The first imaging radars were Side-Looking Airborne Radars (SLARs) developed in the military for reconnaissance purposes. At the wavelengths then used (2 cm), these radars could image the Earth's surface through clouds and atmospheric water vapor under day/night conditions, a considerable advantage for reconnaissance missions. SLAR is known as a "real aperture" radar because its along-track resolution is determined by the size of the physical antenna footprint on the ground, which is given by the ratio of wavelength times the slant range divided by the along-track antenna length--fine resolution in the cross-track direction is obtained through the use of pulse compression techniques. The synthetic aperture radar (SAR) was invented in the 1950s to permit radars to achieve fine spatial resolution in the cross-track direction. As such, SAR's can achieve equal resolution in both directions.
The principal disadvantage of using SLAR is that its along-track resolution is limited by the antenna length. The development of synthetic aperture radar overcame this problem. Like SLAR, SAR used pulse compression techniques to provide fine-range resolution. However, it was shown that if a pulsed coherent radar could be used, then the Doppler-shifted radar returns could be recorded and played back through a coherent SAR image processor to synthesize an along-track antenna length much longer than the antenna's physical length. A further advantage of SAR was that it could be used at longer wavelengths than SLAR.
The first nonreconnaissance uses of these real aperture radars were for cartography and geologic mapping. Radar returns are sensitive to surface structure, surface roughness, surface slope, and the presence of water. Combined with an all-weather capability, this provided geologists with a means to map heavily cloud-covered and previously uncharted regions. Studies were conducted to determine how SLAR images could be used in the emerging field of radargrammetry; i.e., the use of radar images for cartographic and topographic mapping, the functional equivalent to the well established field of photogrammetry. It was also determined that radar images could be used to map different surficial materials, including sand, gravel, and glacial deposits, or to delineate geologic contacts between fans and playas or bedrock and alluvial fans.
The first space-based imaging radar to be used for imaging of the Earth was the L-band SAR on Seasat (Figure 3-1), one sensor in an instrument suite launched into an 800-km altitude near-polar orbit in June 1978. This horizontally polarized sensor operated at a fixed wavelength (23 cm) and at a fixed-look angle (20 degrees from nadir). The Seasat swath width was 100-km and the resolution was approximately 25 meters. While a SAR was included in the Seasat payload primarily for the purpose of ocean wave imaging, during its 3-month lifetime, radar images also were acquired over large areas of the Northern Hemisphere.
Imagery obtained from the Seasat SAR clearly demonstrated its sensitivity to surface roughness, slope, and land-water boundaries. Seasat images have been used to determine the directional spectra of ocean waves, surface manifestations of internal waves, polar ice-cover motion, geological structural features, soil moisture boundaries, vegetation characteristics, urban land-use patterns, and other geoscientific features of interest.
Seasat also established a commercial demonstration program involving private-sector operational users representing a variety of marine applications, including off-shore oil and gas operations, optimum ship routing, deep-sea mining, commercial fishing, and private weather forecasting. The purpose of this program was to permit operational users an opportunity to test the utility of Seasat-derived measurements, including SAR, in commercial ocean applications. Through a series of retrospective case studies, commercial users were able to show the economic impact of the use of Seasat data in daily operations.
Despite its overall technological and scientific success, Seasat's relatively short lifetime precluded the acquisition of a seasonal data set. Moreover, the Seasat SAR was a single-parameter instrument using a fixed wavelength, polarization, and incidence angle. While the near-nadir incidence angle was ideal for acquiring strong ocean returns, it produced severe geometric layover distortions on terrain images of high-relief regions.
The next space-based SAR to follow Seasat was the Shuttle Imaging Radar- A (SIR-A), launched in November 1981 on board the Space Shuttle Columbia on its second mission. Using the Seasat SAR technology (spare hardware), but with a higher incidence angle of 50 degrees, the SIR-A mission was focused on geological research. SIR-A provided much improved image data for geological analysis that were relatively free of the layover distortions in areas of high relief. SIR-A also led to the discovery of buried and previously uncharted dry river beds beneath the Sahara Desert, thus demonstrating the ability of L-band radar to penetrate up to several meters in hyperarid sand sheets.
SIR-B, the next NASA SAR mission, flown in October 1984 on the orbiter Challenger, also used the Seasat/SIR-A technology, but with an articulating antenna which permitted a variable incidence angle over a 15- to 60-degree range. This first multi-incidence angle data set demonstrated the potential for mapping surface features (particularly forests) using multiple-incidence angle backscatter signatures, and for topographic mapping. SIR-B data were also used to demonstrate the sensitivity of L-band radar images to parameters such as soil moisture, geological, structural, and lithologic features, and oceanic directional wave spectra.
The success of the SIR-A and SIR-B missions, supported by aircraft-based SAR research, led to a second generation SAR design, reflected in the SIR-C/X-SAR instrument (Figure 3-2) on the Space Radar Laboratory (SRL) missions. These two missions (SRL-1, launched April 1994, and SRL-2 launched October 1994) incorporated a multifrequency, multipolarization, variable incidence angle SAR and has provided scientific and applications information hitherto unavailable. The multiparameter capability of the SIR-C/X-SAR radar, coupled with the introduction of routine sensor calibration, opened a new regime in SAR-based scientific investigations and applications. The analysis of data from the SRL missions is only beginning, but initial results indicate dramatic new capabilities only possible with the a multiparameter instrument. Polarimetric data has allowed improvements in soil moisture measurements in bare soil areas and canopy water content estimates in vegetation covered areas. Accurate maps of snow cover and snow water equivalence are now possible to aid in water supply forecasting and hydroelectric power management. Oil spills are detectable and the multiparameter capability permits oil type and oil-natural surfactant differentiation to be determined. Ocean surface waves were measured and, for the first time, wave spectra were processed on-board the satellite for direct downlink to operational centers. Cross-polarization data were proven to be a powerful tool for extracting lithographic information in the interpretation of geologic features. Mud flows on Mt. Pinatubo were observed which posed severe hazards to local populations. Volcanic ash deposits and lava flows were observed, and, using interferometric techniques, measurements of surface displacement in volcanic areas were obtained. Such measurements offer the potential for predicting volcanic eruptions and producing damage maps following eruptions and earthquakes. Using calibrated cross-polarization L-band data, the aerodynamic roughness of the ground was measured, thus permitting the assessment of the ability of the wind to initiate dust and sand storms. Multiparameter data allowed a determination of wetland inundation (flooding status) and vegetative cover, biomass estimates, crop monitoring, vegetative mapping, and the monitoring of flooded forests and coastal/low-stature wetlands. While many of these capabilities were the result of extensive airborne SAR research, their extension to space-based platforms was confirmed as a result of the SRL missions. More exciting results will surely emerge as data analysis continues.
Building upon the Seasat experience, the European Space Agency (ESA) embarked upon the development of the ERS-1 satellite (Figure 3-3). Launched in May 1991, this satellite contains a suite of earth/ocean observing microwave instruments, including a C-band SAR. Like Seasat, the SAR has a single-frequency, single-polarization (VV) and fixed incidence angle (23 degrees). Unlike Seasat, the SAR operates as one mode of an Active Microwave Instrument (AMI) which, in another mode, functions as a wind field scatterometer. In its nearly 4 years of operation, the ERS-1 SAR has proven to be an extremely reliable and stable instrument. The calibrated SAR data have proven to be highly useful in polar ice applications where ice-type determinations are possible using fixed look-up tables. In spite of being a single-parameter instrument, the ERS-1 SAR has provided a rich and extensive set of Arctic sea ice, ocean, and land data. The sea ice data are, for the first time, being used to support ice operations in the national centers of the U.S., Canada, and Sweden. The nearly 4-year life of the SAR has permitted the collection of an extended time series of data over several seasons, allowing long-term and seasonal variations of land features, vegetation cover, and sea ice to be measured.
ERS-1 SAR has been used to monitor flood water levels during the great Midwestern floods of 1993 and detect standing water beneath vegetation canopies. ERS-1 SAR has demonstrated a crop monitoring capability using multitemporal data, and has been effective in studying boreal forests on issues related to the terrestrial carbon cycle. With some limitations, ERS-1 SAR has been effective in mapping wet snow. ERS-1 SAR has established most of the scientific utility for ice sheet and glacier work. It has been a powerful tool in measuring the Greenland ice sheet, where the data clearly define the zonation and boundaries of the surface, based on backscatter signatures. Variations in these patterns can provide an indication of changing input conditions on a near-continental scale. In coastal ocean regions, ERS-1 SAR has measured fronts, eddies and internal waves, and has shown variations in the shallow water seabed topography through surface signature modulations. Images of wave refraction and shoaling have also provided indications of the seabed topography. Surface ships and shipwakes have been regularly detected in the SAR imagery, and operational demonstration projects on vessel traffic monitoring have been successfully conducted. Oil spills have been detected and monitored, and several vessels have been cited by pollution authorities for violations based on ERS-1 SAR observations. ERS-1 SAR observations of sea and lake ice have shown that ice motion, ice age, ice/water boundaries, polynyas, leads, and ridges can be accurately measured. While in the 3-day repeat orbit phases of the mission, ERS-1 SAR has been used in repeat-pass interferometric mapping experiments with good results. However, the 23-degree incidence angle creates severe layover distortion in areas of high relief.
Earlier this year, ESA successfully launched the ERS-2 spacecraft, identical to the ERS-1. The two satellites are being flown in a tandem mode (flying one after the other along the same track) to allow interferometric processing of images from both spacecraft to create topographic maps. While this is being done on a very limited demonstration basis, the results to date show promise for this technique which can be expected to be useful in follow-on multi-spacecraft programs.
In 1984, Japan initiated the development of the JERS-1 satellite that was successfully launched in February 1992. The principal sensor in the instrument payload is an L-band SAR with a single polarization (HH) and fixed-incidence angle (35 degrees). The longer wavelength was adopted in order to provide greater penetration of vegetative cover and sand layers, while the relatively large incidence angle was incorporated to reduce the layover distortions in mountainous and high-relief regions. Unlike the Seasat and ERS-1 and 2 satellites, the JERS-1 spacecraft includes two tape recorders that may be used to store SAR data for subsequent playback to selected ground stations. As a result, global SAR data can be acquired independent of a global network of ground stations. Unfortunately, a reduction in transmitter power has limited the use of JERS-1 data.
Brief mention should be made of the short duration mission of the Almaz radar satellite launched by the Soviet Union in 1987. Almaz was a single-frequency SAR sensor mounted on an extremely large spacecraft derived from SALYUT space station components. Although the Soviet government later shared imagery from Almaz, revealing optically-processed data similar in capability to Seasat, the program was originally shrouded in secrecy (the spacecraft identified only as Cosmos 1870 by the U.S. Space Command). In the new political order of Russia, a follow-on Almaz has been announced as almost ready to launch, and support is being collected by Russia for data purchases which can be used to offset the cost of the program.
The Canadian Space Agency (CSA) developed RADARSAT (Figure 3-4), which was launched by NASA from Vandenberg AFB, California on November 4, 1995. This spacecraft carries a single-frequency (C-band), single-polarization (VV) SAR which provides a variety of beam selections that can image swaths from 35 km to 500 km with resolutions from 10 meters to 100 meters, respectively. Incidence angles range from less than 20 degrees to more than 50 degrees. This satellite is the first in a continuing series of RADARSAT spacecraft.
RADARSAT is operated as a quasi-commercial program, with the Canadian Government paying for the development and operation of the spacecraft and a private company, RADARSAT International (RSI), will market data with its revenue helping to defray the cost of operations and production.
Over the ensuing 16-plus years since the launch of Seasat, space-based SAR technology has evolved to a powerful, but still developing level. More scientific research is needed to understand the radar phenomenology involved with some microwave-surface interactions, from which refined and robust algorithms will emerge to permit more accurate measurements of geophysical parameters. The advent of multiparameter SAR and interferometric techniques permit vast new measurement capabilities that can make space-based SAR an ever more powerful remote sensing tool, supporting both scientific research and operations. The experiences to date with ERS-1, ERS-2, JERS-1 and the SIR-C/X-SAR missions, have provided the learning experiences for operational users to develop the skills needed to use SAR data from the satellites to be available over the next decade in support of operational applications. Unlike scientific investigations, operational applications need not have perfectly refined algorithms or data extraction techniques in order to achieve operational utility. In many areas, such as ice operations, coastal zone and wetlands monitoring, flood water level monitoring, fisheries management and enforcement, and oil spill detection and monitoring, the SAR technology, based on single parameter SAR, is sufficiently mature to make operational investments now in data acquisition and utilization useful and cost-effective. Both single and multiparameter SAR systems can yield geophysical measurements under all-weather, day/night conditions at high-spatial resolution over wide surface areas--a unique aspect of this sensor technology. Continuity of data is a major requirement of operational users, which dictates that a series of space-based SAR systems must be approved and launched over a decade or more time period. The current number of approved future SAR missions shown in (Figure 3-5) begins to ensure this continuity of data and is sufficient to justify operational investments.