The Following Are the Articles Reviewed for This Assignment

Leonard Kemp

Intro to Remote Sensing

ES 5053

The following are article titles, authors, and abstracts reviewed for this assignment:

“The Importance of SAR Wavelength in Penetrating Blow Sand in Northern Arizona,” in Remote Sensing of the Environment 69:87-104, (1999), by Gerald G. Schaber and CarolS. Breed

The authors demonstrate the ability of C-band, L- band, and P-band radar to image a sand streak using Airborne Synthetic Aperture Radar (AirSAR). Multifrequency and polarimetric images of the feature are qualitative and quantifiable analyzed from best to worst.

“SAR Studies in the YumaDesert, Arizona: Sand Penetration, Geology, and the Detection of Military Ordnance Debris,” in Remote Sensing of the Environment 67:320-347 (1999) by Gerald S. Schaber

The authors demonstrate the ability of C-band, L- band, and P-band to image subterranean geologic and cultural targets through the use of Synthetic Aperture Radar in the YumaDesert, Arizona. The radar bands were able to penetrate through several meters of blow sand and sandy alluvium to image three distinct geologic features. In addition, the utility of the L-band, and P-band in the cross- polarization modewas used successfully discriminate metallic objects on the surface and subsurface.

Introduction

The following will review two articles related to the capabilities of multifrequency and polarimetric radarto image geologic features in arid and semi-arid environments. Airborne Synthetic Aperture Radar (AirSAR) was flown over two sites in Arizona utilizing the C-band (6-cm wavelength), L-band (24 cm wavelength), and P-band (68 cm wavelength) in all four transmit receive polarization modes and imaged geologic and cultural features. This paper will relate the history of SAR research, the areas of study, data, methods of collection, the conclusion of the studies, and its potential application to archaeology. The following review willprimarily reference Schaber:1999, and Schaber and Breed:1999, except where noted within the text.

History of SAR Research and Area of Studies

The Arizona SAR projects are a continuation of ongoingglobal-scale research which seeks to image geologic features mantled by blow sand and/or sandy alluvium through the use of remote sensing (Schaber and Breed; Schaber). The following will synthesize the history of SAR research in arid environments, a brief outline of Airborne Synthetic Aperture Radar, and the areas of study utilized in these two articles.

Beginning in the 1980’s, NASA instituted Shuttle Imaging Radar (SIR-A, SIR-B) to image subsurface geologic features of the northeastern SaharaDesert utilizing the L-band. Subsequent missions imaged Badan-Jaran Desert of China, the Mohave Desert in California, and the Saudi Arabian peninsula. This research led to the observation that the P-band could image geologic features several meters below the surface. Ongoing research by Schraber et al. documents the use of multifrequency and polarimetric wavelengths to image and create a geologic map of features at Bir Safaf in southern Egypt.

The Airborne Synthetic Aperture Radar (AirSAR) was designed, built, and is managed by the Jet Propulsion Laboratory (JPL). AirSAR is a side-looking radar instrument which can collect multifrequency data at three radar wavelengths: C-band (6-cm wavelength), L-band (24 cm wavelength), and P-band (68 cm wavelength); a DC-8-72 aircraft serves as a platform for this technology. AirSAR isa component of NASA research programs for demonstrating radar technology and data acquisition for the development of radar processingtechniques and applications. AirSAR first flew in 1988 and continues to conduct a minimum of one research flight per year. The preceding was referenced from the following:

Both areas of study are located in Arizona; they are the Ward Terrace study area and the YumaDesert study area. The Ward Terrace study site lies within the southern Great BasinDesert in northern Arizona. The terrace is underlain by sedimentary rocks. The terrace is composed of water impervious silicious layers; the terracecannot support vegetation other then grass. The “sand streak” is 7 km long, and .5 km wide with a max. depth of 2.5 m.; it extends northeast across the terrace where it supplies sand to a dune at the base of the Moenkopi Plateau. Based upon laboratory analysis the sand streak consists of well sorted quartz sand with overall soil moisture of 2.2% (by weight).

The area of study is located in the southwestern corner of Arizona; itis located on the U.S. Marine Corps Air Station in the YumaDesert. The YumaDesert is one of the most arid areas on the North America continent;it receives approx 75mm of precipitation annually. The area of study is a windswept plain with a surface of alluvial pebbles and coarse sand over semi-consolidated alluvium and eolien deposits. The surface is overlain by modern eolian sand sheetswhich in places is several meters thick. The sparse vegetation of the site consists mainly of creosote bushes, and grasses. The moisture sample collected on the date of the flight had moisture content between 2.4% to 3.9% (by weight) at a depth of 5 cmand 40cm below the surface.

Data and Methods

In May of 1991, AirSAR took multifrequency and polarimetric images of the Ward Terrace by NASA/JPL AirSAR platform. These images were collected in the C-band, L-band, and P-band in all polarization modes. The incidence angle over the area of study site ranges from Θ=42.5˚ at the top of the sand streak/plateau to Θ=55˚ at the bottom below the southwest scarp. Due toin-flight navigational computer difficulties, the images were “smeared” although the author states while loss of sharpness was noted, the image was still readable in the study area. The data were processed and calibrated using JPL SAR processor version 3.56.

In March of 1990, AirSAR also took multifrequency and polarimetric images of the YumaDesert study site specifically the Barry Goldwater Bombing and GunneryRange located on the U.S. Marine Corps Air Station between the YumaValley and GilaMountains. Due to electronic interference from military and commercial radars located in this area, the L- and P-bands were contaminated. The incidence angle ranged from Θ=25˚ to Θ=55˚ with an image resolution of 5 m. All SAR data was processed at the JPL using the SAR Processor version 3.55.

The authors noted that for both studies only the HH polarization mode is presented are included in these articles. This is because the HH and VV polarization modes were almost identical in backscatter responses.

Results of Analysis

Subsurface imaging of geologic features from both space and airborne sensors has been of arid environments where elevation is relatively level and contained subsurface structures that backscatter strongly. The Arizona SAR study expands upon this research. It demonstrates that diverse wavelengths and polarization modes can image varied geologic and cultural features in arid and semi-arid environments. The following section presents the qualitative and quantitative analysis of SAR images of Ward Terrace and the YumaDesert sites, respectively, by the authors.

The Ward Terrace sand steak was imaged and successively penetrated deeper in longer wavelengths. In band- C (HH), the sand streak is sharply contrasted due to strong backscatter from surrounding surface of lag-covered bedrock. This lag surface is mainly composed of fragmented chalcedony, chert and calcrete. In band-C (HV), the sand streak remains distinctive. In band-L (HH), the sand streak is discernable but with very poor contrast between terrace surface. Band-L (HV), the sand streak is almost invisible. The authors note this image, band-l (HV), may be compromised by technical problems.Band P penetrates through sand streak approximately 2.5 in depth in both polarization modes and the sand streak appears to be gone. Again note image change of surface from radar bright to radar dark due to increase of scattering and signal penetration. The following utility table details the geologic feature from best to worst image resolution for the Ward Terrace.

(Schaber and Breed: 1999)

Quantitative analysis of Ward Terrace is in accord with preceding qualitative analysis of SAR images. The authors plotted a backscatter coefficient to polarization mode (see below). The most dramatic backscatter contrast occurs between the sand streak and surrounding lag-covered surface in the C-band

(Schaber and Breed: 1999)

The Yuma desert site images succeeded in imaging the following geologic features and the target ranges. The geologic features which are the areas of interest are described as follows:

Sand-mantles alluviumis a sandy to small pebble alluvium mantled by approx 20/ 30 cm to several meters of active blow sand which supports vegetation year- round; scald- are flat surfaces created by eolian deflation of the sand mantle, surface characterized by a lag deposit of small pebbles and dead vegetation, and coarse-grained river gravels- older gravels of the Colorado and Gila Rivers, this is composed of cobble gravels up to 25cmin diameter.

In Band C (HH), the sand-mantled alluvium is relatively radar dark due to low backscatter and minimum depth penetration of 20-40cm below surface at 5,7cm wavelength. In band- L(HH), and band-P(HH), the sand mantled alluvium is imaged as intermediate and radar bright, respectively. Band-P is bright because it is penetrating and backscattering off the alluvium and CaCO3 nodules found in subsequent backhoe trench testing. The scald surfaces in the C-band(HH) image are characterized as strong in contrast to surrounding sand mantled surface; in L(HH), and P(HH), the scald emits a relatively weak backscatter. This is thought due to the relative small size of gravel within these scalds. The coarse river gravels are radar bright in the C(HH) image, and not as distinct on the C(HV)image. On the L(HH) image, the coarse river gravels become undistinguishable from surrounding surface, in the P(HH) image, the bed upon which the gravel rested is penetrated and become radar dark.

Target ranges are used by military pilots for firing practice. The target ranges consists of debris fields of non-explosive military aircraft cannon projectiles, earthen mounds and old automobile tires which are used to enhance the visibility of targets. In addition, road and fence lines were also documented in the Yuma Desert SAR images. According to the authors,the ability to separate metallic ordnance debris from ground clutter varies considerably due to wavelength and polarization. In the P-band, ordnance is highly radar reflected in cross polarization due to reduction of signal strength which reduces the ground clutter from ordnance. Secondly, the use of P-band allows maximum contrast between debris which is radar bright and what can be described as a radar dark desert floor. The following utility table details the geologic feature from best to worst image resolution for the YumaDesert site.

(Schaber:1999)

Again, the quantitative analysis of YumaDesert site is in accord with the preceding qualitative analysis of SAR images. The authors plotted a backscatter coefficient to polarization mode (see below). This analysis confirms the inverse backscatter relationship between the sand-mantled alluvium and the coarse river gravels. In addition, Figure C confirms the supposition that the desert is highly contrasted to the metallic ordnance in the P-band and L-band.

(Schaber: 1999)

Radar Application to Archaeology

The history of archaeological method is of a discipline which has used sophisticated technologies in the search for and understanding of archaeological sites. Indeed, the first aerial image of Stonehenge was taken in 1906 in a military balloon. Today, LIDAR image of Stonehenge are analyzed to detect ancient roadways and other archaeological features (Fowler:2002). The following will briefly discusses the application of SAR technology to image archaeological features.

The discovery of SAR to image geologic features in the early 1980’s, led archaeologist to test the application of SAR technology to image archaeological features in conjunction with NASA. This research led to the discovery that multi-band and multi-polarization radar can interact with phenomena which is associated with human occupation and activities. Human occupation may be defined by geometric shapes that suggest subsurface architectural remains, and areas with dielectric properties which backscatter strongly(Comer, and Blom).In addition, the application of SAR technologies allows archaeologists to predict areas of high probability of prehistoric human occupation based upon relictfluvial features (Comer and Blom). The testing of SAR technology has focused on archaeological areas with architectural features; this research includes the Roman site of Petra, Jordan, and the Khmer site of Angkor, Cambodia.

Recently, Comer and Blomhave proposed that AirSAR can be used in conjunction with GIS to build a model of distribution of prehistoric archaeological sites and features. They have proposed flights over San Clemente Island, California to image prehistoric archaeological sites. The radar image will be correlated to known sites to test the ability of the image to discern them. Finally, a team will be led to ground-truth to verify the validity of the predictive of the model (Comer and Blom). To date, there is no additional information available concerning the results of this research plan.

Conclusion

The Arizona SAR studies demonstrate the ability of multifrequency and polarimetric radar to image both geologic and cultural features in varied arid and semi-arid environments. The authors demonstrate the ability of C-band, L- band, and P-band to image subterranean geologic and cultural targets through the use of SAR at the Ward Terrace site and the Yuma Desert, Arizona.The authors note that much work still needs to be done to quantify electromagnetic properties relative to contaminants in arid and semi-arid environments through laboratory and field work. In addition, the amount of radar penetration relative to wavelength still needs to be researched. The potential of utilizing SAR to image archaeological features may be viewed as a relativelyunusual approach to the investigation of archaeological sites. If the application of airborne radar technology to archaeological research is to be more then a novelty more study needs to be initiated with the results of that research published.

Reference

n.d.Comer, Douglas C. and Blom, Ronald G.

Detection and Identification of Archaeological Sites and Features Using Radar Data Acquired from Airborne Platform,

2002Fowler, Martin J.F.

Satellite remote Sensing and Archaeology: A Comparative Study of Satellite Imagery of the Environs of Figsbury Ring, Wiltshire, in Archaeological Prospect, 9:55-69

1999Schaber, Gerald S.

SAR Studies in the YumaDesert, Arizona: Sand Penetration, Geology, and the Detection of Military Ordnance Debris, in Remote Sensing of the Environment 67:320-347

1999Schaber, Gerald G. and Breed, Carol S.

The Importance of SAR Wavelength in Penetrating Blow Sand in Northern Arizona, in Remote Sensing of the Environment 69:87-104