The Use of Historical Imagery in the Remediation of an Urban Hazardous Waste Site

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, VOL. 4, NO. 2, JUNE 2011 281

The Use of Historical Imagery in the Remediation ofan Urban Hazardous Waste Site

E. Terrence Slonecker

Abstract—The information derived from the interpretation ofhistorical aerial photographs is perhaps the most basic multi-temporal application of remote-sensing data. Aerial photographsdating back to the early 20th century can be extremely valuablesources of historical landscape activity. In this application, im-agery from 1918 to 1927 provided a wealth of information aboutchemical weapons testing, storage, handling, and disposal of thesehazardous materials. When analyzed by a trained photo-analyst,the 1918 aerial photographs resulted in 42 features of potentialinterest. When compared with current remedial activities andknown areas of contamination, 33 of 42 or 78.5% of the featureswere spatially correlated with areas of known contamination orother remedial hazardous waste cleanup activity.

Index Terms—Chemical weapons, historical imagery, photo in-terpretation, World War I.

I. INTRODUCTION

T HE intentional or accidental release of hazardous sub-stances into the environment is an inevitable consequence

of anthropogenic activity. Industrial, commercial, mining, mili-tary, and even domestic activities can result in the release of sub-stances into the air, land, and water that are harmful to environ-mental quality and human health. The combined industrializa-tion and population growth of the twentieth century has resultedin an unprecedented release of fugitive contamination that todaythreatens many plant and animal species and may ultimatelythreaten the survival of the human race [1]. The discovery, detec-tion, and remediation of many hazardous waste problems con-sists of a variety of monitoring and analysis strategies that aretime-consuming and expensive, such as detailed field samplingfor detection.

One of the technologies that has an established and growingpotential to provide a noncontact and cost-effective alternativeto traditional sampling methods is through the use of remote-sensing techniques. The paper documents the critical role thathistorical remote-sensing data played in the remedial investiga-tion of Spring Valley, a Formerly Used Defense Site (FUDS) inWashington, D.C.

Manuscript received October 20, 2009; revised January 28, 2010; acceptedMarch 16, 2010. Date of publication June 01, 2010; date of current version May20, 2011. This work was supported in part by the U.S. Geological Survey andthe U.S. Environmental Protection Agency. Opinions expressed here are thoseof the author and do not necessarily reflect the official views of either agency.

The author is with the U.S. Geological Survey/Eastern Geographic ScienceCenter, Reston, VA 20192 USA (e-mail: [email protected]).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/JSTARS.2010.2049254

Hazardous waste control, monitoring, remediation, and re-lated issues result in a staggering cost to society in terms ofhuman and ecological health effects, negative externalities onreal estate values (i.e., the “stigma” of contamination), and theextraordinary burden of a massive regulatory infrastructure oneconomic productivity.

Substances are considered hazardous wastes if they are ig-nitable (capable of burning or causing a fire), corrosive (ableto corrode steel or harm organisms because of extreme acidicor basic properties), reactive (able to explode or produce toxicgasses), or toxic (containing substances that are poisonous topeople and other organisms) [2]. In the United States (U.S.),the regulatory definition of hazardous substances are detailedin the Resource Conservation and Recovery Act (RCRA) andcan be found under specific listings, along with accepted testingmethods, in Chapter 40, Code of Federal Regulations, Section261 (40 CFR § 261) [2].

The U.S. Environmental Protection Agency (USEPA) esti-mates that complying with hazardous waste regulations costnearly $32 billion in 2000, which is about 20% of the cost forall U.S. pollution control laws [3]. In 1998, the U.S. Agencyfor Toxic Substances and Disease Registry (ATSDR) evaluatedthe medical and lost productivity costs from health conditionsoccurring in U.S. communities located near hazardous wastesites that were contaminated with volatile organic compounds(VOCs). For the 258 sites studied, the annual costs were in ex-cess of $300 million [4].

Most hazardous waste is the by-product of industrial or com-mercial manufacturing processes but significant levels of haz-ardous substances are associated with agricultural chemicalssuch as pesticides. However, even household waste containssubstances such as bleach, gasoline, batteries, and solvents thatqualify as hazardous wastes. Hazardous waste can also be natu-rally occurring substances, such as lead and mercury, which arebrought in much higher than normal exposure concentrationsby anthropogenic processes, such as mineral mining, metal pro-cessing, electroplating, and pesticide production.

The effects of fugitive hazardous waste are multiple andvaried. Hazardous wastes may pollute soil, air, surface water,or underground water. Pollution of soil may affect people wholive on it, plants that put roots into it, and animals that moveover it. Pollution may concentrate in individual organismsand up the food chain, with serious additive effects in highertrophic organisms in processes called bioaccumulation andbiomagnification [5].

The human health effects from exposure to fugitive hazardouswastes are highly variable but can range from acute toxicity andimmediate danger to life to chronic exposures and a wide range

1939-1404/$26.00 © 2010 IEEE

282 IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, VOL. 4, NO. 2, JUNE 2011

of health effects. Many studies have shown that, in residentialcommunities near a hazardous waste site, there is an increasedrisk of birth defects, neurotoxic disorders, leukemia, cardiovas-cular diseases, and respiratory and skin disorders [6]–[9]. Ina comprehensive review of the literature, Vrijheid [10] notedthat there is a general weakness in almost all human health-haz-ardous waste site studies in that there is a lack of direct exposuremeasurement, and this is a significant research need in order tobetter understand and quantify the risk of residential exposureto hazardous waste sites. Hazardous waste sites can have seriouseconomic ramifications on the value of the contaminated prop-erty, as well as the value of surrounding properties and neigh-borhoods.

Research suggests that residential dwellings located near haz-ardous waste sites experience a negative impact on propertyvalues and that this impact is directly related to distance fromthe site, generally disappearing between 3.2–4.8 km from thesite [11]. Interestingly, some research suggests that the USEPA’slisting of a hazardous waste site on the National Priorities List(NPL) actually has a positive effect on surrounding real estatevalues because it leads to a formal cleanup plan and tends to re-move uncertainty that affects market values [11]. Commercialreal estate values are often more severely affected by proximityto a hazardous waste site because of the fear of liability issues.In recent years, some companies have actually been able to re-cover these “stigma damages” through legal action [12].

On a global scale, the issue of hazardous waste is evenmore serious. During the 1980s, the development of strictenvironmental controls on hazardous waste in industrializedcountries, such as the Superfund Act in the U.S., resulted in ablack market for unregulated translocation of hazardous wastesfrom industrial to third-world countries [13]. Internationaloutrage at the activities of “toxic traders” led to the draftingand adoption of the Basel Convention of 1989. Both the BaselConvention and the Rotterdam Convention of 1998 seek tostem the trans-boundary movement of hazardous substancesand hazardous waste, especially to developing countries. Also,the central goal of the Basel Convention is “environmentallysound management” to protect human health and the environ-ment by minimizing hazardous waste production wheneverpossible and promoting an “integrated life-cycle approach”to hazardous waste management which involves promotinginstitutional controls from the generation of a hazardous wasteto its storage, transport, treatment, reuse, recycling, recovery,and final disposal [14].

In the U.S., land disposal of hazardous waste and hazardouswaste cleanup of contaminated soil are primarily regulated bytwo federal statutes that are administered by the USEPA. TheRCRA was enacted by Congress in 1976 in order to: 1) pro-tect human health and the environment from the potential haz-ards of waste disposal; 2) conserve energy and natural resourcesand to reduce the amount of waste generated; and 3) ensure thatwaste is managed in an environmentally sound manner. TheComprehensive Environmental Response, Compensation, andLiability Act (CERCLA) of 1980, commonly known as “Super-fund,” and the 1986 Superfund Amendments and Reauthoriza-tion Act (SARA) provide strict guidelines regarding the respon-sibility of past and present property owners, as well as others,

for the cost of toxic-waste cleanup. Superfund empowers theUSEPA to compel the owner of property contaminated by haz-ardous substances to clean up the site. Alternatively, the USEPAcan conduct the cleanup and obtain reimbursement from the re-sponsible parties at a later date. However, other federal statutes,such as the Toxic Substances Control Act (TSCA), the CleanAir Act (CAA), and the Clean Water Act (CWA) also regulatehazardous substances, especially in air and water mediums.

A special class of hazardous waste problems relates to cur-rent or former military activity at active military facilities and atproperties that were formerly owned by, leased to, or otherwiseutilized by the U.S. under the jurisdiction of the Secretary ofDefense. Such properties are known as Formerly Used DefenseSites (FUDS). The FUDS program is funded under the DefenseEnvironmental Restoration Program (DERP), mandated underSARA, and administered by the U.S. Army Corps of Engineers(USACE). There are over 4400 currently identified FUDS withan estimated cost to complete cleanup of between 15 and 20billion dollars [15]. One of the costliest and most politicallycharged cleanups of a FUDS is currently taking place in north-west Washington, D.C., at the American University and the sur-rounding neighborhood, known as Spring Valley. The objectiveof the research described here is the evaluation of the value ofhistorical aerial photographs for the cleanup of this hazardouswaste site where remediation activity for unexploded ordnance(UXO) and inorganic arsenic is still in progress.

A. Legacy of Chemical Weapons (CW)

Perhaps the most insidious and dangerous forms of haz-ardous waste are those substances that are precisely engineeredand manufactured for the sole purpose of ending human lifeas quickly as possible. Chemical weapons, as defined by theUnited Nations,“… are chemical agents of warfare … whethergaseous, liquid or solid, which might be employed becauseof their direct toxic effect on man, animals and plants” [16].The North Atlantic Treaty Organization (NATO) definition of achemical agent is a “chemical substance which is intended foruse in military operations to kill, seriously injure or incapacitatepeople because of its physiological effects” [17].

Poison gas weapons were first developed and extensivelydeployed during World War I (WWI). Although some accountsdiffer, the first effective deployment of CW occurred at thebattle of Ypres, Belgium, on April 22, 1915. Chlorine gaswas released by German forces and killed over 5000 alliedtroops. Both sides deployed and continued to develop CWthroughout the conflict with a total of over 92 000 killed and1.3 million wounded by the end of the war [18]. Research intonew and more deadly types of CW continued throughout thewar and saw the development of Mustard Gas and deadly formsof inorganic arsenic and arsenical compounds such as arsinegas, Lewisite (2-chloroethenyldichloroarsine), and Adamsite(diphenylaminechlorarsine) [19].

When the U.S. finally entered WWI in 1917, one fact was dis-turbingly apparent; it was woefully behind in the developmentand utilization of CW and countermeasures such as protectiveclothing and gas masks. An intense research and development

SLONECKER: USE OF HISTORICAL IMAGERY IN THE REMEDIATION OF AN URBAN HAZARDOUS WASTE SITE 283

Fig. 1. Spring Valley study area. The boundary of the Spring Valley study sitein northwest Washington D.C. overlain on the 1994 USGS digital orthophotoquarter quadrangles (DOQQ), Washington West NW and Washington West SW.

program was initiated at several locations around the country in-cluding the American University, located in what is now north-west Washington, D.C. The U.S. Army leased the entire campusof the college and an additional 243 hectares (ha) for the purposeof research, development, and field testing of chemical warfareagents and countermeasures [19] (see Fig. 1).

The legacy of that CW testing and development still exists inthe area today in the form of UXO and soil and groundwatercontamination from inorganic arsenic and other chemical com-pounds. The American University and the surrounding neigh-borhood, known as Spring Valley, is currently the focus of amajor FUDS cleanup.

B. Remote Sensing and Hazardous Waste

The process of discovering, characterizing, and remediatingfugitive contaminants in the environment is typically a long andcostly endeavor. The current cleanup of Spring Valley has beenunderway for nearly ten years with at least another five to go andhas cost in excess of $40 million [15]. In the hazardous wasteremediation process, one of the key steps is Site Characteriza-tion, which is the determination of the spatial extent and natureof the contamination. Site characterization is often costly andtime-consuming, requiring extensive field sampling and labora-tory analysis. One technology that has been valuable in cleanupefforts and shows promise in providing an alternative to fieldsampling methods is through the use of remote sensing tech-niques.

Remote sensing is a generic term that encompasses a bodyof noncontact monitoring techniques that measure energy in-teractions to determine the characteristics of a target surface ormedium. Although remote sensing includes a wide variety ofinstruments and methods, such as LiDAR, radar, X-ray tech-nology, and acoustic instruments, it is most often associatedwith overhead imaging techniques, such as aerial photographyand satellite imagery that record energy in the solar-reflectedpart of the electromagnetic spectrum (EMS) between 400–2500nm wavelengths. Remote sensing has a long history of providing

critical information to the process of identifying, characterizing,and remediating hazardous waste problems [20]–[22]. Further,new and emerging remote-sensing techniques show promise forcharacterizing site conditions and providing critical informationto the hazardous waste cleanup process.

C. Aerial Photographic Interpretation

One of the most fundamental techniques of remote sensingis that of aerial photographic interpretation. Aerial photographshave been routinely collected over the conterminous U.S. sincethe 1930s for engineering and agricultural purposes, and theyrepresent a rich archive of historical changes on the landscape.The analysis of aerial photographs has been used for decadesto assist in hazardous waste investigation and remediation, andthere is a long history of successful applications of this form ofremotely sensed data in environmental monitoring.

The USEPA has produced over 4000 historical aerial photo-graphic reports on hazardous waste activity that have been in-strumental in environmental cleanup programs [23]. These re-ports use the interpretation of historical aerial photographs todetail landscape activities, such as burial areas, landfills, vege-tation stress, and ground disturbance, as indicators of possiblesurface and subsurface contamination. While the use of histor-ical aerial photographs has often been employed, there havebeen few, if any, assessments of the accuracy or value of his-torical aerial photographic interpretation in the hazardous wastesite cleanup process. Formal tests for the accuracy and consis-tency of photographic interpretation were developed by Con-galton and Mead [24], but these largely revolved around interan-alyst variability and were dependent on a relatively recent timeframe in which landscape conditions were the same as when thephotograph was taken and data could be field-verified. There areinherent difficulties associated with evaluating subjective inter-pretation of historical photographs, most notably the inability toperform field checks or obtain reliable ground reference data foraccuracy assessment.

The Spring Valley FUDS remediation offers an excellent op-portunity to perform a relative accuracy evaluation of photo-de-rived information by utilizing Geographic Information Systems(GIS) layers of information. By first geo-registering historicalaerial photographs to a common coordinate system, photo-in-terpreted features, such as ground scars and pits, can be digi-tized into GIS format and overlaid on a current map of areasof known contamination. This permits direct spatial correlationbetween features derived from historical aerial photographs andthe areas of contamination requiring eventual remediation. Theresults of this analysis provided a good indication of the valueof historical photo-derived information for hazardous site reme-diation.

II. SPRING VALLEY, ARSENIC, AND THE AMERICAN

UNIVERSITY EXPERIMENT STATION

The historical background and context of the hazardous wastecontamination and cleanup issues in Spring Valley are an im-portant component to understanding the unique problems andchallenges of this type of hazardous waste cleanup. The SpringValley site is one of the highest profile FUDS remediation efforts


Fig. 2. Ordnance Testing at the AUES: Above: A CW-laden mortar shell ex-plodes releasing shrapnel and chemical agent to volatilize in the air. Below: fieldtesting of smoke candles (Photos courtesy USACE).

in the U.S. New discoveries and remediation activities there at-tract the attention of the news media, members of Congress, andDistrict of Columbia officials [15].

WWI was the first global-scale conflict to employ the tech-nology of the industrial revolution on the battlefield. Althoughaircraft, submarines, tanks, wireless communication, and in-creasingly sophisticated artillery changed the face of warfareforever, perhaps the most significant of the emerging tech-nologies was the widespread use of CW by both sides in thisconflict. WWI was the first widespread use of such weaponry inhuman history. When the U.S. formally entered the conflict onApril 6, 1917, military leaders were keenly aware that the U.S.did not possess the technology or expertise to deal with the CWthat was being deployed on the European battlefields [25].

Therefore, in order to rapidly develop and test CW andcountermeasures, the U.S. Federal Government leased theentire campus of the American University and an adjacent 243hectares from five private landowners. The facility was renamedthe American University Experiment Station (AUES) and, byOctober 1917, laboratories, facilities, test ranges, and provinggrounds had been constructed for the development and testingof CW such as Mustard Gas and Lewisite [25]. Research andtesting included both offensive and defensive measures. Thisincluded ordnance and delivery mechanisms such as artilleryshells and grenades, human and animal toxicity, chemical per-sistency in the environment, and poison gas countermeasuressuch as gas masks and protective fabric for clothing, tents,

Fig. 3. Livens Battery Firing at AUES. Live fire tests utilizing Livens mortarsfired from the western edge of American University to the west towards theDalecarlia Woods. (Photographs courtesy of USACE.)

Fig. 4. Examples of UXO. Unexploded 75-mm artillery shells excavated froma pit in Spring Valley. Some shells still contained explosives and CW agents.(Photographs courtesy of USACE.)

and trench curtains. Figs. 2 and 3 show ground photographsof weapons and smoke candle testing being conducted at theAUES during the 1917–1918 operations. Fig. 4 shows some ofthe shells extracted from the various burial pits.

When the armistice was signed on November 11, 1918, theAUES was in full operation and found itself suddenly withoutan urgent mission. Over the next several months, the station wasshut down. Troops and staff were released, buildings were razed,


and much of the chemical and munitions inventory was eithermoved or buried at various locations around the AUES property.As a result, the Spring Valley area was left with a toxic legacy ofburied munitions and soil contamination from inorganic arsenic.

The area today mostly consists of single-family residentialhomes. American University occupies most of the originalAUES ground in the southeast corner of the study area andSibley Hospital is located in the southwest corner (see Fig. 1).In spite of anecdotal evidence, the environmental situationwent generally unnoticed until March 1986 when officialsof American University, because of planned building con-struction, requested that the USACE conduct an investigationinto possible buried chemical munitions on the campus. TheUSACE completed its investigation in October of that year andconcluded that there was no evidence of buried munitions andno further action was necessary.

In 1993, a construction crew accidentally unearthed several75-mm artillery shells near the 52nd Court cul-de-sac in thenorthwest part of the study area. This resulted in a full Su-perfund remedial investigation. Although the USACE removedover 200 pieces of ordnance from a burial pit near the 52ndCourt cul-de-sac, the final recommendation of the Remedial In-vestigation, signed by both the USACE and the USEPA was “Nofurther action is necessary” [26].

In 1998, at the request of the Health Department of the Dis-trict of Columbia (DCDOH), the USACE returned to investi-gate possible munitions on the property of the Korean Ambas-sador on Glenbrook Road. Two major burial pits were discov-ered, resulting in the removal of another 200+ pieces of ord-nance and other items. This resulted in the initiation of a secondSpring Valley investigation (termed Operation Safe Removal)and has resulted in several significant UXO extractions and thediscovery of elevated levels of arsenic in the soils around SpringValley [27]. This work is ongoing, and this research will con-tribute to the cleanup and risk assessment decisions in the cur-rent remediation effort.

III. PHOTOGRAPHIC INTERPRETATION AND HISTORICAL

AERIAL PHOTOGRAPHY

The first known aerial photographs were taken in 1858 onboard a tethered balloon by a Frenchman named Gaspard FelixTournachon. Tournachon, who later became known by the nick-name “Nadar,” successfully photographed the landscape aroundParis, France. Shortly thereafter, the aerial perspective proved tobe so valuable that General George McClellan used tethered bal-loons to photograph and study enemy positions in the U.S. CivilWar [28], [29]. Until the early 1900s, balloons, kites, and evenpigeons were used as platforms to hoist cameras above the landto photograph the surface below [30]. These platforms were,however, relatively stationary, limited in altitude and range andvulnerable [31].

After the advent of the airplane in 1903, the value of air-craft-based photography became readily apparent to many,including Wilbur Wright, who took the first recorded pho-tographs from an airplane in 1909. The airplane soon becamethe primary platform for the acquisition of overhead photog-raphy. Regular use of cameras from airplanes continued untilWWI, when the formal development of reconnaissance systems

and photographic interpretation science became so accuratethat they completely changed the tactics of battlefield warfare[30]. Between WWI and WWII, the continuing developmentof both aircraft and photographic technology made the use ofaerial photographs commonplace for military and intelligenceapplications, domestic mapping, planning and natural resourcemanagement

During WWII, the role of aerial photography and interpre-tation was so critical that it prompted German General Oberstvon Fritsch to predict in 1938 that the nation with the best photo-graphic interpretation will win the next war. It is estimated thatbetween 60%–90% of all Allied intelligence was derived fromphotographic interpretation of enemy-held areas [30]. It wasalso during this period that the scientific discipline and trainingin aerial photography and interpretation began to become estab-lished. Formal lines of study for basic photographic interpreta-tion, photogrammetry, and cartography were developed for themilitary, and after the war, for civil purposes such as geology,agriculture, and forestry.

Universities developed courses on photographic interpreta-tion and photogrammetry and adapted the technology to an ever-increasing base of applications. The art and science of photo-graphic interpretation, along with photogrammetry and cartog-raphy, continued to develop throughout the post WW II periodand to expand into new areas such as nonphotographic imagingsensors like radar and thermal infrared (TIR) imaging. In 1961,the first photographs were taken aboard an orbital platform, andsatellite remote sensing was born. With the application of com-puter graphics and database management software, the revo-lution in GIS technology began in the 1980s, and aerial pho-tographs became a regular part of the typical GIS database de-velopment process.

As new systems, sensors, and technologies became available,it became increasingly clear that the routine use of aerial pho-tographs for a number of purposes had created a tremendousarchive of historical information that could not be duplicatedby the most sophisticated of current technologies. Applicationsof historical imagery began to develop and continue to this daybased solely on the ability of the aerial photographs to freeze,record, and document a moment in time [31].

In the 1970s, environmental awareness of the hazardouswaste problems of the U.S. came to the forefront of publicconscientiousness and led to the passage of the CERCLA(Superfund) in 1980. The main goal of the Superfund programwas to clean up abandoned hazardous waste sites, and, just asin Spring Valley, the archive of historical aerial photographs be-came an invaluable source of information relating to landscapeactivities of the past. Historical aerial photographic analysis ofabandoned hazardous waste sites became, and still is, a routinepart of the CERCLA remedial investigation process.

Titus [32] documented the critical use of historical photog-raphy in the assessment of hazardous waste activity relatingto the groundwater contamination from the textile industries inWoburn, MA. The use of historical aerial photographs and mapsin the study of hazardous waste sites was further developed bya number of researchers where a temporal chronology of land-scape events was reconstructed from the analysis of aerial pho-tographs and maps [33]–[36]. Nelson et al., Philipson et al., and


TABLE IKEY HISTORICAL AERIAL PHOTOGRAPHS OF SPRING VALLEY

Barnaba et al. historical utilized aerial and satellite photographsto perform area-wide inventories of abandoned waste sites aswell as to develop a methodology for prioritizing sites for re-medial action [37]–[40]

IV. METHODOLOGY

A search of government and commercial imagery archiveswas conducted in order to identify historical imagery in a timeframe that may be of value for the identification of landscapeactivities that were the direct result of historical AUES-relatedactivity. Imagery research was conducted at the U.S. NationalArchives and Records Administration (NARA), the U.S. Li-brary of Congress (LOC), American University (AU) and otherplaces for aerial photographs, reports, monographs, newspaperarticles, and any other records that potentially contained infor-mation about the WWI activities at the AUES. Although sev-eral historical aerial photographs were discovered and acquired,three proved to be critical to the eventual hazardous waste inves-tigation and are listed in Table I.

Duplicate negatives of all relevant images and copies of allreports and maps identified in the historical research were ob-tained. Aerial photographic interpretation was accomplished ac-cording to standard USEPA methods [41]. Photo analysis wasconducted by viewing film positives on a backlit light table withdouble ocular microscopes/stereoscopes. When available, over-lapping aerial photographs were analyzed in 3-D stereo, whichgives the analyst a depth and topographic perspective of thelandscape. Specific photo-identified features were extracted firstas graphics markings and annotations on a mylar overlay tothe photographic print, along with the analysts’ written descrip-tion of the features and the surrounding landscape context. Inthe context of hazardous waste, photographic interpretation in-volves the identification of a distinct set of photographic observ-ables, whichare activities or conditions such as ground scars, ac-tive digging, or waste disposal, that are identified and extractedto meet the specific information needs of the investigation. Theformal definitions of the specific photographic observables andstandard operating procedures can be found in the report “PhotoInterpretation: Hazardous Waste Sites. Standard Operating Pro-cedure” [41].

For GIS analysis, all historical photographs were scanned ata minimum resolution of 1300 dpi and registered to the Mary-land State Plane Coordinate System for entry into a GIS data-base. Geo-registration was accomplished using ENVI ImageProcessing software (Version 3.4). Image-to-image registration

Fig. 5. GIS features extracted by aerial photographic analysis. An example offeatures derived from historical aerial photographic interpretation displayed asvectors on a geo-registered image in a GIS environment. Adapted from Kartmanand Slonecker [43].

using DOQQ and other derived historical orthoimagery datasetswas utilized with a second-order polynomial transformation.The overall registration accuracy for the 1918 image was 8.34m. All photo-interpreted features were then digitized as points,lines, and polygons in a GIS format and attributed with descrip-tions.

A. Reference Data From Remedial Activities

As reference points for the evaluation of historical photo-graphic information, a verification dataset was developed froma variety of actual hazardous waste cleanup activities that havetaken place over the years in Spring Valley. These include thefollowing activities/events.

1) UXO burial areas.2) Chemical waste burial areas.3) Properties with high soil arsenic concentrations, from site-

wide soil arsenic screening.4) Properties with detected levels of chemical warfare agents

or their breakdown products.5) Areas of relevant and confirmed contextual activities, such

as the mustard field testing area, derived from historicaldocument research.

6) Areas of unusual human health problems or other activitiesthat were reported and verified by the Area of Interest TaskForce, a multi-agency team established to investigate allpotential problems in Spring Valley.

These features, derived from ongoing remedial activities,provide documentation and defensible information aboutlandscape-level effects of AUES activities and establish anexcellent source of reference information for the evaluation ofhistorical photographic interpretation features. Features derivedfrom aerial photographic interpretation were overlain with thereference data in a GIS environment to determine the spatialcorrelation between photo-interpreted features and the futurerequirements for hazardous waste remediation. An example ofphoto-derived information displayed on a geo-registered imageof part of the American University area is shown in Fig. 5.


Fig. 6. The 42 potential areas of interest identified from photo interpretation of1918 aerial imagery.

Fig. 7. Potential areas of interest identified from photo interpretation of 1918aerial imagery overlain on 1991 image.

The purpose of the investigation of aerial photographs wasnot to provide a rigorous statistical analysis of photographic in-terpretation results, but rather to provide a practical measure,by simple percentage, of the relative value of the informationderived from the acquisition and analysis of historical aerialphotographs. While the analysis of historical photographs issomewhat subjective and very basic in terms of modern scien-tific analysis, information derived by this means has a long his-tory of practical value, and sometimes critical significance, tothe investigation and remediation of hazardous waste. In 1986,the EPA produced a photographic interpretation report, “Histor-ical Aerial Photographic Analysis, American University, Wash-ington, D.C.” [42]. The photo-analysis in that report was per-formed prior to any public knowledge of the nature or extentof the contamination issues in Spring Valley. It was also per-formed without benefit of the extensive collateral informationthat has resulted from 15 years of historical research in libraries,archives and historical holdings of the American University, the

Fig. 8. Enlargement of the northwest area of the 1918 aerial photograph ofSpring Valley and photo-interpretation features.

Fig. 9. Enlargement of the southwest area of the 1918 aerial photograph ofSpring Valley and photo-interpretation features.

U.S. Federal Government and the government of Washington,D.C.

The photo-analyst, using standard hazardous waste imageryanalysis techniques, identified 50 features on the 1918 aerialphotograph that were of potential relevance to hazardous wastedisposal and the current environmental cleanup activities. Fig. 6shows the photographic interpretation features overlaid on the1918 aerial photograph and Fig. 7 shows the same features over-laid on a 1991 image of the Spring Valley area. Figs. 8 and 9depict enlargements of the 1918 aerial image with the photo-graphic interpretation features. These are presented here for im-proved clarity of the landscape detail.

Eight of the photographic interpretation features identified inthe 1986 report were omitted from this analysis because theywere provided in the original report for locational context only.These features include: tents, barracks, support buildings, gen-eral buildings (B1–B10), the Weaver Farm House, Massachu-setts Avenue, and Nebraska Avenue. Of the remaining 42 fea-tures, each was investigated for spatial correlation to any poten-tial environmental contamination or cleanup issue. This was ac-


TABLE IIPHOTO-INTERPRETED FEATURES AND REMEDIAL ACTIVITY

complished by digitizing and geo-referencing the photographicinterpretation features into a vector GIS data layer. AnotherGIS data layer was created consisting of all known areas of

Fig. 10. WWI glass containers of chemical warfare agents and unexplodedmortar and artillery shells that were discovered in the disposal pit along Glen-brook Road. Photo interpretation played a critical role in the discovery and re-mediation of this major environmental threat. (Photos courtesy USACE).

past, current and future remediation activity. These consistedof: 1) all known areas of CW or UXO removal; 2) all knownparcels with arsenic levels exceeding the residential standard; 3)any parcels that specialty sampling results showing detectablelevels of other (nonarsenic) contaminants of concern; 4) areas ofknown contextual activities (areas of documented AUES activ-ities, such as animal toxicity testing) that have not been clearedby subsequent investigations; and 5) verified citizen reports ofunusual activities, events, or documented health problems.

V. RESULTS

Of the 42 features analyzed, 33, or over 78.5%, were spa-tially related to a past or present FUDS cleanup issue. Table IIlists the site-by-site analysis and the contamination issue relatedto each location, and Figure 25 in the table shows this graph-ically. Fourteen of the photo-identified areas of interest wererelated to properties of confirmed soil arsenic contamination.The arsenic contamination thresholds currently being utilizedin Spring Valley are 20 ppm for residential properties and 43ppm for all other properties [27].

Ten of the photo-identified features were related to actual ord-nance or specific areas of known ordnance testing. Three areaswere related to report human health problems and six areas cor-related with known activities of contextual importance, such as


Fig. 11. 1927 and 1991 aerial photographs with GIS overlays of trails featuresthat led to the discovery and remediation of a major hazardous waste disposalarea.

animal toxicity testing and chemical persistency testing. Onearea was related to known mercury and antimony contamina-tion. Ongoing photographic research and interpretation has alsoled to the discovery of two major contamination areas.

First, the location of a third ordnance burial pit, located onGlenbrook Road, was identified and mapped from historicalaerial and ground photographs. After much debate among thepartnering regulatory agencies (USACE, USEPA, DCDOH), anintrusive ground investigation was undertaken in the summer of2002 to determine if a third pit did actually exist. On the firsttest pit excavated, WWI-era mortar shells were discovered. Sub-sequent analysis revealed that many of these shells were filledwith arsine and other poison gas agents. Also, several glass con-tainers filled with CW agents were recovered. These items rep-resented a serious danger to human and ecological health andtheir remediation removes a potentially major environmentalrisk in an urban area.

Second, in 2004, interpretation of a newly discovered histor-ical aerial photo revealed a series of trails leading to a ravine inthe back potion of the American University property. Initial in-vestigations revealed the presence of very high arsenic as wellas laboratory glassware and equipment of the WWI era. Sub-sequent investigations have identified a large chemical labora-

Fig. 12. Buried Glass Laboratory Waste from AUES. A piece of glass tubingfrom one of the AUES chemistry laboratories that was discovered as part of aninvestigation triggered by the identification of a series of trails on a 1927 aerialphotograph.

tory disposal area that has yielded confirmed chemical weaponsagents and other dangerous contaminants.

Information from historical aerial photographs was directlyresponsible for the discovery and removal of these two seriousrisks to human and ecological health and these sites might stillbe undiscovered if it had not been for the utilization and inter-pretation of historical aerial photographs (see Figs. 10–12).

VI. SUMMARY

Although sometimes overlooked in a world of dynamic scien-tific instrumentation and emerging information processes fromsophisticated satellite sensors, historical aerial photographs rep-resent a rich and sometimes invaluable source of landscape-levelinformation related to hazardous waste disposal and remedia-tion or any issue where past activity on the landscape is relevant.

As pointed out by Philipson [44], the importance of photo-graphic interpretation can not and should not be downplayedor minimized because it is a fundamental information processin itself but, perhaps more importantly, it is the foundation ofmany sophisticated, digital, remote sensing analysis techniques.A cursory review of current literature reveals numerous ap-plications of photographic/imagery interpretation in providingcritical land use and land cover information that includes timeframes that predate satellite image availability [45]; verificationof computer model predictions of earthquake damage [46];and as a more reliable [46]; and as a more reliable method ofdeveloping specific landscape metrics in describing holisticaspects of complex landscapes [47].

In this application in Spring Valley, information derived fromthe interpretation of historical aerial photographs was directlyresponsible for the discovery and removal of several seriousrisks to human and ecological health and these sites might stillbe undiscovered if it had not been for the utilization and inter-pretation of this basic form of information.

However, there are some inherent drawbacks in utilizing thistype of historical analysis. One is limited to the historical datesof imagery available and the search and acquisition of this im-agery can be sometimes labor intensive. Also, the results willalways be somewhat subjective and subject to alternative inter-pretations, especially in legal settings.


However, used as a screening tool in the early stages of aremedial investigation, a 78.5% correlation with specific areasand conditions that will eventually require remedial action couldmake this type of information a very cost-effective tool in atime when resources for environmental cleanups are dwindling.A more detailed scientific investigation of the relationship be-tween information derived from historical imagery interpreta-tion and confirmed hazardous waste conditions across a statis-tically significant sample of hazardous waste sites, could reveala quantifiable level of information confidence related to humanand ecological risk. Additionally, it could facilitate a cost-ben-efit type of economic analysis that could document efficienciesand cost savings over other investigative techniques. Finally, astudy of the effectiveness of information provided by aerial pho-tographic interpretation could underscore the need to invest inarchival sciences and preserve invaluable sources of historicalphotographic information.

REFERENCES

[1] J. O. Nriagu, Arsenic in the Environment, Part I: Cycling and Charac-terization. New York: Wiley, 1994.

[2] “Electronic Code of Federal Regulations, CERCLA Hazardous Sub-stances,” GPO (U.S. Government Printing Office), 40 CFR § 261, 2010[Online]. Available: http://ecfr.gpoaccess.gov/

[3] H. Sigman, “Reforming Hazardous Waste Policy”. Essays in PublicPolicy The Hoover Institution [Online]. Available: http://www.hoover.org/publications/epp/2846086.html?show=essay

[4] J. A. Lybarger, R. Lee, D. P. Vogt, R. M. Perhac, Jr, R. F. Spengler, andD. R. Brown, “Medical costs and lost productivity from health condi-tions at volatile organic compound-contaminated Superfund sites,” En-viron. Res., vol. 79, no. 1, pp. 9–19, Jan. 1998.

[5] “Glossary for Chemists of Terms Used in Toxicology: Pure and Ap-plied Chemistry,” Int. Union of Pure and Applied Chemistry, 1993 [On-line]. Available: http://sis.nlm.nih.gov/enviro/glossarymain.html

[6] M. F. Vine, L. Stein, K. Weigle, J. Schroeder, D. Degnan, C. H. J. Tse,C. Hanchette, and L. Backer, “Effects on the immune system associatedwith living near a pesticide dump site,” Environ. Health Perspectives,vol. 108, no. 12, pp. 1113–1124, Dec. 2000.

[7] M. Berry and F. Bove, “Birth weight reduction associated with resi-dence near a hazardous waste landfill,” Environ. Health Perspectives,vol. 105, no. 8, pp. 856–861, Aug. 1997.

[8] H. I. Hall, W. E. Kaye, L. S. Gensburg, and E. G. Marshall, “Residentialproximity to hazardous waste sites and risk of end-stage renal disease,”J. Environ. Health, vol. 59, no. 2, pp. 17–21, Feb. 1996.

[9] N. J. Vianna and A. K. Polan, “Incidence of low birth weight amongLove Canal residents,” Science, vol. 226, no. 4679, pp. 1217–1219,Dec. 1984.

[10] M. Vrijheid, “Health effects of residence near hazardous waste landfillsites: A review of epidemiological literature,” Environ. Health Perspec-tives, vol. 108, no. Suppl 1, pp. 101–112, Mar. 2000.

[11] “What Does the Evidence Say About NPL Listing and Home Prices?,”U. S. Environmental Protection Agency, 2006 [Online]. Available:http://www.epa.gov/superfund/programs/recycle_old/property.htm

[12] C. C. Honigberg and M. Nolan, “Who Pays for the Stigma of Environ-mental Contamination?,” Commercial Investment Real Estate, 2000[Online]. Available: http://www.ciremagazine.com/article.php?ar-ticle_id=406

[13] C. C. de Nava, “World wide overview of hazardous wastes,” Toxicol.Ind. Health, vol. 12, no. 2, pp. 127–138, 1998.

[14] “Minimizing Hazardous Wastes: A Simplified Guide to the Basel Con-vention,” United Nations Environmental Programme, 2002 [Online].Available: http://www.basel.int/pub/simp-guide.pdf

[15] “Many Uncertainties Affect the Progress of the Spring Valley Cleanup.Report to the Subcommittee on the District of Columbia,” U. S. Gen-eral Accounting Office, Environmental Contamination: Committee onGovernment Reform, House of Representatives, 2002, GAO-02-556.

[16] “Chemical and Bacteriological (Biological) Weapons, and the Effectof Their Possible Use,” United Nations, New York, 1969.

[17] “Handbook on the Medical Aspects of NBC Defensive Operations,”HQDA, DASG-HCO, U.S. Army, 1996, Fm 8-9.

[18] C. E. Heller, “The Leavenworth Papers, Number 10. Chemical War-fare in World War I: The American Experience, 1917–1918,” CombatStudies Inst. U.S. Army Command and General Staff College, FortLeavenworth, KS, 1984.

[19] W. D. Bancroft, “Bancroft’s History of the Chemical Warfare Ser-vice in the United States. Research Division, Chemical Warfare Ser-vice, American University Experiment Station,” Res. Division, Chem-ical Warfare Service, American Univ. Experiment Station, Washington,DC, 1919.

[20] S. Titus, “Survey and analysis of present or potential environmentalimpact sites in Woburn, Massachusetts,” in Proc. 48th Annu. MeetingAmer. Soc. Photogrammetry, Bethesda, MD, 1982, pp. 538–549.

[21] J. G. Lyon, “The use of maps, aerial photographs and other remotesensing data for practical evaluation of hazardous waste sites,” Pho-togramm. Eng. Remote Sens., vol. 53, no. 5, pp. 515–519, May 1987.

[22] E. M. Barnaba, W. R. Phillipson, A. W. Ingram, and J. Pim, “The useof aerial photographs in county inventories of waste disposal sites,”Photogramm. Eng. Remote Sens., vol. 57, no. 10, pp. 1289–1296, Oct.1991.

[23] M. J. Benger, “Fact Sheet: The EPA Remote Sensing Archive The U.S. Environmental Protection Agency, Environmental Photographic In-terpretation Center (EPIC),”, Las Vegas, NV, 2004.

[24] R. G. Congalton and R. A. Mead, “A quantitative method to test forconsistency and correctness in photointerpretation,” Photogramm. Eng.Remote Sens., vol. 49, no. 1, pp. 69–74, Jan. 1994.

[25] A. A. Fries and C. J. West, Chemical Warfare. New York: McGraw-Hill, 1921.

[26] “Final Remedial Investigation Report for The Operation Safe Re-moval Formerly Used Defense Site,” U.S. Army Corps of Engineers,Huntsville Division, Baltimore District, Parsons Engineering Science,Inc., Washington, D.C., 1995.

[27] “Engineering Evaluation/Cost Analysis for Arsenic in Soil SpringValley Operable Units 4 And 5,” Parsons Engineering Science, Inc.,Washington, DC, 2003, vol. 3, Contract Daha90-94-D-0010, TaskOrder DA01 DERP-FUDS HTRW Project No. C03DC091802.

[28] C. Babington-Smith, Air Spy: The Story of Photo Intelligence in WorldWar II. New York: Harpers, 1957.

[29] D. S. Simonett, R. G. Reeves, J. E. Estes, S. E. Bertke, and C. T. Sailer,“The development and principals of remote sensing,” in The Manualof Remote Sensing, R. N. Colwell, Ed., Second ed. Bethesda, MD:Amer. Soc. Photogrammetry and Remote Sensing, 1983.

[30] R. N. Colwell, “History and place of photographic interpretation,”in The Manual of Photographic Interpretation, W. Phillipson, Ed.,2nd ed. Bethesda, MD: Amer. Soc. Photogrammetry and RemoteSensing, 1997, ch. 1, pp. 3–47.

[31] E. T. Slonecker, M. J. Lacerte, and D. Garofalo, “The value of historicimagery,” Earth Observation Mag., vol. 8, no. 6, pp. 39–41, 2002.

[32] S. Titus, “Survey and analysis of present or potential environmentalimpact sites in Woburn, Massachusetts,” in Proc. 48th Annu. MeetingAmer. Soc. Photogrammetry, Bethesda, MD, 1982, pp. 538–549.

[33] T. L. Erb, W. R. Phillipson, W. L. Teng, and T. Liang, “Analysis oflandfills with historical airphotos,” Photogramm. Eng. Remote Sens.,vol. 47, no. 9, pp. 1361–1369, Sep. 1981.

[34] D. Garofalo and F. Wobber, “Solid waste and remote sensing,” Pho-togramm. Eng., vol. 40, no. 1, pp. 45–59, Jan. 1974.

[35] J. G. Lyon, “The use of aerial photographs and remote sensing data inthe management of hazardous waste sites,” in Hazardous Waste Man-agement for the 80’s, T. Sweeny, H. Bhatt, R. Sykes, and O. Sproul,Eds. Ann Arbor, MI: Ann Arbor Sci., 1982, pp. 163–171.

[36] J. G. Lyon, “The use of maps, aerial photographs and other remotesensing data for practical evaluation of hazardous waste sites,” Pho-togramm. Eng. Remote Sens., vol. 53, no. 5, pp. 515–519, May 1987.

[37] A. B. Nelson, L. A. Hartshorn, and R. A. Young, “A Methodology toInventory, Classify and Prioritize Uncontrolled Waste Disposal Sites,”U.S. Environmental Protection Agency, Las Vegas, NV, 1983, EPA-600/4-83-050.

[38] W. R. Phillipson, E. M. Barnaba, A. Ingram, and V. L. Williams, “Landcover monitoring with SPOT for landfill investigations,” Photogramm.Eng. Remote Sens., vol. 54, no. 2, pp. 223–228, Feb. 1988.

[39] E. M. Barnaba, W. R. Phillipson, A. W. Ingram, and J. Pim, “County-wide inventory of waste disposal sites: A case study of historic airphotoanalysis,” in Proc 55th Ann. Meeting Amer. Soc. Photogrammetry andRemote Sens., Baltimore,, MD, 1989, vol. 1, pp. 167–174.

[40] E. M. Barnaba, W. R. Phillipson, A. W. Ingram, and J. Pim, “The useof aerial photographs in county inventories of waste disposal sites,”Photogramm. Eng. Remote Sens., vol. 57, no. 10, pp. 1289–1296, Oct.1991.


[41] Photo Interpretation: Hazardous Waste Sites. Standard Operating Pro-cedure. U.S. Environmental Protection Agency. Las Vegas, NV: U.S.Environmental Protection Agency/ Environmental Photographic Inter-pretation Center, 2005.

[42] K. K. Stout, “Historical Aerial Photographic Analysis, American Uni-versity, Washington, D.C.,” U S. Environmental Protection Agency,Environmental Sciences Division, Las Vegas, NV, 1986, vol. 1 and 2,TS-PIC-86001M.

[43] A. S. Kartman and E. T. Slonecker, “Historical Aerial PhotographicAnalysis, Spring Valley Site, Washington, D.C.,” U.S. EnvironmentalProtection Agency, Environmental Sciences Division, Las Vegas, NV,2001, vol. 1 and 2, TS-PIC-20103439S.

[44] Manual of Photographic Interpretation, W. R. Philipson, Ed.Bethesda, MD: Amer. Soc. Photogrammetry and Remote Sens., 1997.

[45] J. Imbernon, “Pattern and development of land-use changes in theKenyan highlands since the 1950s,” Agriculture, Ecosyst. Environ-ment, vol. 76, no. 1, pp. 67–73, Oct. 1999.

[46] S. Marzorati, L. Luzia, V. Petrini, F. Meronia, and F. Pergalani, “Detec-tion of local site effects through the estimation of building damages,”Soil Dynam. Earthquake Eng., vol. 23, no. 6, pp. 497–511, Aug. 2003.

[47] M. Antrop and V. V. Eetvelde, “Holistic aspects of suburban land-scapes: Visual image interpretation and landscape metrics,” LandscapeUrban Planning, vol. 50, no. 1–3, pp. 43–58, Aug. 2000.

E. Terrence Slonecker received the Ph.D. degree in environmental science andpublic policy from George Mason University, Fairfax, VA.

He is a Research Geographer with the United States Geological Survey’sEastern Geographic Science Center, Reston, VA. He has over 30 years of expe-rience in remote sensing, including positions with the U.S. Air Force, the Intelli-gence Community, private industry, and the Environmental Protection Agency.His current research interests include hyperspectral analysis of heavy metals,hazardous substances, and related vegetation stress. He has served as an expertwitness for the Federal Government on remote-sensing-related matters on sev-eral occasions.

Date post:	23-Sep-2016
Category:	Documents
Upload:	et
View:	212 times
Download:	0 times

The Use of Historical Imagery in the Remediation of an Urban Hazardous Waste Site

Documents