Spaceborne and Airborne SAR for Target Detection and Flood Monitoring

Huadong Guo

Abstract The Spaceborne Imaging Radar-c/X-band Synthetic Aperture Radar (SIR-C/X-SAR) mission was a cooperative endeavor of the United States, Germany, and Italy. The SIR-C/X-SAR scientific research program was a large international cooperative pro- gram of radar for Earth observation in which 13 countries participated, including China. SIR-C/X-SAR, with the ability to acquire polarimetric SAR and integerometric SAR data, was the first spaceborne radar to operate simultaneously at several frequencies andpolarizations, representing the most advanced civilian SAR system for Earth observation. This paper will present some results of the SIR-C/X-SAR program made in China. The emphasis is placed on aerial and ground synchronous experiments with SIR-C/X-SAR overpasses, SAR penetration studies for dry sands, and SIR-C/X-SAII data applications in relevant fields and different areas, e.g., discovering the vol- canoes of the Kunlun Mountains, detecting geological features underneath vegetation canopies, and revealing the Great Wall segments of the Ming and Sui dynasties. The paper also in- troduces the use of the Chinese airborne L-band SAR system, developed by the Chinese High Technology Program, for flood monitoring in 1998.

Introduction The Spaceborne Imaging Radar-C/X-band Synthetic Aperture Radar (SIR-C/X-SAR) was placed aboard Space Shuttle flights in April and October 1994 and successfully conducted two 10-day missions for Earth observation with 107 terabits of data re- turned (Stofen et al., 1995). Up to now, all other spaceborne imaging radar systems were characterized by single band and single polarization, including Seasat SAR, SIR-A, SIR-B, Almaz-1 SAR, FRS-ISAR, ERS-IIZSAR, and Radarsat. The SIR-CIX-SAR has the following distinct characteristics: (1) the first multifre- quency imaging radar system to be simultaneously operated in Earth orbit, (2) the first multipolarization imaging radar system, and (3) a system with interferometric measurement ability. This system largely utilizes wavelength, polarization, magni- tude, and phase information of electromagnetic waves, enab- ling further exploitation of electromagnetic resources in the microwave bands. The multifrequency, multipolarization capability creates a new and powerful tool for detecting surface targets more accurately and efficiently.

As an integral part of the SIR-C/X-SAR research program, the Chinese team has conducted multidisciplinary research work for Chinese test sites. A simultaneous experiment with airborne SAR imaging and ground measurements was carried out during the Shuttle Endeavor's overpass. With the advanced SAR data, detection and analysis for terrain surface and subsurface as well as man-made features were conducted. The unprece- dented flood disaster of 1998, which occurred along the mid

Laboratory of Remote Sensing Information Sciences, Institute of Remote Sensing Applications, Chinese Academy of Sci- ences, P.O. Box 9718, Beijing 100101, P.R. China.

PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING

and lower reaches of the Yangtze River, was successfully moni- tored with an L-band airborne SAK system developed by the Chinese High-Technology Program.

SAR System Characteristics and Chinese Test Sites SIRC/X-SAR System Characteristics SIR-C/X-SAR was operated at L-, C-, and X-bands, which can acquire data for the same target at the same time. There are four polarization combinations at the L- and C-bands, i.e., HH, Hv, VH, and w polarizations. h addition, SIR-C can provide phase- difference images of four polarized echoes, from which a full scattering matrix can be derived and full polarimetric SAR information can be obtained. X-band SAR has vv polarization only (Table 1). The viewing angle is adjustable aboard and ranges from 15" to 60". The acquired SIR-C/X-SAR data were mainly recorded and saved onboard; a small amount of data were sent to the ground via the Tracking and Data Relay Satel- lite System for real-time data processing.

SIR-C/X-SAR developed and applied four advanced tech- niques. These are interferometric SAR, ScanSAR, calibration, and SAR data real-time processing techniques. Based on the repeated-pass interferometric-imaging principle, interfero- metric SAR data collection was performed during the last three days of the second SIR-C flight, obtaining a great deal of interfer- ometric SAR data including China's test sites. To meet the needs of a large-scale study for natural resources, environment, and technical experiments, scans^^ images were acquired during two flights of the SIR-CIX-SAR missions with a maximum swath width of 200 km. The radiometric calibration was performed for the SIR-C/X-SAR data. Thus, this allowed the use of the SIR-CI X-SAR data to directly derive radar backscattering coefficients and to conduct a quantitative remote sensing study. Another key technique developed was a real-time SAR processor on the ground, which is a high-speed processor with data rate of 45 Mbps and is capable of continuously outputting SAR image data. Data strips of 7 krn can be processed per second.

SIRC/X-SAR Test Sites in China The design of the SIR-c1x-S~~ test sites in China began in 1990. After studying the SIR-C/X-SAR parameters, six test sites were selected, including the Xinjiang, Inner Mongolia, Guangdong, Hubei, Taiwan, and Huabei (Beijing, Shangdong, and Hebei) areas. The major application fields were geology, forestry, oceanography, penetration studies, etc. According to the dif- ference of terrain features and SIR-C/X-SAR data application

Photogrammetric Engineering & Remote Sensing Vol. 66, No. 5, May 2000, pp. 611-617.

0099-1112/00/6500-000$3.00/0 O 2000 American Society for Photogrammetry

and Remote Sensing

Parameter L-Band C-Band X-Band

Orbital Altitude (km) 225 Wavelength (cm) 23.5 5.8 3.1 Polarization HH, HV, VH, W HH, HV, VH, W W Azimuth Resolution 30 X 30 Range Resolution 13 and 26 m 13 and 26 m 10 and 20 m Swath Width (km) 15 - 90 15 - 60 Look Angle (") 20 - 55 Bandwidth (MHz) 10,20 'llansmit Pulse Length 33.17, 8.5 psec Pulse Repetition Rate 1240 and 1736 pulses per second

40 psec

Data Rate [Mbps) 90 90 45 Data Acquisition Time (h) > 100 NB: except as specified in the table, other parameters in the L and X bands are the same as those in the C band.

objectives, viewing angles were first determined for the central coordinates of each test site, and imaging modes were then selected based on the chosen viewing angles. SIR-clx-SAR has 23 imaging modes (Stofan et al., 1995). Apart from X-band w polarization for X-SAR imaging, the other modes are different combinations of HH, HV, VH, and W polarizations at the L and C bands. Modes 22 and 23 were specifically designed for inter- ferometric SAR. In addition to different radar information being acquired at different modes, the radar information was also affected by the differences in imaging swaths.

During the flights of the Space Shuttle Endeavor in April and October 1994, SIR-CIX-SAR imaged all the test sites in China in accordance with our requirements. The imaging swath usu- ally ranged from 15 to 90 krn in width. The length of each image strip varied for each data take depending upon the imaging time. Some strips were a few hundred kilometers in length but some exceeded 2000 km. A total of 90 minutes of sm-CIX-SAR data were acquired for the Chinese test sites. In addition, the astronauts on board the Space Shuttle Endeavor took hand- held camera photos of the test sites when weather conditions allowed. These photos provided very useful data for per- forming a comparative study with the radar images.

Chinese Airborne LSAR System In 1997, China developed an airborne L-band SAR operational system, with the support of the Chinese High Technology Pro- gram (863-308), with the objective of conducting preliminary research for a Chinese spaceborne SAR system to be launched following the turn of the century.

The system has been placed aboard the Citation SIII air- craft owned by the Chinese Academy of Sciences. The flight altitude varies from 6,000 to 10,000 m, and the flight velocity is 550 kmlh. The L-SAR system is operated at 1250 MHz with HH and w polarizations and left and right viewing directions. Table 2 gives the operational modes of the L-SAR system.

The LSAR system has a wide application potential for detecting targets beneath canopies forest and dry deserts, min- eral exploration, offshore exploration, studying oceanic waves, mapping underwater topography, etc. The system has demon-

strated its power in monitoring the flood disaster that occurred in the mid and lower reaches of the Yangtze River.

SpacebomeAirbome Ground-Based Radar Remote Sensing Experiment One quantitative remote sensing method is to make ground measurements during the flight of airborne or spaceborne remote sensing experiments. We conducted this experiment at a Beijing test site during the overpass of SIR-CIX-SAR. From 18 to 20 April 1994 when SIR-CIX-SAR imaged the test site, an X- band airborne SAR developed by the Chinese Academy of Sci- ences (CASSAR) was imaging the same area at the same time. Meanwhile, a truck-mounted scatterometer was measuring the ground surface's backscattering coefficients. The surface roughness was measured and samples were collected.

The WSSAR operated in the X-band and HH, W polariza- tions. Its resolution is 10 m by 10 m. The flight altitude was 7200 m. The azimuth and range directions were designed to coincide with the first mission of SIR-CIX-SAR (SRL-I). The inci- dence angles were 73.93" to 83.12'. CA,~XAR included both an optical processor and a digital real-time processor for data pro- cessing. However, during the operation the optical mode was used, and the data were not calibrated. The truck-mounted scat- terometer operated in the X-band with four polarizations. The antenna was fixed on a boom about 12 m high. It could illumi- nate at any azimuth direction and elevation angle from 0' to 84" with a 6" interval. A Lunenburg sphere was used for external calibration prior to measurement. The measured distributed targets included bare soil, winter wheat, rice staple, water bod- ies, etc. During the observations, vegetation and soil parameters such as biomass, soil moisture, and surface roughness were measured in real-time (Wang et al., 1996).

During the first SIR-CIX-SAR mission, the truck-mounted scatterometer data acquired in real-time were used for calibra- tion of the Chinese airborne C A ~ ~ A R data and SRL-1 survey image in the Beijing test site. In the image linear range, the regression precision is less than 2 dB (except for the L-band data). However, due to the limited dynamic linear range of dis-

TABLE 2. OPERATIONAL MODES OF THE L-SAR SYSTEM

Modes A B1 B2 C1 C2

Altitude (km) 6 6 6 6 6 Slant range (km) 6.4 - 10 9 - 18 9 - 27 15 - 24 15 - 33 Resolution (m) 3 x 3 3 x 3 3 x 3 (Image produced on the ground) Resolution (m) (Image produced aboard in real-time) 3 x 10 6 x 10 3 X 10 6 X 10

812 May 2000 PHOTOORAMMETRJC ENGINEERING & REMOTE SENSING

tributed targets, it is better to use active calibrators to extend the dynamic range of calibration. Analysis of the relationship between soil moisture, water content of winter wheat, and image gray level indicates that there is a good linear correlation between soil moisture and image gray level, but relatively poor correlation between the water content of winter wheat and image gray level. This indicates that there are different radar responses to the two terrain features.

SIR-C/X-SAR Penetration Study Using SIR-A and SIR-B imaging radar data, the author studied and revealed the phenomena of L-band SAR penetrating dry sands in the Gobi desert of Inner Mongolia (Guo et al., 1986; Guo, 1986). Thus, this area was selected as a test site for the SAR penetration study. The area is at the eastern edge of the Badan- jilin Desert. The climate of the area is extremely dry. The sur- face is very flat. Just before the overpass of SIR-CIX-SAR, we placed 1 2 corner reflectors in two sites of the area. Site 1 was selected to the west of the triangular bedrock area reported by Guo (1986), where about a 1-meter thickness of sand filled the depression area of bedrock. But in some areas there are still some outcropped bedrock. The SAR images, especially the L- band images, clearly revealed the surface morphology of the triangular bedrock. There were four corner reflectors buried in natural sands, the bottom depths of which were about 1 m. The purposes of selecting site 1 were two-fold: (1) to observe the pen- etration capability of microwaves in natural conditions, and (2) to understand the process of microwave attenuation in natural sands. Site 2 was selected in an area of the Gobi Desert where about 20 cm layer of dry sand covered the unconsolidated Cenozoic alluvium. Here, six corner reflectors were buried in the spaces to depths varying from 0.93 to 2.73 m with dry sands taken from other places. The reason for selecting site 2 was to study the attenuation process of microwaves in dry sands. TWO corner reflectors were placed on the surface of the Gobi Desert.

Because the penetration depth of SAR is closely related to the moisture of the materials on the ground, the measurement for the moisture and complex dielectric constant of sands was performed for the samples collected from the two sites. By mea- suring the experimental penetrating depths and determining the penetration conditions, the following conclusions have been drawn:

(1) The theoretical maximum penetrating depth of SIR-clx-SAR is 2.82 m for dry sands in Inner Mongolia, but it is possible to detect objects deeper than 2.82 m (Shao et al., 1995); and

(2) The reasons for L-band SAR'S ability to detect bedrock under- neath the sand layer are (a) the sand layer is about 1 m thick; thus, the contact zone of sand and bedrock within 0.65 m is detectable; and (b) there is some energy < l /e backscattered to radar antennas due to its intensive interaction with the contact of rock and sand; thus, the returned energy can be detected within the radar's sensitivity.

Target Detection and Recognition Detecting the Volcanoes of the Western Kunlun Mountains The western Kunlun Mountains lie at the conjunction of the Tarim plate and the Tibet-Indian plate. For many years they have received a lot of attention from geologists in the world. The volcanic activity in the Kunlun Mountains since the Ceno- zoic period is most significant for studying the movement of plate tectonics and the evolving history of the continents. While analyzing SIR-C/X-SAR images acquired on 17 April 1994, a group of volcanoes northeast of Aksayqin Lake in the western Kunlun Mountains were identified (Guo et al., 1997). The vol- canoes are situated at an elevation of over 5300 m above sea level. There are well-exposed volcanic cones and lava flows in the area. Rvo types of lava flow - pahoehoe and blocky lavas

- and nine cinder cones were mapped, which had not pre- viously been reported. Among them, four cinder cones were easily identified because they occur in an open valley and have high topographic relief. The other five cones have positive topography and circular form, but they occur on top of the lava flows and therefore are not easily detected. The squint SAR geometry accentuates the cinder cones and enables them to be identified. Most lava flow areas are occupied by pahoehoe lava, which forms a relatively smooth surface. The surface of a pahoehoe flow is rather flat and mainly composed of smaller fragments, about 10 to 20 cm in size. The radar signature of this type of flow appears relatively dark due to specular scattering from the surface, whereas radar returns from the toe of the flow appears very bright, particularly when the toe faces toward the radar illumination. The surface of blocky lava flows is notice- ably rough with irregular fragments of clinker ranging from 30 cm to more than one meter in size. The lavas have been eroded and weathered. Vesicular and amygdaloidal rocks are the most obvious features on the flows. The rough surfaces give rise to strong diffuse scattering of the electromagnetic wave.

Comparison of L-, C-, and X-band images with different polarizations indicates that the L-band HV polarization is the best for identifying cinder cones and for discriminating the two different types of lava flows (Figure 1). In contrast to the contri- bution from dielectric constants of surface materials to the radar echoes, surface roughness and topography play the most important role in the observed radar returns, while depolariza- tion of the lava surface contributes mostly to the differences for the L-band HV image with co-polarized images.

Geochemical analyses of volcanic rock samples collected from the field indicate that the potassium content is very high. K20 is commonly higher than 4 percent, and K,O is greater than Na20. The samples are enriched in the trace elements Rb, Sr, Ba, Nb, and Ta and in the light rare earth elements. They basically do not show Eu anomalies. The rocks can be classified into the shoshonitic series, which is associated with intracontinental subduction. The K-Ar isotopic ages range from 7.45 to 3.97 Ma, indicating that there were at least two volcanic eruptive phases. This provides further evidence of the presence of neotecton- ism in the Qinghai-Tibetan plateau since the Cenozoic era.

Detection of Uthology Beneath the Vegetation Canopy In a subtropical region, it is very difficult to carry out geological surveying and mapping due to thick depths of soil and dense vegetation cover, poor exposure of bedrock, and inaccessibility. Synthetic aperture radar images, especially the multi-parame- ter SAR, may detect geological features beneath vegetation cover (Guo et a]., 1996). Some of the information extracted from SAR images is impossible to obtain from visible remote sensing data.

On 18 April 1994, the SIR-C/X-SAR flew over the Zhao Qing region in the Guangdong Province of southern China and acquired one data take at mode 11 (L-HH, L-Hv, C-HH, C-HV, and x- w). The research work in Zhao Qing with SIR-CIX-SAR data demonstrates the advantages of multi-parameter SAR technol- ogy in detecting the geological structures and lithology beneath a vegetation canopy. This study was aimed at providing funda- mental information for geological surveying and mapping at scales of 1:100,000 and 1:200,000.

It is difficult for color infrared imagery to distinguish sedi- mentary rocks and determine their strike and dip orientation, which are essential elements for geological surveying and map- ping. The multifrequency, multipolarization color composite SAR image for the Zhao Qing test site reveals the beds of sedi- mentary rocks beneath the dense vegetation cover. On a false- color SIR-c composite image using L-HH (R), L-HV (G), and c-HV (B), there are large bands of cyan, and magenta to grayish brown stripes. The stripes are parallel to each other and have clear boundaries and a sawtooth shape. These are only visible

PHOTOQRAMMmRIC ENGINEERING & REMOTE SENSING May 2000 613

Illumination 0 4 8 km

Figure 1. L-, C-, and X-band and Hv and w SAR imagery of the volcanoes in the western Kunlun Mountains.

in the SAR image; no traces of the stripes are found in the Land- sat TM image or in the color infrared airphoto at a scale of 1:35,000 due to vegetation cover. The distribution pattern of these colored stripes is the same as the occurrence of sedimen- tary clastic rocks of the Devonian Guitou Formation (D2-3) on the north slope near Zhao Qing city. These color stripes can be identified as beds of Devonian sedimentary rocks because of the V-shaped outcrop of two adjacent formations, shown in the SAR image in a sawtooth pattern. A V-shaped outcrop of a geo- logical boundary represents a line along which a contact between two adjacent formations or beds intersects the Earth's surface. Field-work confirmed that these color stripes repre- sent beds of sedimentary rocks of the D2-3 formation. There are a few outcrops of these Quartzose sandstones and siltstones covered with dense vegetation growing in a thin layer of soils.

The formation of the color stripes comes mainly from the L-band image. Sun and Ranson (1995) pointed out that the backscatter received by the SAR system from a vegetated area is composed of five components. One is the trunk-ground com- ponent. L-band microwave penetrates the crown layer of the trees and reaches the ground, so there are distinct variations of intensity from the top of the slope to the bottom in the L-HH and L-HV images. It is not clear in the C-band image and there are no variations in the X-w image. L-band SAR images show the geo- logical features beneath a vegetation canopy better than does the X-band image. The vegetation canopy is comparable to an enveloping surface over the land surface conforming to the topography. The radar return from the X-band image comes from the top of the tree canopy whereas the longer-wavelength microwave penetrates through the vegetation layer and collects the sub-canopy information.

Archaeology The Great Wall segments in a desert region of north central China, within the boundaries of Ningxia and Shanxi prov- inces, were revealed by the SIR-CIX-SAR images (JPL, 1996). Three Walls were identified, two of them built in the Ming dynasty and one in the Sui dynasty. The Great Wall of the Sui

dynasty was built in the year 585 with rammed earth. It is now discontinuously elongated, about 4 m in width and 1 to 3 m in height. The Great Wall of the Ming dynasty was built in 1531 with rammed earth. The remnant wall is about 6 to 8 m in height and 6 to 8 m in basal width.

Figure 2 shows a small portion of the Great Wall in four black-and-white images representing SIR-C's four radar chan- nels. The radar beam was perpendicular to the Great Wall, resulting in a corner reflector effect which enabled the SAR to detect two generation walls effectively. In the area where sand covers the Wall segments, long wavelength SAR'S penetration ability reveals the Wall's features and displays them on the SAR images. In terms of wavelength, the L-band is obviously better than the C-band for revealing the Wall's features. In terms of polarization, HH polarization is better than H v polarization. On the L-HH and C-HH images, the Great Wall of both the Ming and the Sui dynasties are shown, but it is difficult to see that both of them existed simultaneously. Figure 3 shows visualization models of radar backscattering intensity for the two genera- tion's of the Great Wall (on the right of the L-HH image) and the roads to the left of the Wall (only seen on the L-Hv image). Trees lining a road parallel to the Wall show up as bright rectangles because the L-HV and C channels are sensitive to complex vege- tation structure.

L-SAR Applications The Chinese airborne L-band SAR system has demonstrated its capability in monitoring natural hazards, especially the unprecedented flooding events of the Yangtze River. When the third flood peak passed through Jiu Jiang City of jiangxi Prov- ince, the L-SAR system was flown over the area, acquiring 27,700 km2 of real-time, high-resolution data for the Poyang Lake area, which experienced the highest flood stage ever recorded. SAR images of the Dongting Lake area were also acquired from 27 to 31 July, 1998. Researchers at the Institute of Remote Sensing Applications, Chinese Academy of Sciences promptly processed and mosaicked the SAR images. On the basis of human-computer interactive analysis, the total

614 Mav 2000 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING

L H H L-HV

I C H H C H V

Figure 2. L- and C-band HH and Hv SAR response from the segment of the Great Wall in Shanxi Province. On the image, 1 refers to the Great Wall of the Ming Dynasty, 2 refers to the Great Wall of the Sui Dynasty, and 3 refers to the road.

flooded area and the individual flooded areas by county (city) were calculated in detail. The inundated area was further clas- sified into farmland, towns, residential areas, grassland, and damaged fish ponds (Plate 1). Maps of the inundated areas were produced and provided to decision-making authorities.

Conclusions This paper presents research work on the applications of sIR-c/ X-SAR data in the test sites of China as well as the applications of Chinese airborne L-band SAR data for flood monitoring. In

L-mi L-HV

C-IBt

Figure 3. Radar backscatter intensity visualization model of the Ming and Sui Great Wall segments and a road parallel to the Great Wall.

the spaceborne-airborne ground-based remote sensing experi- ment, a linear relationship was derived for terrain features and image gray level. In the SAR penetration experiment for desert sands, it was found that the long wavelength SAR is able to pene- trate dry sands from a depth of several centimeters to a few meters. Five volcanoes and two types of lava flows have been discovered with SIR-C/X-SM data. Geological structures and rock types beneath a forest canopy were detected by SIR-C/X-SAR data in a subtropical region. The Great Wall segments of the Ming and Sui dynasties in a desert area were revealed with SIR- c 1 x - S ~ ~ data. An airborne L-band SAR system developed in China was used for monitoring flood events which occurred in the middle and lower reaches of the Yangtze River in 1998. Our study indicates that SAR data is a powerful tool for detecting the Earth's resources, studying the Earth's environment, and moni- toring floods and other natural disasters.

Acknowledgments The author wishes to thank Drs. Wang Changlin, Shao Yun, Liao jingjuan, Liu Hao, Wang Shixin, Wang Chao, and other colleagues at the Institute of Remote Sensing Applications, Chi- nese Academy of Sciences, for their contribution to the research work. The research discussed in this paper was sup- ported by the key research program of the Natural Science Foundation of China (49989001 and 49493403), the Hi-Tech Research and Development program of China (863-308), and the special support program of the Chinese Academy of Sci- ences (KZ951-A1-302, KZ95T-03). The SIR-CIX-SAR data were provided by NASAIJPL under an agreement between NASA and the Institute of Remote Sensing Applications, Chinese Acad- emy of Sciences. Thereby, special thanks are given to the Natu- ral Science Foundation of China, the Chinese Academy of

Legend I==l rn Flooded

Residential Amr I hags dah w n roqulred by Chlncne akbom L-band 8UI durlw Aw. 23. I-.

woiihr: cloudy

Plate 1. Chinese L-band airborne SAR imagery showing the flooding situation of Poyang Lake in 1998.

616 la i ,)(>nu PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING

Sciences, the Hi-Tech Research and Development program of China, and the National Aeronautics and Space Administra- tion of the United States of America for their support.

References Guo, Huadong, 1986. The Response of Shuttle Imaging Radar to the

Desert and Subsurface of Alashan Plateau, Environmental Remote Sensing, 1(1):34-43.

Guo, H., G.G. Schaber, C.S. Breed, and A.J. Lewis, 1986. Shuttle Imaging Radar Response from Sand and Subsurface Rocks of Alashan Plateau in North-Central China, Int. Syrnpos. Remote Sensing for Resources Development and Environmental Management 7th Pm- ceedings, ISPRS Commission W, Enschede, Netherlands, A. A. Balkema, Boston, pp. 137-143.

Guo, Huadong, Lianpu Zhu, Yun Shao, and Xinqiao Lu, 1996. Detection of Structural and Lithological Features Underneath a Vegetation Canopy Using SIR-C/X-SAR Data in Zhao Qing Test Site of South- ern China, Journal of Geophysical Research, 101(E101):23101-23108.

Guo, Huadong, Jingjuan Liao, Changlin Wang, Chao Wang, T. Farr, and D. Evans, 1997. Use of Multifrequency and Multipolarization Imaging Radar for Detection of Volcano in Western Kunlun Moun- tain Region, Remote Sensing of Environment, 52(2):364-374.

Jordan, R.L., B.L. Huneyeutt, and M. Werner, 1995. The SIR-CIX-SAR Synthetic Aperture Radar System, BEE 7kansactions on Geosci- ence and Remote Sensing, 33(2):829-839.

JPL, 1996. Space Radar Reveals Ancient Segments of China's Great Wall, Earth Observation Magazine, 8(3):56.

Shao, Yun, Huadong Guo, and Hao Liu, 1995. An Experiment on Pene- tration Study of Shuttle Imaging Radar SIR-C/X-SAR, New Prog- ress of Remote Sensing Science, Science Press, Beijing, pp. 161-168.

Stofan, E.R., D.L. Evans, C. Schrnullius, B. Holt, J.J. Plaut, J, van Zyl, S.D. Wall, and J. Way, 1995. Overview of Results of Spaceborne Imaging Radar-C/X-Band Synthetic Aperture Radar (SIR-CIX- SAR), IEEE k s . Geosci. Remote Sensing, 33(4):817-828.

Sun, G.Q., and K.J. Ranson, 1995. Forest Ecological Studies and Radar Backscatter Modeling, Microwave Remote Sensing for Earth Observation, Science Press, Beijing, pp. 213-223.

Wang, Chao, Haudong Guo, and Lin Li, 1996. Spaceborne-Airborne- Ground-Based Remote Sensing Experiment, Science Bulletin, 41(4):377-380.

(Note: The customary western practice of listing author's family names last, except in the list of references where only the first author's name is listed family name first, is followed herein.)

Special Session on ASPRSINASA-CRSP Ten-Year Remote Sensing Industry Forecast

Wednesday, May 24 3:30-5~00 PM

Mark your calendar to attend a special session on the ASPRS/NASA-CRSP Ten-year Remote Sensing Industry Forecast. The session, to be held during the ASPRS Annual Conference in Washington, DC, will be moderated by ASPRS Primary Data Acquisition Division Director Charles Mondello, EarthData Technologies.

ASPRS and the NASA-Commercial Remote Sensing Program are conducting a 10-Year Remote Sensing Industry Forecast, which began in October 1999. The study will provide a baseline to determine the current state of the remote sensing industry, and advance during the next 5 years to establish a "rolling horizon" of forecast data for remote sensing applications, business development, technology opportunities, and investment guidance.

The panel session will report the progress of the study: Study Group Organizations and Staff Industry Definitions Forecast Study Areas

Definitions of Customer Segments Forecast Framework Review of the ASPRS Membership Survey Document

Panelists: Ron Rabin, Lockheed Martin Space Operations-Stennis Prog.

William Piper, NASA-CRSP Mike Renslow, ASPRS President

Nate Boyer, Eastman Kodak Mindy Brown, Space Imaging

Alan Mikuni, US Geological Survey John Schott, Rochester Institute of Technology

Len LaFeir, Autometric, Inc.

PHOTOGRAMMETRIC ENGINEERING 81 REMOTE SENSING May 2000 617