially, it was proposed touse a larger heat shield which demanded exhaustive studyon the aerodynamics of the altered configuration of thelauncher. Later, the RISAT-1 configuration was optimized(Figure 2) by swapping the locations of its payload(the specific instruments/equipment housed in the satellite)and the satellite bus (the standard structure of thesatellite), enabling accommodation of RISAT-1 wellwithin the existing proven payload fairings (Figure 3).This was a major preparatory step towards launchingRISAT-1 on-board PSLV (Figure 4).Assessment on launch padPSLV-C19 was the first PSLV-XL to be launched fromthe First Launch Pad (FLP) of Satish Dhawan SpaceFigure 3.Accommodation of RISAT-1 within PSLV heat shield.Figure 2.Final configuration of RISAT-1.Figure 4.RISAT-1 before closure of heat shield.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 473

SPECIAL SECTION: RADAR IMAGING SATELLITE-1assembly access. A trial of S12 strap-on motor assemblywas carried out at FLP, demonstrating the total feasibility.Figure 5 shows the S12 assembly in PSLV-C19.Another aspect examined in detail was the vehicle liftoffdynamics with respect to the physical gap between thePSLV-XL and the nearby ground structures (namely thelaunch pedestal and the umbilical tower) at FLP. Analysiswas carried out on the launch pad acoustics and thermalaspects as well. All parameters were confirmed to bebenign and acceptable.First polar mission of XL variantThough PSLV-C19 was the third launch of PSLV-XLversion, after the PSLV-C11/Chandrayaan-1 and thePSLV-C17/GSAT-12 missions, this was the first polarorbit venture of the variant. The trajectory of the vehiclewas designed such that all the constraints such as downrangesafety and environmental parameters (dynamicpressure, angle of attack, tail-off thrust of separating bodies,etc. during lower stage separation events) were met.With respect to acquiring the telemetry signal from thelaunch vehicle, analysis showed a visibility gap betweenthe down-range tracking stations at Thiruvananthapuramand Mauritius, due to the low injection altitude of 480 km.An additional tracking station was provided at RodriguesIsland (using transportation terminal) so that completetelemetry signal could be received in real time withoutany data loss.behaviour was influenced by the heaviest satellite attachedatop. Its impact on control capability of the vehicle had tobe well understood. Extensive vibration tests were carriedout with a stacked configuration of a flight-identical PS4stage with simulated satellite mass of about 2000 kgfor validation of the structural dynamics characteristics(Figure 6).It was observed that due to the heavy spacecraft thebase-fixed lateral frequencies were lower than the vehiclerequirements. In the vehicle stack level, the second andthird bending mode frequencies were lower from the control–structure–interactionpoint of view. Also, there was alocal bending in the yaw plane at inertial sensor areawhich could prompt unnecessary control action. Based onexhaustive analysis of the test results and a series ofreviews, the Digital Auto Pilot (DAP) design was finetuned.In the new design of DAP, compensators weretuned to provide extra attenuation to second bendingmodes during first stage flight regime in the yaw plane byreducing the rigid-body band width by a small margincompared to earlier missions. The changes were made forinitial flight regimes up to the burn out of strap-onmotors (named zone-1 and zone-2). A comparison of theUpper stage structural characteristics and itsimpact on autopilot designThe upper stage of the vehicle (PS4) was meticulouslylooked into with respect to the way in which its structuralFigure 5. Assembly of S12 strap-on motors in PSLV-C19 at firstlaunch pad.474Figure 6. Vibration test of PS4 stage with simulated RISAT-1 massfor dynamic characterization.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 1.Mission simulation: various levelsSimulation test bed6D simulationsOILS (On-board computers-In-Loop Simulations)HLS (Hardware-in-Loop Simulations).Actual inertial sensors in the loopALS (Actuator-in-Loop Simulations)PurposeTo validate trajectory, Closed Loop Guidance (CLG) and Digital Auto Pilot (DAP) designsin autonomous mode using the mathematical models for vehicle, control, navigation andenvironmental parameters.To validate the on-board software design. Stress tests are devised and carried out to ensureerror-free on-board software by perturbing all the parameters well beyond 3σ dispersions.To validate flight computers and Inertial Navigation Systems (INS) in closed loop bymounting INS on Angular Motion Simulator (AMS).In these simulations, the response and performance of the flight control actuators areassessed in the closed loop mode.compensator characteristics before and after the modificationsfor zone-2 regime (from 25 to 70 sec after ignitionof first stage) is shown in Figure 7.The process towards successFigure 7. Comparison of compensator characteristics before and afterdigital auto pilot modifications during zone-2 in yaw plane (first stageregime).All the regular and the new aspects of PSLV-C19 werethoroughly assessed by the launch vehicle design reviewteams concerned. All the mandatory mission simulationswere carried out with mission hardware such as on-boardcomputers, control electronics, actuators and also theassociated software on-board. Table 1 shows the variouslevels of mission simulations. Nearly 100 such simulationcases were exercised for PSLV-C19.The realized propulsion, mechanical, electrical, avionicsand software systems were all critically verified bythe mandatory mechanisms of Flight Readiness Review(FRR) in the system level and the launch vehicle level.Quality assessment and audit teams, again in system aswell as launch vehicle levels, confirmed the acceptabilityof all the flight elements.The elaborate testing of the launcher was carried outduring different phases of vehicle integration, supervizedby the Mission Readiness Review (MRR) team. On successfullaunch rehearsal (Figure 8) and confirmation ofall the health parameters, the final clearance for goingahead with the mission was issued by the LaunchAuthorization Board (LAB). PSLV-C19 lifted off withRISAT-1 on 26 April 2012 at 05.47 h IST.Post-flight analysisFigure 8. PSLV-C19 with RISAT-1 on the launch rehearsal day.PSLV-C19 injected RISAT-1 into a polar orbit with470 km perigee (distance from the Earth to the nearestpoint on the orbit) and 479 km apogee (distance from theEarth to the farthest point on the orbit) with an inclinationof nearly 97.6°. The orbit was well within the dispersionsspecified. The vehicle body rates at satellite separationwere less than 0.5°/sec as planned.RISAT-1 was later raised to 536 km SSPO using itsthrusters on-board. After the orbit raising, about 62 kg ofpropellant (out of 100 kg initially loaded) remained in theCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 475

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 2.PSLV-C19 flight profileTime (sec) Local altitude (km) Inertial velocity (m/sec)Event Flight Prediction Flight Prediction FlightStage-I ignition 0.00 0.02381 0.02381 451.89 451.89Stage-I separation 114.20 70.093 72.130 2147.31 2160.86Stage-II ignition 114.40 70.326 72.712 2146.45 2159.04Heat shield separation 155.20 116.235 117.938 2388.70 2387.36Stage-II separation 265.86 226.021 227.202 4098.79 4115.84Stage-III ignition 267.06 227.223 228.212 4096.11 4113.33Stage-III separation 512.72 434.574 446.450 5871.01 5881.90Stage-IV ignition 522.72 440.729 452.630 5862.01 5873.10Stage-IV cut-off 1027.74 485.340 485.390 7618.74 7616.00RISAT-1 separation 1064.74 486.143 486.200 7623.20 7622.83Table 3. Comparison of the first two global lateral modes with thepredicted valuesFirst mode (Hz)Second mode (Hz)Analysis Flight Analysis Flight2.11 2.25 4.72 5.01satellite because of the precise orbital placement byPSLV and the performance of the satellite thrusters,extending the satellite life further.Uninterrupted telemetry data was available as a resultof deployment of additional station at Rodrigues Island.The Post Flight Analysis (PFA) teams have gone throughthe telemetry data on various vehicle components such asthe propulsion systems, structural systems, avionics systems,control systems, separation systems, etc. and confirmedsatisfactory performance (Table 2).All the propulsion systems performed almost in thenominal range. The estimated specific impulse values ofthe stages, derived from trajectory matching, are withinthe dispersion bounds specified. No control–structureinteraction was seen during the first stage regime (withrespect to the effect of the heaviest satellite). All theseparation events were clean and imparted only negligibledisturbance to the ongoing stage.Benign vehicle loads and control force requirementswere seen from PFA. The aerodynamics parameters, viz.Mach number, dynamic pressure, relative velocity, altitude,angle of attack and Q-alpha during the first-stageregime were reconstructed from position, velocity andvehicle orientation angles in Earth centred inertial frameand also from wind profiles measured at T + 15 min(T denotes the lift-off time).The peak dynamic pressure observed in PSLV-C19flight was 69.77 kPa (upper limit 90 kPa) and the maximumangle of attack was around 0.84° (upper limit 3°)during the maximum dynamic pressure regime. Themaximum Q-alpha was 649.89 Pascal-radians (upperlimit 4000 Pascal-radians) at 60 sec into the flight.The effectiveness of DAP compensator tuning based onstage-level vibration test was assessed during PFA bycomparing the first two global lateral modes with thepredicted values and good matching was observed (Table3). Body rates showed no control–structure interactionduring the entire flight regime.ConclusionPSLV-C19/RISAT-1 mission was the 20th consecutivesuccess of the workhorse launcher of Indian Space ResearchOrganisation. The satellite was the heaviest andbiggest among the various spacecraft carried till date byPSLV. Typical aspects that called for closer attention andanalysis have been pointed out in this article. Also mentionedis the way in which each specific requirement wascatered to. The major findings from PFA have beensummarized. The professionalism and adequacy of launchvehicle management in terms of planning, preparationand pursuit are brought out.ACKNOWLEDGEMENTS. We thank all the System DevelopmentAgencies of PSLV in various Centres of ISRO who extended full supportin meeting all the special requirements with respect to PSLV-C19in terms of new extensive studies, trials, characterization tests, assessmentsand the specific attention on control aspect. We also thankthe senior executives of ISRO for providing valuable guidance towardsthe successful accomplishment of PSLV-C19/RISAT-1 mission.476CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Ground segment for RISAT-1 SAR missionV. Mahadevan 1, *, T. V. S. R. K. Prasad 2 , D. S. Jain 3 , Santanu Chowdhury 4 ,M. Pitchamani 2 and N. M. Desai 41 ISRO Satellite Centre, Indian Space Research Organisation, Bangalore 560 017, India2 ISRO Telemetry, Tracking and Command Network, Indian Space Research Organisation, Bangalore 560 058, India3 National Remote Sensing Centre, Indian Space Research Organisation, Hyderabad 500 625, India4 Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, IndiaRISAT-1 carries a C-band multi-mode SyntheticAperture Radar (SAR) operating in Stripmap, Scan-SAR and Sliding Spotlight modes and mainly caters tocivilian land applications related to agriculture, forestryand disaster management. Considering themulti-resolution and multi-polarization requirementsof RISAT-1 SAR, on-board programmability andflexibility in SAR payload as well as fairly autonomousoperation have been major mission requirements.The necessary intelligence and sophisticationhave been built into the on-board SAR subsystems tofulfill these essential requirements, apart from intelligentcontrol and coordination of active antenna ofRISAT-1 SAR payload.The complexity and large size of on-board SARinstrument have demanded a matching and equallyinnovative approach for ground segment operationsfor RISAT-1 mission. Since the launch of RISAT-1satellite by PSLV-C19 flight on 26 April 2012, the satelliteand SAR payload performances as well as missionand ground segment operations have been foundto be nominal and satisfactory. This article highlightsthe features and achievements of various RISAT-1ground segment systems and the activities carried outduring pre-launch, launch and Early Orbit Phase andnormal operating phases of RISAT-1 SAR mission,data reception chain, quick look processing, offlinedata product generation and dissemination chains.Keywords: Mission operations, offline data processingand dissemination, quick look processing, SyntheticAperture Radar.IntroductionRISAT-1 Synthetic Aperture Radar (SAR) is India’s firstindigenous, active, antenna-based microwave radar sensorin space. It is a C-band multi-mode spaceborne SAR withdifferent operating modes, viz. Stripmap, ScanSAR andSliding Spotlight and varying swath coverage of 10–225 km with imaging resolution of 1–50 m. It meets allthe basic civilian applications like agriculture, vegetation,forestry, flood mapping, disaster management, geology,*For correspondence. (e-mail: vm@isac.gov.in)ocean applications, pollution monitoring, etc. and feedsinto the database for India’s national natural resourcemanagement system 1 .A multi-mode system like RISAT-1 SAR generates alarge and variable volume of data, in view of differentswath coverage and resolution modes. Block AdaptiveQuantizer (BAQ)-based SAR data compression and variabledata rate formatting before SAR data transmissionthrough spacecraft X-band downlink add to the complexityof subsequent ground processing chain. The unique approachof overall SAR payload control and coordination operationsusing a separate on-board radar Payload Controller(PLC) has also resulted in the changes in the missionTelecommand and Telemetry Segment which interfaceswith the On-Board Computer (OBC) and Bus ManagementUnit (BMU) of the spacecraft.RISAT-1 satellite health maintenance and SAR payloadoperations are carried out from the Mission OperationsComplex (MOX) of ISRO Telemetry, Tracking andCommand Network (ISTRAC), Bangalore, using variousmission computers and associated mission software andcommunication links. The Telemetry, Tracking and Command(TTC) functions of the satellite in S-band are alsosupported by a network of ground stations. The recentlyoperationalized Integrated Multi-mission Ground segmentfor Earth Observation Satellites (IMGEOS) facility atNRSC, Shadnagar, Hyderabad Complex carries out theautomated execution of entire ground-processing tasksfor RISAT-1 mission beginning with SAR payload programming,data acquisition and SAR signal and imagedata processing (DP) to SAR raw data and data productdissemination with fast turn-around times (TATs). Unlikeoptical sensors, SAR image or data product generationinvolves elaborate pre-processing of SAR raw data aswell as complex, two-dimensional radar-matched filteringor focusing apart from other motion correction tasks, allof which have been implemented and operationalized bySIPA/SAC team. A Hardware Quick Look SAR Processor(HWQLP)/Near Real Time SAR Processor (NRTP)has also been built by the MRSA/SAC team at Ahmedabadand installed at IMGEOS, NRSC, Shadnagar.Considering the fact that RISAT-1 is the heaviest satellitelaunched by PSLV till date, elaborate and complexground segment systems and operations were put in placeCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 477

SPECIAL SECTION: RADAR IMAGING SATELLITE-1for successful operationalization of SAR payload. Startingfrom mission operations to SAR data product generation,many new systems and innovative techniques havebeen developed and established specifically for RISAT-1SAR. Thus, the complexity and large size of on-boardSAR instrument have demanded equally innovative andchallenging developments for ground segment operationsfor RISAT-1 mission. Since the launch of RISAT-1 on 26April 2012, the satellite and SAR payload performances aswell as mission and ground segment operations have beenfound to be nominal and satisfactory. The sections belowgive the details of various RISAT-1 ground segment activitiesrelated to mission operations and management as wellas data reception, processing and dissemination.RISAT-1 mission planning and managementRISAT-1 has been a challenging mission to operate withhigh power requirements of SAR payload, calling forjudicious battery and power bus management, thermalmanagement, high transmission data rates and the largeinertia of the satellite demanding manoeuvring capabilitieson the control system supported by high torque reactionwheels apart from the other bus subsystems. Thus,the mission planning and management involved manynew elements with the OBC providing extensive faulttolerantfeatures pre-programmed in the on-board memoryand supplemented by additional software programmedfrom the ground to provide safety features for in-orbitoperations besides taking care of the many observationson the various subsystems of the satellite in the groundcheck out and integration phase. The sections below detailthe planning aspects for the mission, summary ofmission operations and performance analysis.Orbit specifications478For RISAT-1 mission, a polar Sun-Synchronous Orbit at536.38 km altitude with an inclination of 97.554° waschosen considering the overall weight and power requirements,atomic oxygen effects on solar panels and thermalmaterials, atmospheric drag and systematic coverage requirementsof SAR payload. This orbital configurationgives a repetivity cycle of 377 orbits in 25 days with alocal time of 6 a.m. at descending node and path-to-pathdistance of 106 km, which ensures sufficient overlap forcontiguous swath even with the uncertainties due to attitudepointing and ground-track shifts. Global coverageis achieved twice in the repetivity cycle, both by thedescending as well as ascending passes, as SAR is amicrowave payload with no illumination constraints.A number of simulations had been carried out prior tolaunch to ensure smooth conduct of the various missionoperations, performance of communication links andnetwork stations. Operational procedures were exhaustivelyprepared to facilitate the end-to-end conduct of themission from lift-off to normalization of all subsystemsand commissioning of SAR payload in all its operationalmodes. Each of these operations involved online performanceevaluation of the system, sighting solutions andtheir implementation. Contingency procedures were alsoworked out in view of limited satellite-to-station visibilityfor operations. Activities spanning various workcentres called for extensive interfacing to ensure smoothcommunication flow. The required network supportduring various stages of the operations was worked outfor the commissioning phase qualifying them throughnetwork simulations prior to launch. A dynamic ‘softwarespacecraft simulator’, an important component in theoverall mission planning, was developed for testing andvalidation of ground segment facilities (networks) andoperation procedure, training of operations personnel, etc.Spacecraft initial phase operationsRISAT-1 satellite was launched on 26 April 2012 at 00 : 17 :05 UT (05 : 47 h IST) from the Satish Dhawan SpaceCentre (SDSC), Sriharikota launch pad. Immediately afterthe spacecraft injection into its polar Sun-SynchronousOrbit, the automatic deployment of solar panels and SARantenna deployments were carried out by the on-boardtimers triggered by the launcher and the initial acquisitionwas initiated over Troll ground station near South Poleusing the pre-loaded attitude Quaternions, followed bythree-axis attitude acquisition using ground commands.When the negative pitch axis of the spacecraft was madeto point towards the Sun, the solar panels were rotated togenerate the power and Earth Sensor and digital Sun Sensorswere switched ON. After the attitude convergence,Earth acquisition with yaw capture mode was commandedafter confirming Earth presence signal. In orbit-2 over thenetwork visibilities from Svalbard, Lucknow, Bangalore,Mauritius and Trolls, further activities for normalizationof the spacecraft were carried out. Wheels were switchedON and run at nominal control speed (3500 rpm) to getbetter dynamic friction estimation, which was subsequentlychanged to recommended nominal 1500 rpm in orbit-3.Both the star trackers were switched ON and normalizedto get star updates. GPS-based On-board Orbit DeterminationSystem (GOODS) was initialized after confirmationof the Satellite Positioning System (SPS) tracking thesatellites. Thermal heaters and auto temperature controllimits were fine-tuned with respect to on-orbit configuration.After confirming star sensor updates, the spacecraftwas put in normal mode with star sensor in loop followedlater by star Kalman filter mode. The safety featureson-board the spacecraft – hardware safe mode, wheelover-speed logic, spurious speed logic, auto reconfigurationlogic for wheels, failure detection logic of solar arraydrive, battery temperature control, etc. were enabled 2 .CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1PSLV-C19 launcher was required to place RISAT-1 at476 km altitude in view of the large mass of the satellite(1858 kg). The final injection orbit achieved was 470 kmand had benign dispersions of about 6 km less in thesemi-major axis and an inclination of + 0.06° with respectto the nominal injection parameters. A series (four) oforbit manoeuvres by firing five thrusters for cumulative1844.371 sec duration and utilizing 37 kg of on-boardfuel on the satellite ensured that the final mission orbit of536.6 km with inclination 97.562° was achieved, asplanned, within the first two days after launch.SAR payload commissioningRISAT-1 SAR payload commissioning-related exercisesstarted from 29 April 2012 onwards after the missionorbit of 536.6 km was reached. Unlike other satellites, forthe first time in RISAT-1, a single X-band carrier is beingused to transmit V and H polarization data in RHCP andLHCP modes. Thus, systematic characterization of theground reception systems was carried out with data handlingtests using RHCP mode alone in one pass, LHCPmode alone in another pass and then both together in yetanother pass. SAR payload commissioning started in aplanned manner by operating the payload in Fine ResolutionStripmap-1 (FRS-1) mode with single beam operationand then Medium and Coarse Resolution ScanSAR(MRS/CRS) modes with multiple beam operations. Thenear beam(s) and far beam(s) energizing exercises wereconducted for various modes and their power profileswere characterized in-orbit. About 27 test cases, includingon-board calibration operations were exercised duringthe in-orbit commissioning of SAR payload. Solid StateRecorder (SSR) was also commissioned with recordingand downloading of the PN sequence, followed by imagingsessions that required SSR recording. The High ResolutionSliding Spotlight (HRS) mode will be characterizedwithin the next few months, due to various satellitemanoeuvring operations involved.Eclipse operationsIn RISAT-1, due to 6 a.m./6 p.m. Sun-Synchronous536 km orbit, the orbital eclipse is only seasonal (2 May–12 August) with maximum eclipse duration of about22 min (around 23 June, when the Sun declination is23.5°), unlike other IRS missions where eclipse is encounteredin every orbit. Regular monitoring and managementof solar array and battery resources, and thermalcontrol of on-board systems are carried out to match withthese variable eclipse periods and seasonal variations.Daily uploading of the eclipse start time and duration tothe on-board systems is necessary for the initiation ofSolar Array Drive Assembly (SADA) auto capture in caseof panel non-tracking during sunlit period. During orbitaleclipse period, SAR payload (P/L) operation is avoided asthe full load of the payload along with the mainframe isrequired to be supported only by the battery. Similarly,during the solar eclipse time (lunar shadow on Earth)also, the SAR P/L operation is not planned. These operationaldisciplines are integrated in the P/L planning S/Wsystems itself.<strong>Special</strong> operationsAfter SAR payload commissioning through In-OrbitTests (IOTs), special operations for SAR payload calibrationwere also planned and executed. To perform externalcalibrations of SAR payload, Amazon rainforests areideal sites apart from other sites identified at Shadnagar,Nargoda, SAC–ISRO campus, Nalsarovar and Little Rannof Kutch, Gujarat. Apart from this, RISAT-1 also participatedin the oil-spill experiments organized by KSAT,Norway. The on-board SSR was also tested in playbackmode before using TROMSO station near North Pole forplayback of SAR payload data.Safety featuresAny SAR payload operation in RISAT-1 requires thesupport of both batteries and solar panels to meet itsoverall power requirements. There are software-basedchecks to ensure that the battery charge and solar powergeneration are sufficient to meet the payload and satelliteoperations. Based on the communication between SARPLC and OBC, SAR imaging is carried out only if thehealth parameters of all the payload subsystems are fine.OBC aborts the operation if any non-nominal condition isobserved in any of the payload or spacecraft elements.Since the payload operations require biasing of the satellite,which is performed with wheels, momentum ischecked continuously to ensure that it is kept within operationallimits. Remote programming to OBC throughground commanding is carried out to trigger event-basedoperations for many contingencies foreseen by systemanalysis and actions for recovery are also pre-programmed.OBC has features like Configurable Command Blocks(CCBs) and event-based programming, where CCBs canbe programmed from the ground to execute any given setof commands. In RISAT-1, CCBs were extensively used toswitchoff heaters during payload and switch them onagain after payload operation, for load management.Event-based programming is also performed to keep batterycharge within acceptable limits and switch off allsystems, except the minimum essential to keep up theattitude and allow for battery charging.RISAT-1 mission nominal operations and healthmaintenanceThe complexity of RISAT-1 mission entailed properplanning, execution and critical monitoring of on-boardCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 479

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 1. Organization of ground segment operations.payload subsystems. RISAT-1 mission control and healthmonitoring operations during pre-launch, Launch andEarly Orbit Phase (LEOP), initial and normal phases areprovided from the MOX, ISTRAC, Bangalore. Figure 1shows the functional organization of RISAT-1 groundsegment operations. The ground elements participating inthe mission operations and their interfaces are checkedduring the pre-launch simulation exercises. Mission scenariodata recorded during RISAT-1 satellite integrationand testing at ground checkout were used as test data duringsimulations. The simulation exercises includedautonomous tests, data flow tests, integration tests andfull dress rehearsals. RISAT-1 satellite being the first ofits kind had many new elements to meet the high powerand high data rate requirements of SAR payload. Thesatellite health monitoring included critical observationand analysis of status and performance of these subsystemelements during various payload modes.The analysis of launch trajectory revealed that therewas a requirement of station to fill the visibility gapbetween Trivandrum and Mauritius. A TTC TransportableTerminal (TT) at Rodrigues Island near Mauritius wasintegrated, tested, transported and operationalized successfullyfor the RISAT-1 launch support. SHAR1 andSHAR2 ground stations provided RISAT-1 satellite datamonitoring and command support during launch padoperations. The transportable terminal (TT1) at RodriguesIsland, Mauritius and TROLL station provided the LEOPoperations support during which the SNAP signal, autosolar panel deployment and initiation of SAR antennadeployment operations were monitored. TTC stations atBangalore, Lucknow, Mauritius, Biak and Svalbard providedthe normal phase operations of the satellite. Themonitoring and control system (MCS) at ISTRAC NetworkControl Centre (INCC) also provides the remotemonitoring and control of all TTC ground station equipments.480Figure 2 shows the TTC ground station configuration.ISTRAC TTC stations are equipped with antenna subsystemwith Transmit–Receive (TR) feed, TTC Processor(TTCP), Station Computer (STC), and Monitor and ControlSystem. The station is configured to support S-bandcarrier reception with polarization diversity mode forauto track and ranging functions. It receives both RCPand LCP signals simultaneously and combines them optimallybefore data detection. ISTRAC communicationnetwork provides the real-time voice/data/fax connectivityfor the mission operations between the MOX,Vehicle Control Centre, TTC stations, payload dataacquisition and DP centres. Communication is establishedusing satellite links, terrestrial links and dedicated fibrelinks. The pre-launch, launch and initial phase operationsare supported from the Mission Control Room (MCR)and Mission Analysis Room (MAR). The regular normalphase operations are being supported from DedicatedMission Control Room (DMCR). Network communicationsoftware in conjunction with the dedicated communicationlinks connects all the supporting TTC stationsthrough which the ISTRAC operations team also interfacesclosely with the mission team, subsystem designers,payload data acquisition and processing teams spreadacross various ISRO work centres. Scheduling networkstations and spacecraft and payload operations is animportant activity in multi-mission environment. Flightdynamics operations of the satellite include the orbitdetermination using on-board SPS data, generation anddissemination of orbital elements and orbit manoeuvreplanning.Spacecraft configurationThe Attitude Orbit Control System for RISAT-1 is configuredwith the 4π Sun sensors, Magnetometers, InertialReference Unit (IRU), Star Sensors, Earth Sensors,CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 2. Configuration of TTC ground station.Digital Sun Sensor and Solar-panel Sun Sensor. All theSun Sensors, Star Sensors and Magnetometers provideattitude information in the form of absolute attitude andattitude errors. The 4π Sun Sensor is used for Sun-pointingthe spacecraft during the initial acquisition after injectionand in the safe mode attitude holding. Magnetometers areused for updating the magnetic torquers, which in turn areused for the desaturation of momentum of the wheels.The star sensor is used as the prime attitude sensor forattitude measurement. Orbit information on-board is derivedfrom GPS receiver called SPS. Nominal attitudecontrol is through four reaction wheels with a capacity of0.3 Nm torque and 50.0 NMS @ 4410 RPM mounted in atetrahedral configuration about pitch axis. The OBC containsthe spacecraft control software as well as powerrelatedsafety logics. Its remote programming feature hasbeen extensively used for introducing additional on-boardsoftware patches into it. Thermal control of spacecraft isthrough 131 heaters operated in auto thermal controlmode with temperature limits as uplinked to the satellite.Due to the 6 a.m./6 p.m. orbit of RISAT-1, the solar panelslie in orbit plane. Sun tracking solar panels are drivenby unified SADA for autonomous acquisition, sensorclose loop mode and profile mode. Through micro steppingof SADA, the periodic disturbance to the spacecraftis minimized and the solar panels are also rotated alongthe roll axis during payload operations. The OBC drivesSADA to track the amplitude and phase profile. The threeminiature gyros in the configuration drive the spacecraftplatform with frequent attitude updates from star sensors.The on-board computed azimuth and elevation of theground station with respect to on-board Phased Array Antenna(PAA) is used to download the data. PAA has aField-of-View (FOV) blockage of +15° to +165° in azimuthand elevation of 82.5–100° due to SAR antenna.Through ground software, these blockage zones duringdata transmission requirement are estimated and appropriatelyeither real time or recording of imaging operationis planned. The Basedband Data Handling (BDH) mode isdecided based on the payload data rate and selected prior toimaging operation. SAR payload operations are controlledthrough an on-board PLC. For every operation, the SARpayload requires beam definition parameters, which arestored in remote uplinkable locations of OBC and passedonto PLC at pre-defined timelines. The on-board payloadsequencer complements the ground plan by providing thenecessary commands according to the time line for SARpayload, attitude manoeuvre and control, solar panel offsetand tracking, handling of varying high rate payload dataand management of on-board SSR and X-band systems.The payload sequencer is organized to carry out operationsin terms of sessions (max 128) and strips (max 16). A sessionis defined as the duration between nadir position to 36°roll bias and again return to nadir position of the spacecraft3 . Unlike earlier missions, the power generation patternof RISAT-1 varies over an orbit with variable sunaspect angle. Orbit average power estimations are usedwhile programming payload operations. To meet the highpower requirement, a set of thermal heaters is kept OFFduring payload operations. Pre-defined sequences in OBCCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 481

SPECIAL SECTION: RADAR IMAGING SATELLITE-1are used for thermal management of Payload and Batteryheaters during payload operation to meet the overallpower requirements. Power safe logics are verified at thestart and during every payload operation. In case of anypower eventuality, automatic payload abort sequence isinitiated to switch OFF the payload.Roll bias and attitude steering for zero DopplerSince the SAR payload is a side-looking radar, there is aroll bias requirement of ± 36° for left and right-lookingconfiguration. Also to nullify the Doppler due to Earth’srotation and the Doppler variations due to eccentricity ofthe orbit and oblateness of Earth, yaw and pitch steeringof the spacecraft have to be performed. The necessarysteering coefficients and bias commands are uploaded tothe satellite through ground commanding. Both yaw andpitch steering coefficients are computed on ground anduplinked to the satellite. The residual Doppler (50–150 Hz) can be estimated either from SAR raw data orusing spacecraft attitude data and corrected during processing.SAR payload operationsRISAT-1 SAR payload operations are carried out usingthe on-board payload sequencer by transmitting thecommands generated by the Command Sequence Generator(CSG) based on the Request file received from NRSCData Centre (NDC), Hyderabad by the Payload ProgrammingSystem (PPS). PPS is a ground-based operationalsoftware system to efficiently plan user imageacquisition requests and generate spacecraft payloadsequencing commands for imaging the area of userrequest. It also helps image maximum number of userrequested Areas of Interest (AOI) in pass-wise sequence,by arranging the user requests in one orbit and optimallyusing the spacecraft resources. PPS consists of severalmodules, viz. Proposal Generator (PG), Clash Checker(CC), Day-wise sequencer (Master Scheduler), Scheduleand Status Generator (SSG) and CSG. The various userrequirements for SAR P/L data acquisition are consolidated,prioritized and optimized for maximum number ofservicing in a day. Thus, PPS is utilized to generate theP/L operations on a given day, including SSR recordingoperations elsewhere in the world. These consolidatedP/L plan is sent to the Spacecraft Control Centre (SCC),ISTRAC, Bangalore for command generation throughCSG system. CSG is responsible for the generation ofconfiguration and timing information and beam parametersfor conducting SAR payload operations. SAR P/Loperation commands are up-linked one day in advance.Thus, PPS and CSG are important components of theMission Management System (MMS).482Health monitoringThe health monitoring and control of spacecraft constituentsis carried out through telemetry analysis and associatedtele-commands. Each of the telemetry sources hasbeen identified with a series of parameters and expertsmonitor these parameters to analyse the overall spacecrafthealth. Different techniques such as tabular display,graphical display, alarms, pictorial and mimic representationsin a single Graphical User Interface (GUI) window,developed using Spacecraft Health Monitoring and ControlSoftware (SCHEMACS) package, are utilized to presentthe spacecraft health parameters. For telemetryprocessing in real-time and offline, SCHEMACS core isorganized into three loosely coupled layers of data acquisition,data processing and data presentation. The commandgeneration and uplinking of these commands forcontrolling the spacecraft are taken up by the commandchain. The chain has various elements like command filecreators, verifiers, data command converters, telecommanddisplays, etc. to carry out different functions at differentlevels. The offline chain carries out archival,retrieval and presentation of data day-wise in offlinemode. Payload aux display developed on Linux gives theoverall picture of payload and active antenna (TR modules)health. Mimic display for SAR payload and SSR–BDH chain gives the pictorial representation of connectivity,interfaces and their behaviour in real-time andoffline. Web technology-based solutions enable thespacecraft health monitoring and analysis system usableand interoperable over different operating systems. Thiswould also allow the spacecraft subsystem experts fromremote stations/centres to access the spacecraft healthdata for analysis and operations with simple web browsersover the ISRO net. Figure 3 shows a snapshot of atypical SCHEMACS screen display page and display ofhealth monitoring parameters as well as mimic displays.Payload auxiliary data received through 2 Mbps data linkfrom HWQLP at Shadnagar, NRSC is also used to evaluatethe overall performance of SAR active antenna duringbuilt-in calibration and imaging operations.Ground reception, processing and disseminationsystemThe engineering challenges of acquiring and processingradar data from SAR sensor are complex and demandstate-of-the-art high computing infrastructure. The IntegratedIMGEOS implemented at NRSC, Shadnagar hasprovided the right platform to facilitate RISAT-1 datareception, processing and dissemination in an integratedmanner. The IMGEOS facilitated process re-engineeringof the entire data chain from payload programming todata dissemination with an integrated multi-missionapproach and has consolidated data acquisition andCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 3. Typical RISAT-1 health monitoring and mimic displays.processing systems with scalable architecture to cater tocurrent and future EO missions. It features a world-classdata centre with three-tier SAN, secured networks,improved product accuracy, enhanced user servicesthrough implementation of CRM and resolution-basedonline data ordering and dissemination. RISAT-1 is implementedin IMGEOS architecture and operationalizedfrom the first day of data acquisition. The RISAT-1 systemsconfiguration is worked out in accordance with theIMGEOS objective of integrated automated process flowfrom data reception to dissemination with combinedthroughput of 1000 products/day and TAT of 1 h for anemergency product and 24 h for a normal product.Data Reception SystemData Reception System (DRS) comprises four 7.5 mantenna systems with dual polarization configured inmulti-mission mode to track and receive data from anyremote sensing satellite. It is equipped with the state-ofthe-artbore-site facility for validating the data receptionchain both in local loop and radiation mode. Figure 4shows one of the RISAT-1 DRS chains configured underIMGEOS architecture. It comprises of Antenna andTracking Pedestal, Dual Polarized Feed and RF systems,Digital Servo and Automation system, IF and Base-Bandsystem and Data Ingest System. The composite S/X feedis dual circularly polarized in both S- and X-bands withthe capability to receive LHC and RHC polarized signalssimultaneously using frequency reuse technique. TheS-band Telemetry Data and Tracking signals are downconvertedto 70 MHz IF. The down-converted X- andS-band tracking IF signals are fed to a three-channelIntegrated Tracking System (ITS). The IF outputs fromfirst data down-converter (two carriers) and S-band dataIF are driven to the control room through a multi-coreoptical fibre cable and fed through programmable IFmatrix to the second down-converter and then to highdata rate digital demodulator. The data and clock signalsfrom high rate digital demodulators are driven throughLVDS interface to the data ingest system for further processingand product generation.The salient features of RISAT-1 DRS are as follows:• 7.5 m Cassegrain antenna system with G/T of32 dB/°K @ 5° EL.• Simultaneous RHC and LHC polarized signal reception@ 8212.5 MHz with dual polarized S/X-bandcomposite Feed using the frequency reuse technique.• Feed and front-end system realized single channelmono pulse tracking.• Two data reception chains at 720 MHz IF, each with320 MHz bandwidth.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 483

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 4. RISAT-1 data reception chain.• X-band auto track either through RHCP or LHCP carrier.• QPSK modulated RF carrier with 160 Mbps data rateeach in I and Q channels.• Synthesized up/down converter with additional channels.• IF link for transfer of high data rate modulated IFspectrums.• High data rate demodulators at 320 Mbps (I + Q) datarate.Figure 5 shows the software and hardware processingchains for data product generation. The major constituentsare described in the following sections.Data Ingest SystemIt consists of PC servers with RAID for real-time dataingest and PCI-Front-end Hardware (FEH) cards connectedto the demodulators. SAR data are acquired in oneor two streams in real time, at 320 Mbps data rate foreach stream and archived stream-wise and channel-wiseonto RAID. The raw data are then transferred to SAN innear-real time for level-O processing and product generation.Ancillary Data ProcessingAncillary Data Processing (ADP) system generates AncillaryData Interface File (ADIF), Browse and Formattedraw data (FRED). The ADIF is populated in the database484for subsequent access and pre-processing of data. Thebrowse images are automatically screened for quality witha provision for manual certification before releasing tointernet for user access. The FRED data are stored ontoonline data storage SAN for further processing.Data processingIn the IMGEOS environment, the DP Schedulers are classifiedinto optical, microwave and non-imaging categories.RISAT-1 products are routed to Microwave DPscheduler. The DP servers of each category operate inmulti-mission mode for optimum resource utilization.The data centre features state-of-the-art centralized threetierSAN storage configured with high reliability andredundancy for consolidation of data of all the satellitesand to facilitate online data archival and retrieval. Thecomputer systems are connected to SAN by FC and gigabitethernet for instant access and processing. The highendRISAT-1 systems with multi core and multi-CPU aresized and configured keeping in view the complex processing,storage and throughput requirements. The standardproducts with UTM/WGS-84 are generated everyday for all the passes and archived in product archives asoff-the-shelf products in addition to the user-orderedproducts.Workflow ManagerThe Station Workflow Manager software provides centralizedscheduling of all the antenna reception systems,CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 5. Configuration of software and hardware data ingest and processing systems.event monitoring and control functions for station operationswith appropriate interfaces for user order processingsystems. A centralized Event Monitor and Control (EMC)is integrated in the production chain for monitoring allthe events.Data dissemination and user servicesAll the processes involved in product generation areautomated right from online data ordering to data disseminationto user through FTP. The emergency productsare delivered within 1 h from the time of acquisition andnormal products within 24 h.User Ordering Processing SystemThe services provided through User Ordering ProcessingSystem (UOPS) are data browsing, ordering and futurecollections. Browsing and ordering service is prerequisiteinformation provided to the users for convertingthe required AOI into scenes and checking the data availabilityfor the required AOI. In case of RISAT-1, themeta information is populated to enable the user to verifycoverage. The browse facility has been integrated withdata ordering and PPSs. Different browsing options basedon map, AOI, path and date are provided for data search.Payload programmingPayload programming system (PPS) caters to the userneeds and for optimized utilization of on-board and groundresources. The requests for future data can be placed inadvance up to T-2 day (where T is the target day). All therequests from the users move automatically to the passplanning server. The clashes are resolved at PPS andschedules are generated and sent to ISTRAC, Bangalorefor command up-linking. The data are acquired accordinglyand provided to the user.SAR off-line DP and products generationThe off-line operational DP for RISAT-1 SAR is carriedout at NRSC, Shadnagar in an IMGEOS environment onsix SMP nodes with each node having four 8 coremachines.The basic steps of SAR DP can be summarizedas follows:• Block adaptive quantization decompression.• Correction for I and Q imbalance.• Doppler centroid estimation.• Range compression.• Range cell migration correction.• Azimuth compression.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 485

SPECIAL SECTION: RADAR IMAGING SATELLITE-1• Single-look complex or multi-look data generation.• Slant range to ground range conversion.• Geocoding.Figure 6 shows the basic data flow diagram for SARprocessor. The request for data product generation isingested through Data Product work flow managers. Masterand slave schedulers execute on separate hosts. Once awork-order arrives, the software automatically routes it toa free slave node and generates the outputs. The status ofwork-orders, viz. running, suspended, aborted, scheduled,error or completed for a particular scheduler session canbe known from GUI. The data products generation facilitycaters to Stripmap, ScanSAR and Spotlight imagingmodes of RISAT-1 satellite with the following productlevelspecifications.Raw Signal Products (level-0)This product contains raw SAR echo data in complex inphaseand quadrature signal (I and Q) format. The onlyprocessing performed on the data is the stripping of thedownlink frame format, BAQ decoded (optional) andre-assembly of the data into contiguous radar range lines.Each range line of data is represented by one signal datarecord in the RAW CEOS product. Auxiliary datarequired for processing are also made available alongwith echo data.Geo-Tagged Products (level-1)The image is geo-tagged using orbit and attitude datafrom the satellite. This allows latitude and longitudeinformation to be calculated for each line in the image.The earth geometry is assumed to be the standard ellipsoid.Each image line contains auxiliary information whichincludes the latitude and longitude of the first, mid andlast pixels of the line. The raw radar signal data are processedto provide SAR image data pixels. The image pixeldata are represented by a series of CEOS processed datarecords, each record containing one complete line ofpixels lying in the range dimension of the image. Theproduct can be obtained as slant range data (16 bit I and16 bit Q) or ground range (16 bit) amplitude data. Additionally,an auxiliary file containing a dense grid of geolocationsis associated along with the data file.Terrain-corrected Geocoded Products (level-2)This product contains terrain-corrected and geocodeddata. Provisions exist for UTM (default for systematicpojection) and polyconic map projection. The pixel spacingin the product will depend on SAR operating mode,number of looks and look angle. The options for productformats are CEOS and GEOTIFF.Figure 7 shows typical RISAT-I SAR data products forvarious SAR operating modes.Hardware quick look and near real-time SARprocessorsConsidering the growing user demand and inevitable necessityof real or near-real time SAR DP, the design anddevelopment of a HWQLP/NRTP was pursued as one ofthe mission goals of RISAT-1 ground segment. TheHWQLP/NRTP has been built, to the extent possible,using only Commercial-Off-The-Shelf (COTS) DigitalSignal Processor (DSP) and other hardware plug-in moduleson a Compact PCI (cPCI) platform. Thus, the majorthrust for the HWQLP has been on working out multi-DSP architecture and algorithm development and optimization.The HWQLP is currently installed at NRSCShadnagar, ground receive station with system configurationas shown in Figure 5, and is mainly used for dataFigure 6. RISAT data flow for SAR processor.Figure 7.RISAT-I data products for various SAR modes of operation.486CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1archival, SAR sensor performance evaluation and real/near real-time browse product generation 4 .RISAT-1 SAR image generation throughput requirementsare of high order and involve BAQ data decompression,two-dimensional complex signal compressionbased on radar matched filtering algorithms along boththe range and azimuth directions, as well as motion sensingand compensation tasks like Doppler CentroidEstimation and AutoFocus. The two-dimensional SARprocessing tasks are performed by DSP boards and theprocessed SAR images after mosaicing are displayed onthe monitor screen as well as stored on a suitable recordingmedia.The HWQLP/NRTP system is configured to operate inthe following modes:(a) Quick Look Processor (QLP) and Data ArchivalMode: QLP mode is exercised during the satellitepass over India. In this mode, RISAT-1 raw SARdata available from the Ground segment FEH aredirectly received by HWQLP and archived on dedicatedJBOD/RAID recorder and SAR image generationis accomplished in real time.(b) NRTP Mode: In this mode, the data archived duringthe satellite pass are played back from the archivalsystem to NRTP at a slower rate. In this mode,NRTP may utilize the ADIF/OAT files availablefrom the ground segment processing chain to generateprecision SAR images.(c) Payload Performance Evaluation (PPE) Mode: Inthis mode, HWQLP/NRTP system is used to evaluatethe performance of the RISAT-1 SAR payload,operating in calibration (CAL) mode. The payloadperformance evaluation includes data format verification,raw data statistics, antenna performanceevaluation, etc.For FRS-1 Stripmap mode and MRS/CRS conventionalScanSAR imaging modes, frequency domain range-Doppler algorithm is employed in HWQLP/NRTP. Individualsub-swaths of MRS/CRS are processed by zeropaddingin the burst gaps in azimuth direction and all thebursts are processed at once using a full-aperture matchedfilter. Therefore, the compression algorithm is similar tothat of continuous case and additionally includes scallopingremoval and range mosaicing. SPECAN or Derampingalgorithm is also being implemented in HWQLP/NRTP for MRS/CRS and FRS-2, which is quad polarizationand burst mode SAR imaging mode. Spotlight orSliding Spotlight SAR processing algorithm for HWQLP/NRTP is a variant of the range-Doppler algorithm withadditional processing steps like time-domain bulk RCMcorrection and reverse RCMC.The overall HWQLP/NRTP system consists of twounits corresponding to V and H receive chains of RISAT-1 SAR payload. Each HWQLP/NRTP system is configuredaround a 16-slot cPCI chassis with a host SBC andanalog devices 112 TigerSHARC, TS201S processors.Additionally, each HWQLP/NRTP has its own archivalsystem consisting of a cPCI recording blade along with 2TByte JBOD/RAID-based disk array. HWQLP/NRTPsystem caters to various RISAT-1 spacecraft data transmissionmodes, namely Real Time (RT) transmissionmode, stretch mode and SSR playback mode. SAR processingalgorithm and other software utilities for HWQLP/NRTP have been coded using VC++, Visual DSP++,Matlab and DSP assembly language. One of the challengingaspects in the design and development of theHWQLP/NRTP system is the development of real-timeDSP software for inter-processor communication betweenprocessing modules without the use of any centralizedReal Time Operating System (RTOS). The HWQLPsoftware is a fully automated system wherein it automaticallyreads the pass schedule file and configures itself forRT data archival and processing. Figure 8 shows a photographof HWQLP/NRTP system at NRSC, Shadnagar andsome of the results of RISAT-1 SAR images generatedusing this system.As mentioned earlier, the other off-line application ofHWQLP/NRTP is for the evaluation of sensor performanceof RISAT-1 SAR using calibration (CAL) modedata. A GUI-based software package called ‘RISAT-1SAR Payload Performance Evaluation (QLP/NRTP)’ inVC++ has been designed and developed. This softwareevaluates major auxiliary parameters related to SAR payloadsubsystems and spacecraft subsystems. It also measuresthe quality of video data received in each of thereceive chains and monitors the health and performanceof 288 pairs of TR modules used in the RISAT-1 SARactive antenna by computing the gain and phase responseof each TR module. The in-orbit gain and phase responseof each TR module is then compared with the previouslystored ground reference. This software also generatesdifferent TR module health information files which aretransferred from NRSC, Shadnagar over a dedicated link(2 Mbps) to ISTRAC, Bangalore for further analysis atthe Mission Control Centre.ConclusionsThe ground segment operations for RISAT-1 SAR missionare more complex and technologically challenging comparedto any other previous earth observation missionshandled by ISRO. It comprises of handling the multimodehigh power, variable and high data rate microwaveSAR payload and management of spacecraft on-boardmainframe systems during variable eclipse periods. Thepreparedness of mission operations planning, necessaryTTC and control centre hardware and the required missionsoftware ensured the smooth functioning of RISAT-1during launch, initial phase and normal phase operations.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 487

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 8. RISAT-1 SAR HWQLP/NRTP system and results.488CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1The major challenge in this mission was the synchronizedand coordinated execution of the various tasks forRISAT-1 SAR payload operation and the role of missionplanning, control and monitoring was therefore crucial.This apart, realization and successful operation of dualpolarization high data rate (640 Mbps) reception linksand quick look real/near-real time and off-line SAR DPchains, SAR payload programming, etc. have also beensignificant achievements.It is envisaged that these new technology developmentsare significant steps in standardization of ground segmentand mission operations for ISRO’s future spaceborneSAR missions like L-band SAR and X-band SAR. Thisapart, the involvement of indigenous Indian industries insome of these developments, has also been a significantachievement, which will help ISRO in faster execution ofsuch complex missions of similar complexity, nature andvolume.1. RISAT SAR payload-detailed design review documents, SAC/ISROinternal reports, February 2009.2. RISAT-1 mission operations plan, ISRO-ISAC-RISAT-1-PR-1652.3. RISAT Spacecraft Configuration Data Book, ISRO-ISAC-RISAT-1-PR-1663.4. Quick look/near real time SAR processor system for RISAT-1: detaileddesign review document. SAC/ISRO internal report, November2011, p. 92.ACKNOWLEDGEMENTS. We thank the Chairman, ISRO andDirectors of various ISRO Centres and Units for their continuous guidance,support and encouragement towards developmental activitiesrelated to RISAT-1 SAR mission in general and its ground segment, inparticular. We also thank the RISAT-1 project teams at various ISROCentres and other colleagues and staff members as well as privateindustries in India and abroad, for their support.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 489

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Initial results using RISAT-1 C-band SAR dataManab Chakraborty 1, *, Sushma Panigrahy 2 , A. S. Rajawat 1 , Raj Kumar 1 ,T. V. R. Murthy 1 , Dipanwita Haldar 1 , Abhisek Chakraborty 1 , Tanumi Kumar 1 ,Sneha Rode 1 , Hrishikesh Kumar 1 , Manik Mahapatra 1 and Sanchayita Kundu 11 Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, India2 Formerly with Space Applications Centre, Indian Space Research Organisation, Ahmedabad 380 015, IndiaImaging radars provide information that is fundamentallydifferent from sensors that operate in thevisible and infrared portions of the electromagneticspectrum. The Indian Space Research Organisation(ISRO) has launched a multi-mode, multi-polarizationSynthetic Aperture Radar (SAR) on-board Radar ImagingSATellite-1 (RISAT-1) on 26 April 2012. Variousdata products from RISAT-1 SAR are now goingthrough calibration–validation (cal–val) phase andsoon will be available for the global users for operationaland research purposes. In this regard, algorithmsare being developed to retrieve variousparameters in diverse application areas. This articledeals with the in-house algorithm development forstudying different resources using initial availabledata of RISAT-1.Keywords: Mangrove ecosystem, ocean surface wind,rice acreage, RISAT-1, ship detection, Synthetic ApertureRadar, wave spectra.Introduction*For correspondence. (e-mail: manab@sac.isro.gov.in)490RICE, a major staple food crop in India, is grown mostlyduring the kharif season. Extensive cloud coverage hindersoptical sensors to sense the crop growth, particularlyduring monsoon season. Thus SAR data remain the onlyviable option. Previous studies have demonstrated thatthe signature of rice is unique and dynamic when monitoredusing SAR data due to the presence of water backgroundin most parts of the country throughout majorityof its growth phase. Backscatter shows a steady increasewith time since transplanting until the heading stage andthereafter it maintains a constant value 1 . Results usingmulti-temporal ERS/RADARSAT imagery have confirmedthat C-HH backscatter can detect differences incrop type, crop growth stage and crop indicators. Crosspolarizedradar returns (HV or VH) result from multiplereflections within the vegetation volume and thus can addan extra dimension to the existing single polarizationmode. RISAT-1, the country’s first indigenously developedSAR satellite, has come up with huge potential tomonitor rice acreage. In the present study rice crop identificationand classification was attempted using RISAT-1C-band two-date data with central incidence angle 37°,HH/HV polarization. During initiation of the nationalkharif rice monitoring and acreage estimation project,rice signature has been developed from multi-temporalSAR dataset by Panigrahy et al. 2 . The model is suitablyused to study the rice crop of recently launched RISAT-1.The present study is an attempt to monitor the ricegrowingareas with the initial datasets (MRS, MediumResolution SAR) from this satellite, which will be theSAR workhorse in near future.Various oceanographic applications have shown thepotentiality of SAR images during the past three decadeswith the successful launch of a range of spaceborne SARs(SEASAT in 1978, ERS-1 in 1995, ERS-2 in 1995,RADARSAT-1 in 1995, ENVironmental SATellite(ENVISAT) in 2002, etc.). The ocean features commonlyseen on SAR imagery include surface waves, mesoscaleocean circulation structures such as eddies and currents,oil slicks, ships and wakes, internal waves and coastalbathymetry. The SAR is so sensitive to the interaction ofwind with the ocean surface that, in addition to windspeed, patterns and structures within the atmosphericboundary layer produce identifiable surface imprints 3 . Analgorithm has been developed for the retrieval of veryhigh resolution ocean surface winds, ocean wave spectraas well as detection of coastal and deep sea ships usingRISAT-1 SAR data.Sensitivity of radar to water, due to its high dielectricconstant, is extremely valuable to the remote sensing ofwetlands. It is not only sensitive to soil moisture, but canalso differentiate between moist soil and standing water 4 .High sensitivity to standing water and soil moisturemakes radar an efficient tool for determining hydro pattern5,6 . SAR technology with improved spatial resolutionallows regional wetland mapping. In a study by Baghdadiet al. 7 , it was suggested that C-band data are useful inwetland mapping and monitoring. Studies conducted withShuttle Imaging Radar (SIR-C) and Japanese EarthResources Satellite (JERS)-1 L–HH band imagery confirmedthis finding 8,9 . Slatton 10 had shown that polarimetricmultiband SAR has potential for mapping the majorsub-environments associated with coastal herbaceouswetlands. Discrimination of different types of marshes,swamp thickets and swamp forests is also possible withCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1radar imagery due to varying stem height and density 11 .The use of multi-temporal radar data combined via anintensity, hue and saturation (IHS) transformation hasalso been found to improve wetland mapping 12 . Theopportunity to explore the potential of radar data willimprove as many SAR satellites are currently beinglaunched or will soon be launched. These extra satelliteswill not only increase the amount of data available foranalysis, but will also increase collection frequency andthe variety of polarizations and frequencies available 13 .Forest inventory and monitoring have been routinelydone by satellite remote sensing techniques. While opticalremote sensing is widely used for this purpose, it isoften affected by sun angle and atmospheric parameters(e.g. clouds, haze and aerosols) and hinders quantitativemodelling. In this context, the active radar remote sensing,having particular significance for all-weather, day–night imaging capability and responsiveness to moisturecontent and structural attributes of vegetation, holdspromise for quantitative analysis in studies on forestry.One of the unique potentials demonstrated by spaceborneradar images is their penetration capability throughshallow sand cover in arid regions and detecting subsurfacegeological and archaeological features, in particularburied river channels and imprints of archaeological sitesoccurring in the adjoining areas. Images acquired innortheastern Sahara in 1981 during the first NASA ShuttleImage Radar mission (SIR-A) demonstrated the capabilityof the L-band (wavelength 24.5 cm) to penetrate1–2 m of loose sand and return information about geologicand geomorphologic features covered by sand 14,15 .Subsequently, radar images in L-band by Seasat, SIR-Aand SIR-B acquired in the Baden-Jaran Desert of China 16 ,in Saudi Arabia 17,18 and in the Mojave Desert of California19,20 confirmed the benefit of radar images to studybedrock features beneath few metres of loose sand.Analysis of Seasat, SIR-A and SIR-B images revealedigneous features (dykes) buried as much as 2 m beneathalluvium in the Mojave Desert 19 . The radar energy haspenetrated through the shallow sand cover and subsurfaceigneous dykes have acted as subsurface rough surfacesand provided high backscatter to be detected by the radarsensor. Ancient drainage pattern cut in the bedrock is notvisible in the optical images because of the sand cover.The buried palaeodrainage network appears as dark featuresdue to smoother channel fillings and is identifieddue to contrasting brightness of hard substrate of theadjoining region 14,15,21–24 .Kharif rice acreage using two-date RISAT-1 dataDataset and study areaRISAT-1 C-band (5.35 GHz) dual-polarization (HH, HV)MRS L2 data (18 m pixel spacing and 37° incidence angle)are used for estimating kharif rice acreage estimation.Data have been collected for two different dates sufficientlyspaced in phenology to identify rice fields. Spacingbetween the data is minimum of 12 days for the Odishacoastal area and maximum 50 days for Allahabad. Detailedinformation about the dataset used is given in Table 1.Rice is predominantly grown in the entire Indo-Gangetic plain during monsoon season. Uttar Pradesh,Bihar, Odisha and West Bengal are considered as the ricebowl of the country. Two-date scenes having overlap inmajor rice-growing regions of these states are selected asthe test area (Figure 1).MethodologyProcessing of the RISAT-1 data was carried out usingPCI Geomatica (ver. 9.0) software. L2 product is availablein .tiff format. HH and HV images are imported tonative raster format of PCI Geomatica (.pix) and transferredto a single file as two separate channels. Reprojection(if needed) is done and then the image is filtered byenhanced Lee adaptive filter with kernel size 5 × 5, asestablished by standard procedures 1 . Most of the sceneshad abrupt brightness variations in the near and far rangeside and banding at equal interval along the track due tomosaicing of data of beams used to form the MRS data.These unwanted patterns were removed by marking atransect across the image and adjusting the brightness byplotting the variations in brightness across the pixels. Thecoefficient and central pixel values are used to correctthis pattern. Calibration for HH and HV polarization amplitudeimage were carried out using the calibration constantin product.xml as:Value in dB (in 32-bit real channel)= 20 × log10 (DN) – calibration constant.The two-date images were co-registered by fitting manualGCPs with about 25 well-distributed points throughoutthe image scene. A second-order polynomial model wasfitted to register the second date image with respect to thefirst date image.On this two-date HH–HV image, the vectors in theform of sample-segment or district boundary were overlaid.Classification of the rice area is performed insidethese segments to derive the crop proportion and finallydistrict rice acreage. Addition of HV channel takes careof the fallow and river border region otherwise classifiedas rice using two-date HH data.ObservationRice, being a semi-aquatic crop, generates unique temporalbackscatter profile unlike other crops. During transplantation,backscatter is quite low due to puddled fieldsin standing water condition which reflects more energy inCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 491

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 1.Details of RISAT-1 scenes used for the studyIncidence Temporal PixelArea Date Polarization angle interval (days) spacingAllahabad, Uttar Pradesh 17 July 2012 HH 35.09 50 18 × 185 September 2012 HV 36.58Baleswar, Odisha 2 August 2012 HH 35.04 12 18 × 1814 August 2012 HV 37.49Aurangabad, Bihar 18 July 2012 HH 34.30 38 18 × 1826 August 2012 HV 36.56Figure 1. Area overlapped by two-date RISAT-1 scenes.forward direction and minimal in the backscatteringdirection. Only HH data-based assessment for rice fieldalso includes waterbodies and moist fallow lands ascommission error. The overestimated area can be removedby introducing HV data with previous set. In many cases,it has been noticed that signature generated from HHbackscattered coefficient (in dB) of rice fields has beenmixed with waterbody or moist fallow land in the borderingfields, whereas HV backscattering for these featuresdiffers. HH–HV difference is higher for cropped fieldscompared to fallow land or waterbodies. Two-date datasetin Allahabad, Aurangabad and Odisha areas can successfullyidentify early and late transplanted rice cropwith combination of HH and HV data.In two-date FCC, early transplanted rice fields appearas bright green patches. During first-date data acquisition,as transplantation had taken place, rice fields had sufficientlylow amount of backscattering which had subsequentlyincreased in second-date acquisition. For late ricetransplanted fields, there is a sudden dip in backscatteredcoefficient in the second date. So, energy available ingreen and blue channel is less and it appears as bright redpatches in the images (Figure 2).492Figure 2. Two-date RISAT image covering a part of rice-growingarea of Aurangabad district, Bihar. a, Overview of two-date overlappedarea in FCC (R: First date HH, G: Second date HH, B: Second dateHV). b, Rice sample segments. c, Classified rice as early (yellow) andlate (magenta) transplanted.Sample segments are used to find the crop proportionof the study region. Rice is classified within the agriculturalsample segments. Crop proportion is calculated bytaking ratio of total number of pixels and total number ofpixels classified as rice. Crop acreage is estimated bymultiplying this value with total population of each district.The procedure can be summarized as follows.Crop proportion = Pixel classified as rice withinsample segments/Total number of pixels withinsample segments.Acreage = Crop proportion × N × Segment area insuitable units.where N is the total population of agricultural segments.Crop proportion for early transplanted rice in Aurangabadarea is recorded as 0.094. On the other hand, it isCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-10.30 for late transplanted rice. Misclassification due tomoist soil and waterbodies in two-date single polarizationdata has been solved by adding HV polarization in bothdates. HV is more sensitive to above-ground volume andhence backscattering differs. Due to late onset of monsoon,most of the area has transplanted rice lately, whichclearly showed up in the second-date data.Dataset for coastal Odisha comprises one ascendingpass and another descending with 10 days temporal interval.Though the interval is too small to identify early andlate transplanted rice, the SAR data successfully pickedup rice fields because majority of transplanting operationscoincided with these dates. Crop ratio for this areais estimated as 0.10 and 0.31 for early and late transplantedrice respectively. The FCC composed by two-dateHH polarization shows a similar picture of the study area(Figure 4).In the Allahabad scene (Figure 3), analysis shows thatmost of the rice-growing areas have early transplantedrice, only few fields have sown it late. Crop proportion ofearly transplanted rice is estimated as 0.30 and for latetransplanted rice it is 0.08. Two-date HH dataset was alsoanalysed and results have been compared. Overestimationof late transplanted rice is clearly identified, which estimatedcrop ratio of late transplanted rice as 0.15. Thelarge data gap (50 days) between the two dates renderedsome difficulty in picking up the late transplanted paddyas it was missed in the 10 August scene. The fully-growncrop was picked up in the third date, but critical transplantationphase for some parts was missed.Figure 3. a, classified rice sample segment in Allahabad region. b,classified rice in coastal Odisha.Figure 4. Comparison between two-date HH FCC of RISAT-1 andRadarsat-2.The present study is a preliminary attempt to utilize theRISAT-1 data for kharif rice monitoring, which was thelargest national project using microwave data from a foreignsatellites till the launch of RISAT-1. The feasibilityof RISAT-1 data for monitoring the national rice acreagehas been demonstrated in this study. Suitable datasetsfrom RISAT-1 would improve classification of the presentstudy and also boost the monitoring of other kharifcrops in future. The low data cost would be a boon for themicrowave remote sensing community globally in allarenas of application.Oceanographic applications of RISAT-1 SARDataIt is well known that for oceanographic applications wideswath (scanSAR)-mode data are best suitable. Hencein-house development of all the retrieval algorithms hasbeen carried out using Wide Swath Medium resolution(Wide Swath Mode, WSM) data from Advanced SAR(ASAR) on-board ENVISAT. The algorithms are testedon Fine Resolution (FRS) and Medium Resolution(MRS)-mode data from RISAT-1 SAR (due to nonavailabilityof wide swath data from RISAT-1 while preparingthis manuscript). Brief descriptions of the variousdata used in this study are given below.(1) ENVISAT–ASAR: ENVISAT was launched by theEuropean Space Agency (ESA) on 1 March 2002 in aSun-synchronous orbit of altitude 799.8 km with inclination98.550°, as a successor to ERS-1 and ERS-2. The orbitalperiod of ENVISAT is 100.59 min with a repeatcycle of 35 days and varying imaging frequency from1 to 3 days. ASAR is one of the ten payloads carried byENVISAT. ASAR uses in-phase array with an incidenceangle range of 15–45° and is able to operate in five differentpolarization modes (VV, HH, VV/HH, HV/HH,VH/VV) in C-band (5.3 GHz, 5.6 cm). Among the fiveoperating modes of ASAR (Image Mode, AlternatingPolarization, WSM, Wave Mode and Global Monitoring),only the WSM, using the scanSAR technique, offers awide enough swath (around 400 km) with a spatial resolutionadapted to accurate regional monitoring (with apixel spacing of 75 m). The incidence angle in eachimage ranges from 17° to 42°.(2) RISAT-1 SAR: RISAT-1 was launched on 26 April2012 by ISRO in a Sun-synchronous dawn–dusk orbit ofaltitude 536 km with inclination of 97.55° and orbitalperiod of 95.5 min. RISAT-1 carries a C-band (5.35 GHz)SAR as the sole payload. The RISAT-1 SAR is capableof imaging the Earth’s surface in different modes, e.g.HRS (spotlight scanning, resolution less than 2 m), FRS-1 (Stripmap scanning, resolution 3 m), FRS-2 (StripmapScanning, resolution 6 m, quad-pol capability), MRSCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 493

SPECIAL SECTION: RADAR IMAGING SATELLITE-1(scanSAR scanning, resolution 25 m) and CRS (scanSARscanning, resolution 50 m) with variable swath widthsranging from 30 to 240 km.Methodologyimagery, a single SAR image frame is extracted. In caseof RISAT-1 FRS-1 mode, such a frame is comprised of512 × 512 image pixels. Since each pixel represents a3 m × 3 m area, the image frame corresponds to 1.5 km ×SAR achieves its azimuth resolution by coherent processingof radar returns. Due to this coherent processing,noises (speckles) appear on the SAR images. Before theprocessing of SAR images for any retrieval purposes,such speckles have to be removed. In our methodology,speckle removal is performed using a 5 × 5 gamma MAPfilter. In case of coastal SAR images, the land portions ofthe images are to be masked so as to eliminate land contaminationin the resultant product. We do this using30 arcsec global topography (GTOPO30) data availablefrom the US Geological Survey (USGS). At this end theSAR image is ready for retrieval of oceanographicparameters.Retrieval of ocean surface winds: Over the oceans, theroughness explicitly depends on the winds blowing overthe surface. Greater the wind speed, the higher is the surfaceroughness and so is Normalized Radar Cross Section(NRCS) 25 . The variation in surface roughness causesvariation in backscattered power and hence in SAR imageintensity, which is directly proportional to the NRCS fora calibrated SAR image 26 . Both the SAR images fromENVISAT and RISAT-1 are calibrated using the methodproposed by Rosich and Meadows 27 considering theeffect of SAR digital counts, external calibration constantand local incidence angles. The calibrated SAR image isthen inverted to obtain the wind speed using a C-bandmodel function (CMOD5) and using auxiliary wind directioninformation from numerical weather model (NationalCentre for Environmental Prediction, Global DataAssimilation System, NCEP-GDAS) 28 . This auxiliaryinformation of wind direction is necessary to invertCMOD5, because single-antenna SAR cannot directlymeasure wind direction. All the CMODs are developedbased on VV polarization SAR images. However, thewind retrievals from HH polarized images are performedusing an azimuth and incident angle-dependent parameterizationfor the effective polarization ratio as givenby Mouche et al. 29 . Figure 5 shows SAR images fromENVISAT and RISAT-1 and their corresponding retrievedwinds. Figure 6 shows the validation results ofretrieved wind speeds from ENVISAT. The validationresults in Table 2 show that the accuracy of the output ofthe algorithm is suited for operational use 30 .Retrieval of ocean wave spectra: At typical incidenceangle between 20° to 70°, SAR interacts with the oceansurface through Bragg’s resonance 31 . Within this incidenceangle range ocean wave spectra are retrieved fromSAR images. From the speckle-free SAR intensity494Figure 5. Wind retrieval from (a) coastal winds from ENVISATASAR (output resolution 975 m) and (b) open ocean winds fromRISAT-1 SAR (output resolution 900 m).Figure 6. Validation of ENVISAT ASAR-derived ocean surface windspeeds with OSCAT, TMI, SSMI and JASON-2. Table 2 shows therespective validation results with OSCAT, TMI, SSMI, Jason-2 andtheir statistics.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 2.Validation results with OSCAT, TMI, SSMI, Jason-2 and their statisticsData No. of points Bias (m s –1 ) Standard deviation (m s –1 ) ROSCAT/ASAR 629 0.87 1.17 0.94TMI/ASAR 1641 0.48 1.51 0.89SSMI/ASAR 634 0.38 0.99 0.93Jason-2/ASAR 217 0.04 1.33 0.88(OSCAT + TMI + SSM/I + Jason-2)/ASAR 3121 0.35 1.40 0.91Figure 7.Retrieval of ocean wave spectra from a subset of RISAT-1 SAR image 952F1_S58_RV.Figure 8. Ship detection from RISAT-1 SAR image 21343_VV. The adjacent table shows the numberof detected ships and their corresponding geo-location for the subset.1.5 km patch on the ocean surface. The frame size providesa sufficiently large area that at least 10 cycles ofvery long surface waves, up to 150 m in length, can beincluded in a single frame. At the same time, the framesare also small enough that the ocean surface can beassumed homogeneous within a frame.The mean intensity is subtracted from the image frameand the resultant is normalized by dividing by the meanintensity. This image of fractional modulation is thenFourier transformed and squared to produce image intensity-variancespectrum as a function of azimuth and rangewavenumber. The image is then corrected for stationaryresponse.Since the resulting image has only two degrees of freedom,the value of the spectrum at each two-dimensionalwavenumber bin is a noisy representation of the underlyingspectrum. Hence from the 2D spectrum, 20% of theensemble average is subtracted after convolving with aGaussian-shaped kernel to remove the noise 32 . After this,the 2D spectra are divided by magnitude of modulationtransfer function (MTF) that accounts for the modulationof capillary wave by underlying gravity waves 33 . Figure 7shows a RISAT-1 FRS-1 image and retrieved wave spectra.Ship detection: Real-time detection of coastal as well asopen ocean ships using SAR images provides a valuableaid for building a space-based surveillance system. Fromthe SAR imagery ships can be detected by means of theirintensity contrast relative to the immediate background.In a SAR image, a 20 × 20 pixels window is selected. Ifthe intensity of a pixel is greater than the threshold (computedby the mean and standard deviation of 400 pixelsCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 495

SPECIAL SECTION: RADAR IMAGING SATELLITE-1within a single window and assuming a constant falsealarm rate, CFAR) of the window enclosing the particularpixel, the pixel is detected as a ‘potential ship pixel’ 34 .The window is then moved through the whole SAR imagesearching all the potential ship pixels. When the searchingoperation is completed, ships are identified based on theship pixels using a scanning-based ship cluster algorithm.Figure 8 shows ship detection results from one RISAT-1image. The algorithm has been validated using informationfrom Director General, Shipping Corporation, Mumbaiand found to be very sound for operation purposes.Future scopeThis article discusses the algorithm development forretrieval of high-resolution ocean surface winds, oceanwave spectra and ship detection from SAR images foroperational utilization. All these algorithms were developedusing ENVISAT ASAR data and then implementedon RISAT-1 SAR data. Since RISAT-1 is now commissionedfor cal–val phase, much of RISAT-1 data couldnot be used, particularly wide swath data. After the completionof cal–val phase, more data will be available tothe users and then extensive validations of these algorithmsare planned. Also, several new aspects of RISAT-1 SAR, e.g. circular polarization, etc. can be utilized forfurther improvement of these algorithms.Analysis of structural components of Wular Lake –Ramsar Site, India – based on RISAT-1 dataStudy areaThe study area – Wular Lake, Jammu and Kashmir (J&K)is an internationally important wetland under the RamsarConvention. The lake, along with the extensive marshessurrounding it, is an important natural habitat for wildlife.It is also an important habitat for fish, accounting for60% of the total fish production in J&K. The lake is asource of livelihood for the large human population livingalong its fringes. Encroachments resulting in the conversionof vast catchment areas into agriculture land, pollutionfrom fertilizers and animal wastes, poaching ofwaterfowl and migratory birds and weed infestation arethe main threats to the wetland.The objective is to study application of initial set of SAR(C-band HH/HV) data from RISAT-1 for extraction ofstructural components of Wular Lake by way of delineationof open water and various vegetation types/densities.The satellite includes RISAT-1 (C-band HH/HV, MRSwith 35.07° incidence angle and 18 m pixel spacing) georeferenceddata of 12 July 2012. The field data comprisesvegetation-related information.MethodologyThe methodology involves a standard approach wherein:• The data were subjected to speckle removal.• The GPS-aided ground truth information is convertedinto a spatial point-layer with the updated attributes offield data.• Classification of the SAR data after defining the trainingclasses based on ground truth as stored in the spatiallayer (point layer).ResultsRISAT-1 data comprising C-band HH/HV with 35° incidenceangle and pixel spacing of 18 m in MRS mode hasshown reasonably good geometric fidelity and uniformityin contrast without much speckle (Figure 9).The Wular Lake has an area of 11,377 ha (Table 3).The classified image shows that Trapa natans and Trapabispinosa are dominant in the wetland, occupying an areaof 4427 ha. Due to lower density, water beneath Phragmitescommunis could be delineated which accounts for1206 ha area. Fringes of the wetland are observed to havebeen planted or have natural vegetation comprising willow,of two density classes – medium density (1543 ha)and high density (1767 ha). Paddy fields and fallow areobserved on the fringes, which together with willow indicateencroachment into the wetland. Details of aerialextents of various classes are given in Table 3.Preliminary decision rule classification ofmangrove ecosystem using single-date VVpolarization SARMangrove forests are found in the intertidal zones oftropical and subtropical coastlines, and exist as an ecosystemcomprising estuaries, lagoons, creeks andObjectives and data used496Figure 9. (a) Wular Lake RISAT-1 SAR image (C-band HH/HV) of12 July 2012 and (b) the corresponding classified image (colour codeswith their associated description are given in Table 3).CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Table 3.Area under various structural components in Wular LakeColourcodeClass descriptionArea(ha)Open water 580Low density Trapa (Trapa natans and T. bispinosa) 2,249Medium density Trapa (T. natans and T. bispinosa) 809High density Trapa (T. natans and T. bispinosa) 1,369Medium density willow 1,543High density willow 1,767Water beneath the emergent vegetation (mainly Phragmites communis) 1,206Habitation 947Paddy fields 168Fallow 739Total 11,377of studies on radar remote sensing has been reviewed inthe field of mangroves 38 . Though most of these studiesgive immense knowledge in understanding radar interactionwith mangroves, one rarely finds large area applicationsleading to operational forestry requirements. Studiesregarding the characterization of Indian mangrove ecosystemsusing SAR are minimal. The objective of thepresent work is to study the usefulness of single-date VVpolarization C-band SAR for broad classification of mangroveecosystem.Study area and datum usedFigure 10. LISS III false colour composite (FCC) showing IndianSundarbans as the dark red zone. Yellow colour depicts the boundariesof the districts. The hollow rectangle overlaid on the FCC shows thestudy area.intertidal mudflats 35 . The mangrove forests consist ofspreading trees with numerous arched aerial roots andpneumatophores. The system is sensitive to changes inthe local hydrological environment, and the changes aretypically manifested through alterations in their species/communitycomposition, structure and biomass.Essentially, the backscatter coefficient of a mangrovecanopy depends upon the interaction of microwaves withleaves, branches, trunks, above-ground roots/pneumatophoresand the underlying mud/water. Thus, apart fromsensor parameters (polarization, frequency and incidenceangle), the type and structure of mangrove vegetation affectthe backscatter 36 .Studies have already demonstrated that microwavescattering and attenuation in C-band SAR result from interactionswith tree canopy and small secondary branches, butC-band backscatter from tree trunks is small due tominimal canopy penetration, resulting in a larger amountof signal absorbed and less signal returned 37 . A summaryThe study area is Dhanchi Island (Figure 10), lyingbetween 21°36′N–21°43′N lat. and 88°25′E–88°28′Elong. and located in the confluence of the Thakuran Riverin the east, the Jagdal Ganga in the west and the Bay ofBengal in the south. There are two prominent creeks inthe island; one originating from the east and the otherfrom the west.RISAT-1 Fine Resolution Stripmap Beam (FRS) datumhaving an incidence angle of 17.84° was utilized. Singledatedatum acquired on 10 May 2012 of ascending mode(around 1709 h local time) was used. The datum characteristicsare provided in Table 4. The acquisition time ofthe image corresponded with low-tide conditions in theisland.MethodologyThe processing of the single-date RISAT-1 datum for thisstudy involved downloading, speckle reduction and calibration.Mangrove ecosystem is composed of mangrove forests,creeks/channels and mudflats. Thus, as a first step, onehas to discriminate mangrove forests from these associatedclasses. Mangroves being evergreen forests, the canopyundergoes changes during the phenological eventsof flowering and fruiting 39 . Thus, one can expect littledeviation in backscatter signature during a particular season.However, the intra-class variability within mangroveforests is significant, mainly due to species composition,density, age and gradient from tidal influx.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 497

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Single polarization (VV) SAR datum was used and thesignatures of the different components of mangrove ecosystemof the island were noted. Based on the backscattersignatures and ancillary information, a knowledge basewas developed which culminated into decision rules thatwere used for classification. The training sets wereselected using the mangrove ecosystem map of NationalWetland Inventory and Assessment, showing mangroveforests, creeks and intertidal mudflats 40 . Considering theintra-class variability within mangrove forests, care wastaken to consider training classes covering these variabilities.These categories were grouped into sub-classes. Foreach class training samples were appropriately selected.A decision rule algorithm was developed based on 65%of training class signature and validated on the rest of35% samples. A mask of the mangrove region was createdusing available Sundarban maps. Within this maskdifferent ranges of amplitude backscatter values of single-dateVV polarization were used for rule development.Using macro language of EASI/PACE, the decision ruleswere generated to classify the single-date image.high backscatter (Figure 11). The mangrove forests registeredbackscatter values ranging from –12 to –7.5 dB.Preliminary decision rule classificationThe mangrove ecosystem of the island could be classifiedinto mangrove forests, intertidal mudflat and creeks/channels (Figure 12). The mangrove forests could bedivided into two broad classes, viz. dense and sparsemangrove based on canopy closure (< 40% closure forsparse mangrove and > 40% for dense mangrove). Thus,dense mangrove corresponded to very dense and moderatelydense forests, and sparse mangrove to open forests,scrub and non forest zones of the forest cover classificationscheme provided by Forest Survey of India 41 . Sincelow-tide conditions prevailed during the image acquisitiontime, mudflats could be demarcated along both theeastern and the western creeks and several other minorResults and discussionVV image and corresponding backscatter valuesIn the image intertidal mudflats appeared dark due to lowbackscatter and creeks/channels appeared bright due toDatum characteristicsTable 4.Datum characteristicsSpecificationProduct formatRISAT-1-GeoTIFFSAR band and polarizationC band, VV polarizationBeam modeFRS1Sampled pixel spacing (m) 9 × 9Number of looks (range × azimuth) 2 × 4Incidence angle (degree) 17.84Pass directionAscendingMap projectionUTMDate of acquisition 10 May 2012Acquisition time (local)Around 17 : 09 h (IST)Figure 12. Preliminary decision rule classification of mangrove ecosystemapplied on the mask of mangrove region of Dhanchi Island onVV image.Figure 11. Mean backscatter values with standard deviation bars (ofthe population means) of VV polarization of different mangrove ecosystemcomponents of Dhanchi Island.498Figure 13. Proportion of different classes of mangrove ecosystem ofDhanchi Island.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 14.Ground photographs of different classes of mangrove ecosystem of Dhanchi Island.island was of the order dense mangrove > sparse mangrove> intertidal mudflats > creeks/channels (Figure 13).Major portions of the sparse mangrove occurred in thearea above the eastern creek (Binidar creek) and in thearea between the two creeks. Figure 14 shows groundphotographs of the different classes.Potentials of RISAT SAR in detectingpalaeo/buried channelsFigure 15. Parts of the palaeochannel of the lost Sarasvati aroundAnupgarh, northern Rajasthan as seen on (a) RISAT SAR acquired on12 July 2012; (b) Landsat ETM FCC and (c) FCC of RISAT SAR andLandsat ETM merged product.creeks and channels. Mudflats could be also discriminatedalong the eastern coast of the island. The proportionof the different mangrove ecosystem components in theThe region of northwestern India (covering the states ofPunjab, Haryana, Gujarat and Rajasthan) and floodplainsof River Indus and its tributaries in Pakistan aregeographically diverse, geologically active and rich inarchaeological sites of Harappan Civilization (2500–500 BC). In the past few decades a large amount of workhas been carried out to map palaeo/buried channels in thisregion using multi-sensor satellite data, including radardata and understand their migration and evolution 42–50 .These studies have shown evidence of a prominent riversystem, which has become buried under sand cover of theThar Desert sometime during late Holocene. This majorriver has been identified as Sarasvati, a legendary rivermentioned in ancient Indian texts. Late Quaternary climaticchanges and neotectonics have significantly modifiedthe ancient drainage courses in the region and it isdifficult to interpret buried/palaeo channels due to presenceof vast spread of aeolian sand cover in opticalimages. Sahai 51 , and Rajani and Rajawat 52 have attemptedto understand spatial distribution and number of Harappansites along the identified courses of the River Sarasvatiby superposing location of Harappan sites on thepalaeochannels of the River Sarasvati in a GIS environment.The study could demonstrate that the clusters ofHarappan settlements from the mature period to laterperiod have moved in the same direction as the migrationCURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 499

SPECIAL SECTION: RADAR IMAGING SATELLITE-1Figure 16. Palaeochannel detected on RISAT SAR and Landsat ETMmerged FCC covering parts of Barmer, Jalor and Pali districts in thesouthern region of Luni River.of the River Sarasvati. The analysis of ERS-1/2 SAR datacovering parts of Thar Desert in western Rajasthan ledto identification of hitherto unknown buried channels,relict valleys and shallow, sand-covered limestoneareas 47,49,50,53,54 . Buried channel was identified near anarchaeological site, Talakadu (Mysore District, Karnataka)situated on the banks of the River Cauvery is southernIndia using RADARSAT-1, C band, VV polarization,fine-beam SAR data 55,56 of 22, 25 April and 19 May2008. RISAT SAR data in conjunction with optical imagesare being analysed to detect buried channels and associatedarchaeological sites followed by field studies forvalidation in the Thar Desert. Figures 15 and 16 demonstratethe potential of RISAT SAR data to detect palaeochannelsin parts of Rajasthan.These are known palaeochannels, and distinct high tomoderate backscatter is observed on RISAT SAR imagesdue to higher soil moisture compared to the adjoiningsand-dune areas. In addition, lower topography along thepalaeochannel is better enhanced due to oblique illuminationof SAR. Vegetation along the palaeochannel is distinctlyseen on Landsat ETM FCC. Merged RISAT SARand Landsat ETM FCC integrate the complementaryinformation content of both sensors and the signature ofpalaeochannel is distinctly enhanced. Work is in progressto detect hitherto unknown palaeochannels in other partsof the Thar Desert.1. Chakraborty, M., Panigrahy, S. and Sharma, S. A., Discriminationof rice crop grown under different cultural practices using temporalERS-1 SAR data. ISPRS Photogramm. Remote Sensing, 1997,52, 183–191.2. Panigrahy, S., Chakraborty, M., Sharma, S. A., Kundu, N., Ghose,S. C. and Pal, M., Early estimation of rice acre using temporalERS -1 Synthetic Aperture Radar data – a case study for Howrahand Hughly districts of West Bengal, India. Int. J. Remote Sensing,1997, 18, 1827–1833.3. Alpers, W. and Brummer, B., Atmospheric boundary layer rollsobserved by the synthetic aperture radar aboard the ERS-1 satellite.J. Geophys. Res., 1994, 99, 12613–12621.4. Kasischke, E. and Bourgeau-Chavez, L., Monitoring South Floridawetlands using ERS-1 SAR imagery. Photogramm. Eng. RemoteSensing, 1997, 63, 281–291.5005. Kasischke, E., Bourgeau-Chavez, L., Smith, K., Romanowicz, E.and Richardson, C., Monitoring hydropatterns in South Floridaecosystems using ERS SAR data. In Proceedings of the ThirdSymposium on Space at the Service of our Environment, Florence,Italy, 1997, pp. 71–76.6. Dwivedi, R. S., Rao, B. R. M. and Bhattacharya, S., Mapping wetlandof Sunderban delta and its environs using ERS-1 SAR data.Int. J. Remote Sensing, 1999, 20, 2235–2247.7. Baghdadi, N., Bernier, M., Gauthier, R. and Neeson, I., Evaluationof C-band SAR data for wetland mapping. Int. J. Remote Sensing,2001, 22, 71–88.8. Hess, L., Melack, J. and Simonett, D., Radar detection of floodingbeneath the forest canopy: a review. Int. J. Remote Sensing, 1990,11, 1313–1325.9. Hess, L., Melack, J., Filoso, S. and Wang, Y., Delineation of inundatedarea and vegetation along the Amazon floodplain with theSIR-C synthetic aperture radar. IEEE Trans. Geosci. Remote Sensing,1995, 33, 896–904.10. Slatton, K. C., Crawford, M. M., Gibeaut, J. C. and Gutierrez, R.,Wetland vegetation using polarimetric SAR. http://www.csr.utexas.edu/projects/rs/bol1.html (accessed on 19 April 2006).11. Pope, K., Rey-Benayas, J. and Paris, J., Radar remote sensing offorest and wetland ecosystems in the Central American tropics.Remote Sensing Environ., 1994, 48, 205–219.12. Kushwaha, S., Dwivedi, R. S. and Rao, B. R. M., Evaluation ofvarious digital image processing techniques for detection ofcoastal wetlands using ERS-1 SAR data. Int. J. Remote Sensing,2000, 21, 565–579.13. Ramsey, E. and Laine, S., Comparison of Landsat Thematic Mapperand high resolution photographs to identify change in complexcoastal wetlands. J. Coast. Res., 1997, 13, 281–292.14. McCauley, J. F. et al., Subsurface valley and geoarchaeology ofeastern Sahara revealed by Shuttle Radar. Science, 1982, 218,1004–1019.15. Schaber, G. G., McCauley, J. F., Breed, C. S. and Olhoeft, G. R.,Shuttle imaging radar: physical controls on signal penetration andsubsurface scattering in the eastern Sahara. IEEE Trans. Geosci.Remote Sensing, 1986, GE-24, 603–623.16. Guo, H., Schaber, G. G. and Breed, C. S., Shuttle imaging radarresponse from sand and subsurface rocks of Alashan Plateau innorth-central China. In International Symposium on Remote Sensingfor Resources Development and Environmental Management,7th, Proc. ISPRS Comm. VII, Balkema, Boston, 1986, pp. 137–143.17. Berlin, G. L., Tarabzouni, M. A., Al-Nasser, A., Sheikho, K. M.and Larson, R. W., SIR-B subsurface imaging of a sand buriedlandscape: Al Labbah Plateau, Saudi Arabia. IEEE Trans. Geosci.Remote Sensing, 1986, GE-24, 595–602.18. Avery, T. E. and Berlin, G. L., Fundamentals of Remote Sensingand Airphoto Interpretation, MacMillan, New York, 1992, 5thedn, p. 472.19. Blom, R. G., Crippen, R. E. and Elachi, C., Detection of subsurfacefeatures in SEASAT radar images of Means Valley, MojaveDesert, California. Geology, 1984, 12, 346–349.20. Farr, T. G., Elachi, C., Hartl, P. H. and Chowdhury, K., Microwavepenetration and attenuation in desert soil: a field experimentwith the Shuttle Imaging Radar. IEEE Trans. Geosci. RemoteSensing, 1986, GE-24, 590–594.21. Breed, C. S., Schaber, G. G. and McCauley, J. F., Subsurfacegeology of the Western Desert in Egypt and Sudan revealed byShuttle Imaging Radar (SIR-A). In First Spaceborne ImagingRadar Symposium, JPL Publ. 83-11, Jet Propulsion Laboratory,Pasadena, CA, 1983, pp. 10–12.22. Davis, P. A., Breed, C. S., McCauley, J. F. and Schaber, G. G.,Surficial geology of the Safsaf region, south-central Egypt,derived from remote-sensing and field data. Remote Sensing Environ.,1993, 46, 183–203.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013

SPECIAL SECTION: RADAR IMAGING SATELLITE-123. Elachi, C. Roth, L. E. and Schaber, G. G., Spaceborne radar subsurfaceimaging in hyperarid regions. IEEE Trans. Geosci. RemoteSensing, 1984, GE-22, 383–388.24. McCauley, J. F. et al., Palaeodrainages of the Eastern Sahara –the radar rivers revisited (SIR-A/B implications for a Mid-TertiaryTrans-African drainage system). IEEE Trans. Geosci. RemoteSensing, 1986, GE-24, 624–648.25. Brandt, P., Alpers, W. and Backhaus, J. O., Study of the generationand propagation of internal waves in the Strait of Gibraltarusing a numerical model and synthetic aperture radar images ofthe European ERS 1 satellite. J. Geophys. Res., 1996, 101, 14237–14252.26. Horstmann, J., Koch, W., Lehner, S. and Tonboe, R., Wind retrievalover the ocean using synthetic aperture radar with C-bandHH polarization. IEEE Trans. Geosci. Remote Sensing, 2000, 38,2122–2131.27. Rosich, B. and Meadows, P., Absolute calibration of ASAR level1 products generated with PF-ASAR. Technical Note, ENVI-CLVL-EOPG-TN-03-0010, European Space Agency, <strong>Issue</strong> 1 revision5, 7 October 2004, p. 26.28. Hersbach, H., Stoffelen, A. and Haan De, S., An improved C-bandscatterometer ocean geophysical model function: CMOD5. J.Geophys. Res., 2007, 112, 1–18; doi:10.1029/2006JC003743.29. Mouche, A. A., Hauser, D., Daloze, J. F. and Guerin, C., Dual polarizationmeasurements at C-band over the oceans: results from airborneradar observations and comparison with ENVISAT ASAR data. IEEETrans. Geosci. Remote Sensing, 2005, 43, 753–769.30. Monaldo, F., Thompson, D. R., Beal, R. C., Pichel, W. G. andClemente-Colon, P., Comparison of SAR-derived wind speed withmodel predictions and ocean buoy measurements. IEEE Trans.Geosci. Remote Sensing, 2001, 39, 2587–2600.31. Wright, J. W., Backscattering from capillary waves with applicationto sea clutter. IEEE Trans. Antennas Propag., 1966, AP-14,749–754.32. Goldfinger, A. D., Estimation of spectra from speckled images.IEEE Trans. Aerosp. Electron. Syst., 1982, AES-18, 675–681.33. Engen, G., Vachon, P. W., Johnsen, H. and Dobson. F. W.,Retrieval of ocean wave spectra and RAR MTFs from dualpolarizationSAR data. IEEE Trans. Geosci. Remote Sensing,2000, 38, 391–403.34. Eldhuset, K., An automatic ship and ship wake detection systemfor spaceborne SAR images in coastal regions. IEEE Trans. Geosci.Remote Sensing, 1996, 34, 1010–1019.35. Kathiresan, K. and Bingham, B. L., Biology of mangroves andmangrove ecosystems. Adv. Mar. Biol., 2001, 40, 81–251.36. Lucas, R. M., Mitchell, A. L., Rosenqvist, A., Proisy, C., Melius, A.and Ticehurst, C., The potential of L-band SAR for quantifying mangrovecharacteristics and change: case studies from the tropics.Aquat. Conserv.: Mar. Freshwater Ecosyst., 2007, 17, 245–264.37. Kasichke, E. S., Melack, J. M. and Dobson, M. C., The use ofimaging radars for ecological applications – a review. RemoteSensing Environ., 1997, 59, 141–156.38. Proisy, C., Mougin, E. and Fromard, F., Radar remote sensing ofmangroves: results and perspectives. In Proceedings of IGARSSConference, Sydney, Australia, 9–13 July 2001.39. Kumar, T. and Patnaik, C., Characterization of mangrove vegetationusing temporal C-band synthetic aperture radar data: a casestudy of Sundarbans. In Proceedings of International Conferenceon Geospatial Technologies and Applications ‘Geomatrix 12’, IITMumbai, CD-ROM, 2012.40. National Wetland Atlas. SAC/EPSA/ABHG/NWIA/ATLAS/34/2011, Space Applications Centre, ISRO, Ahmedabad, 2011.41. Forest Survey of India, Forest cover mapping. Ministry of Environmentand Forests, FSI, Dehradun, 2009; http://www.fsi.org.in/scheme_of_classification.html42. Ghose, B., Kar, A. and Hussain, Z., The lost courses of the SarasvatiRiver in the Great Indian Desert: new evidence from theLandsat imagery. Geogr. J., London, 1979, 145, 446–451.43. Ghose, B., Kar, A. and Hussain, Z., Comparative role of Aravalliand the Himalayan river systems in the fluvial sedimentation ofthe Rajasthan desert. Man Environ., 1980, 4, 8–12.44. Pal, Y., Sahai Baldev, Sood, R. K. and Agarwal, D. P., Remotesensing of the ‘lost’ Saraswati river. Proc. Indian Acad. Sci.(Earth Planet. Sci.), 1980, 89, 317–331.45. Sood, R. K. and Sahai, B., Hydrographic changes in northwesternIndia. Man Environ., 1983, 7, 166–169.46. Ramasamy, S. M., Bakliwal, P. C. and Verma, R. P., Remote sensingand river migration in western India. Int. J. Remote Sensing,1991, 12, 2597–2609.47. Rajawat, A. S. et al., Potentials of radar (ERS-1/2 SAR) and highresolution IRS-1C data in reconstructing palaeodrainage networkof western Rajasthan. Mem. Geol. Soc. India, 1999, 42, 245–258.48. Rajawat, A. S., Sastry, C. V. S. and Narain, A., Application of pyramidalprocessing on high resolution IRS-1C data for tracing themigration of the Saraswati river in parts of the Thar desert. Mem.Geol. Soc. India, 1999, 42, 259–272.49. Rajawat, A. S., Verma, P. K. and Nayak, S., Reconstruction of palaeodrainagenetwork in northwest India: retrospect and prospectsof remote sensing based studies. Proc. Indian Natl. Sci. Acad. PartA, 2003, 69, 217–230.50. Rajawat, A. S., Reconstruction, evolution and significance ofpalaeodrainage network in western Rajasthan: an investigationbased on multisensor satellite data and Geographic InformationSystem (GIS). Ph D thesis, Vikram University, Ujjain, 2006,p. 331.51. Sahai, B., Unraveling of the ‘lost’ Vedic Sarasvati. Mem. Geol.Soc. India, 1999, 42, 121–141.52. Rajani, M. B. and Rajawat, A. S., Lost courses of River Sarasvatiand spatial distribution of Harappan settlements – observationsbased on multisensor satellite data. In Remote Sensing for Archeologyand Cultural Heritage Management (eds Lasaponara, R. andMasini, N.), Rome, Italy, 2008, pp. 383–386.53. Mehta, N. S., Rajawat, A. S., Bahuguna, I. M., Mehta, D. S. andShrimal, A. K., Geological potential of ERS-1 SAR data: observationsin parts of Aravalli and Thar Desert (western India). In Proceedingsof the Second ERS-1 Symposium – Space at the Serviceof Our Environment, Hamburg, Germany, ESA SP-361, 11–14October 1993, pp. 931–936.54. Mehta, N. S., Rajawat, A. S., Bahuguna, I. M., Mehta, D. S.,Bhatnagar, P. and Shrimal, A. K., Remote sensing for mineralexploration: geological potentials of ERS-1 SAR data coveringparts of Aravalli Hills and Thar Desert (Rajasthan, India). ProjectReport, SAC/RSA/RSAG/ERS-1/GEO/PR/01/94, Space ApplicationsCentre, ISRO, Ahmedabad, 1994, p. 67.55. Rajani, M. B., Rajawat, A. S., Krishna Murthy, M. S., Kamini, J.and Rao, S., Demonstration of the synergy between multi-sensorsatellite data, GIS and ground truth to explore the archaeologicalsite in Talakadu region in South India. J. Geomatics, 2012, 6, 37–42.56. Rajani, M. B., Bhattacharya, S. and Rajawat, A. S., Synergisticapplication of optical and radar data for archaeological explorationin the Talakadu region, Karnataka. J. Indian Soc. Remote Sensing,doi: 10.1007/s12524-011-0102-6, published online: 3 June 2011.ACKNOWLEDGEMENTS. This work was carried out under theRISAT-1 Utilization Programme funded by ISRO. We thank A. S.Kiran Kumar, Director, Space Applications Centre, ISRO, Ahmedabadand Dr J. S. Parihar, Deputy Director, Earth, Ocean, Atmosphere,Planetary Sciences and Applications Area for their encouragement.CURRENT SCIENCE, VOL. 104, NO. 4, 25 FEBRUARY 2013 501