OEFS Abstracts - Fall 2004

Jim Cronan – Forest succession stages after wild fires in Alaska

Wildland fires are a prominent natural process in the boreal forests of Alaska and the dominant force of disturbance. As the population of Alaska expands at a higher rate than most other states in the nation wildland fires pose a high risk to property. As forests mature, vegetation composition and distribution changes, transitions that have a substantial influence on fire behavior. In general it is believed that fire behavior in the black spruce (Picea mariana) forests of interior Alaska increases with age during the first 40-60 years of development after which the relationship between fire behavior and stand age is unknown. Thus stand age, at least in the first half century of forest growth can serve as a proxy for fire hazard posed by forests and can contribute to understanding fire dangers to rural and remote landowners, tactical decisions in wildland firefighting operations, and in delineating fire management options used by the Alaska Interagency Wildland Fire Coordinating Group to define levels of fire suppression statewide. There are two areas of knowledge that can be expanded on to meet these fire management needs in Alaska. The first is to gain a better understanding of the relationships between seral stages in black spruce forests and potential fire behavior, and the second is to map the spatial extent of these different seral stages. The field of remote sensing is well suited to increasing the level of knowledge for the second point.

The goal of this project is to use remote sensing to determine if seral stages following wildland fire in the black spruce forests of interior Alaska are identifiable through satellite imagery and if so, if the temporal windows for each seral stage can be delineated. Ancillary goals include the evaluation of various classification methods and a comparison of alternate sources of burned acreage data for Alaska to the results of this project.

Andrew Niccolai – Evaluation of forest cover estimation approaches derived from remotely sensed imagery for the Dewey Wills Wildlife Management Area, Louisiana

Forest ecosystem dynamics modeling, geographical information systems, and remote sensing data analysis form a powerful network of decision support tools in natural resource management. Remote sensing presents an excellent opportunity for providing data such as forest cover information at appropriate spatial and temporal scales needed to run forest models like the Landscape Management System (LMS). The optimal exploitation of decision support systems for forest resource management lies in the availability and accuracy of information pertaining to forest attributes on a landscape. This project explores the accuracy of several methods to estimate forest cover types on the Dewey Wills Wildlife Management Area (DWWMA) in central Louisiana. Several combinations including multi-temporal, multi-spectral, as well as different vegetative indices will be tested in order to produce the best possible map of forest cover types. Remote sensing platforms including Landsat TM, ETM+, and ASTER images will be used to create these different temporal, spectral, and vegetative indices combinations. Forest cover type data previously collected on 4,983 hectares (12,313 acres) was used to initially train a series of supervised classifications and then later to test the classification accuracy. Comparison of the accuracies amongst the various remote sensing combinations will be used to determine the method to recommend for future consideration in forest cover classification of bottomland hardwood ecosystems.

Rafael Bernardi – Assessing the current situation and evolution of fragments of coastal sand forests in Uruguay

The main single land cover in Uruguay is grasslands (14.3 million ha), roughly 82% of the total surface of the country. Grazing lands with sheep or cattle are fifty-two percent of the total area of the country. Extensive patterns of land use (like grazing) and rates of extraction of natural resources (native forests were heavily logged for fuel in the first half of the century) are behind a massive reduction of natural habitat and native forests that has occurred since cattle were introduced centuries ago, leaving around 3.5% of native forests remaining, which are under constant threats of new economic activities. Initial estimates for forests were between 6% and 25%, although there are not conclusive studies about past forest covers. Therefore, a high-resolution monitoring system needs to be implemented to assure proper protection and restoration of these forests.

The specific objective of this project is to assess the current situation and evolution of fragments of coastal sand forest in Uruguay. The general objective will be to produce and test a classification for Uruguay’s native forests types.

Dan Braden – Quantifying forest cover change in the western area forest of Sierra Leone

Jutting into the Atlantic Ocean from the coast of Sierra Leone, the Freetown peninsula contains not only the capital city of one of the world’s least developed nations, but also one of the last stretches of coastal montane forest in West Africa, the Western Area Forest. As the only intact rainforest in Western Sierra Leone, home to over 300 species of birds as well as one of the last chimpanzee habitats left in Sierra Leone, the Western Area Forest has enormous biotic importance. It also provides and protects the water supply for Freetown and its 1.5 million residents (via the Guma Valley Reservoir) and provides filtration for the world’s third-largest natural harbor, the surrounding area of which has been included on the Ramsar List of Wetlands of International Importance (2004). The forest was declared a reserve in 1913 and gazetted as a protected area (the Western Area Forest Reserve) by the Sierra Leonean government in 1975.

Little if any quantitative analysis has been done to study land cover change on the peninsula and its forest. The Ministry of the Environment of Sierra Leone is currently developing a multi-use resource management plan for the forest reserve, and needs such information. This purpose of this study is to fill the information gap by quantifying forest change on the Freetown peninsula, and explore its potential causes.

Luisa Fernanda Lema Vélez – Assessing mangrove restoration in the Ciénaga Grande De Santa Marta

The reduction of water flow from the ocean and the Magdalena river towards the system of lagoons of Ciénaga Grande de Santa Marta, Colombia, resulted in the alteration of its regular hydrodynamic and in large areas of dead mangroves. A project of restoration of the ecosystem started in 1992, running up to 1998. I evaluated the evolution of the mangrove ecosystem from 1989 to 2000 using several techniques for remote sensing imaging classification. My results show that mangrove loss kept advancing during that period.

Vincent Medjibe – Examining changes in the Dzanga-Sangha region of the Central African Republic from 1986 to 2001

Remotely sensed data have been used in many studies in order to detect the changes in land cover and land use, to detect floristic association in mixed tropical forest, and also to detect forest loss in protected areas. The detection of land cover and land use changes is an important aspect for the management of protected areas and forest reserves. Changes due to logging activities create socio-economical problems in a particular area assigned for wildlife conservation. Studies have shown that selective logging has an important effect on the change in forest structure, species composition, and change in microenvironment in the Central African forest. Also roads built within the forest for logging has an impact on floristic and wildlife diversity in the Dzanga-Sangha Reserve.

The objective of this project was to detect, classify, and examine the change in land cover/land use in this region. Data from of Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) (thirty meters spatial resolution) were used for this project. Classification technique (supervised and unsupervised) was performed to classify land cover. Band combination was composed in order to detect the changes as well as the comparative Normalized Difference Vegetation Index (NDVI). The results of this project can be applied in the Dzanga-Sangha area in order to evaluate the effect of human pressure land cover/land use changes and assess the nature of future changes in the areas.

Rolando Mendez-Treneman – Development of a remote sensing procedure to identify wetlands

On a world scale, wetlands occupy about 4% of the Earth’s ice-free land surface. As of 1997, the wetland acreage across the conterminous United States was 105.5 million acres and during the period of 1986 to 1997, the annual loss of wetland habitat is reported by Dahl at 58,500 acres. Ninety-five percent (95%) of the remaining wetland resource is of the freshwater type and within this freshwater group, forested wetlands form the largest single category covering 50.7 million acres. Dahl goes on to report that about 1.2 million acres of forested wetlands were lost during the 1986 to 1997 period. The causes for wetland habitat loss have included urban / rural development (51%), agriculture (26%), and silviculture (23%). Forest wetlands contribute to moderation of downstream flooding, maintenance of water quality, diverse wildlife habitat, and pollution control.

Management activities, e.g. forestry operations, and land development can have significant effects on wetland condition and effectiveness. Law, policy, and management practices are designed to protect and or enhance this resource. A critical datum for effective resource management is the location and extent of wetlands. The objective of this project is, through the use of remote sensing data, to determine the presence and distribution of wetlands (open water and vegetated) in northern Connecticut.

Laura Cuoco – Studying the expansion of shrimp farms in the Chone River estuary of Ecuador

The coastline of Ecuador has changed dramatically over the past thirty years. This is evidenced in the alarming rate of mangrove deforestation primarily due to the construction of shrimp farms. One of the underlying factors contributing to the massive destruction of mangrove forests was the 1980s shrimp ‘boom’. In this decade, production of shrimp rose by 600 percent with the United States, Japan and several European countries being the largest consumers. By 1999, total deforestation rates had reached over fifty percent nationwide and close to ninety percent in the Chone River estuary in the Province of Manabi.

Laws were passed in the 1970s to protect mangroves against the ever encroaching shrimp farms. In addition, several governing bodies have been formed and responsibilities delegated to further their protection. However, enforcement has not occurred and mangroves continue to be destroyed. Most of the data documenting shrimp pond and mangrove acreage is provided by CLIRSEN, the Ecuadoran clearinghouse on remote sensing of natural resources. Thus, the purpose of my project was to quantify shrimp pond expansion and mangrove deforestation between 1986 and 2001 as a way to assess CLIRSEN’s data.

Melissa Andersen – Land use change in southeastern Iraq: An analysis of pre- and post-Gulf War marshlands

Mesopotamia, the Garden of Eden, Babylon–terms from a region identified as the cradle of civilization, home to some of the most ancient cultures ever discovered, today describe a destroyed landscape, ruined by a megalomaniac ruler of a war-torn nation. The marshlands of Mesopotamia (southeastern Iraq), once the most expansive wetland system in the entire Middle East are today a desiccated, salt-crusted landscape, the result of both damming and diversion projects upstream meant to improve water storage, hydro-electric programs and agriculture, and also of a vicious campaign designed to systematically destroy the marshlands and drive out its inhabitants in order to suppress civil unrest.

Though much research has been completed on the loss of marshlands in Iraq, and efforts to restore the marshes are already in process, I was interested to investigate the changes on the landscape myself, using available satellite images. I wanted to know if I could find a way to see and describe the changes, both qualitatively and quantitatively, and determine which method of change detection was most helpful for the features in which I had the most interest (marsh vegetation, water, and salt deposits). Understanding the methods that are most useful for observing land cover change in this region will help to understand where the most serious problems on the landscape are, and where resources for restoration should focus. The difficulty in accessing these regions as of late, both under the Hussein regime, and now the current unstable situation in Iraq, make the use of remote satellite imagery all the more valuable.

Rebecca Barnes – Nitrogen export from the Quinnipiac watershed

Coastal waters often receive high nutrient loads and thus suffer from eutrophication, the increased supply of organic matter to the ecosystem resulting in significant changes in ecosystem structure and function. This can lead to increased production in algal biomass, nuisance and harmful algal blooms, decreased dissolved oxygen levels (hypoxia and anoxia), decreased biodiversity, and loss of habitat such as sea grass beds in coastal systems. Increased NO₃ loads, the most bioavailable form of nitrogen that is typically the limiting nutrient in coastal waters, often leads to estuarine eutrophication.

Long Island Sound’s western basin is usually hypoxic from mid-July through September, where there are higher nitrogen loads and less mixing of oxygenated waters. According to the Long Island Sound Study, more than 60% of nitrogen entering the Sound comes from sewage treatment plants; while non-point sources, including atmospheric deposition, fertilizers, animal waste, and soil microbial nitrification contribute the remaining 40%.

In order to effectively manage this problem it is necessary to know the relative importance of various non point sources. The land cover and land use of a watershed determines the type and amount of nitrogen non-point source pollution exported to streams. Remote sensing techniques are needed to accurately assess and classify these watersheds into land cover classes that satisfy two criteria. First and foremost the land cover classes must reflect differences in nitrogen transport. Secondly, the land cover classes must be coarse enough to provide a realistic framework for management. In this project I will classify the land cover of the Quinnipiac River watershed at 3 spatial scales and use a variety of methods and images.

Linda Kramme – Evaluating landcover and landuse changes near Gunung Palung National Park, West Kalimantan, Indonesia

In this study, remote sensing image interpretation was used to evaluate land cover and land use changes over a twelve-year period in an area of West Kalimantan, Indonesia, in southwestern Borneo. Portions of two Landsat images taken in June 1989 and June 2001 were evaluated. The following techniques were applied:

Supervised classification for the two images, followed by qualitative and quantitative comparison of these classifications.
Unsupervised classification for the two images, followed by qualitative comparison of these classifications.
Comparison of 2001 supervised and unsupervised classifications.
Comparative NDVI analysis to evaluate vegetation change over time.

The goal of this study was to develop a better understanding of the benefits and limitations of remote sensing as a tool to evaluate land cover and land use change in this region that is incredibly biologically diverse but extremely threatened.

Kathryn Dana – Irrigation and tourism impacts on the Segura River region of Spain

The Tajo-Segura transfer canal was completed in 1979 in response to water shortages in southern Spain. Since then, water demand has increased significantly, and existing transfer systems are proving insufficient to address the problem. The Tajo-Segura transfer system is one for the most long-lived and successful canals in operation today, and has served as a model for numerous irrigation proposals. For this reason, I set out to evaluate the changes that have come about as a result of the canal’s presence.

Irrigation methods have been implemented in the Iberian Peninsula since the time of Roman occupation. For two thousand years the scope and scale of agriculture has increased, thanks to the introduction of new crops from the Americas and exportation of goods. However, today tourism has replaced agriculture as the largest contributor to Spain’s GNP; this has had serious consequences on water use. As population steadily increases and tourism continues to be on the rise, water demand has followed the trend.

Wei-Chien Lai – Investigating the impact of economic development on the coastal environment of Taiwan

Taiwan’s southwest coast used to be home for coral reefs and migrating birds before industry and economic development. As the rapid developments coming along since mid twentieth century, mountainous areas have suffered from landslides and floods and coastal environment has confronted unprecedented degradation.

The coastal environmental issues are drawing more and more attention of Taiwan’s public. Among the cities, there are clean water acts, clean beach calls, and protests against the mountain cableways construction. It is exciting to see the public’s awareness is rising up. However, if we take advantage of the satellite imagery to assist the analysis of landscape management and coastal conservation mission, we can urge a better, more efficient decision-making and convince the public the importance of environmental conservation by demonstrating the breathtaking images taken from the space.

Brandon Berkeley – Exploring seasonal and long-term changes in vegetation cover in the Sahel region of western Africa

Miriam Clinton – An investigation of the relationship between climate change and patterns on Native American reservations in South Dakota

Landscape studies are a growing field in the social sciences. Landscape archaeology, in particular, has only recently come into its own as a separate archaeological discipline. One of the principle debates in all of the social studies, especially landscape archaeology, is the question of environmental determinism. How much was culture created by environment, or, conversely, how much did humans change the environment to fit their lifestyle?

Many theorists, such as Keith Basso, turn to ethnographic analysis to understand “primitive” cultures and their relationship with the earth. According to such theorists and to popular belief, the indigenous or non-Western cultures of the world were more “in-tune” with the earth and related to it in ways that are rapidly being lost as peoples westernize. However, such archaeologists also criticize theories that attempt to relate the earth’s changes to changes in cultures, arguing that such theories are overly deterministic and do not give fair scope to the cultural strength of indigenous peoples. I hope to address this conflict through my examination of settlement patterns and their relation to climate change in the Lakota culture.

Victoria Thompson – Urban sprawl in Austin, TX between 1987 and 2001

The focus of my studies at the Forestry School is on climate change policy in the United States, but I am also interested by extension in the land-use patterns of American cities. Most growth of US cities in the last 60 years has been car-dependant “sprawl” (the more people drive, the more greenhouse gases are emitted). Sprawl also has many negative local environmental impacts. It increases impervious surface area, increasing runoff and erosion, and decreasing recharge of aquifers (Austin sits over an important recharge area for the Edwards Aquifer). Sprawl eats up farmland and woodland. While this may often be beneficial to certain animal species that co-exist easily with humans (e.g. raccoons), it often has the effect of diminishing the numbers of many native and/or endangered species of plants and animals. Increased car travel from far-flung locations also increases air pollution and traffic congestion, in addition to increased emissions of greenhouse gases.

Austin is the capital of Texas, the seat of Travis County, and a burgeoning center of industry and culture in the state. Like many cities in the American South and West, it has experienced rapid growth over the past several decades. I chose Austin because of its demographic qualities and my association with the city. My aim in this project was to determine, quantitatively and qualitatively, the extent of urban sprawl in the Austin metropolitan area during the 1990s.

Jocelyn Hittle – Investigating landcover change in the Seattle metropolitan region between 1990 and 2000

In the United States, urban land consumption is occurring at about twice the rate of population growth. The rapid development of land that was previously forest, farmland, prairie or wetland has created a serious sprawl problem. Natural areas are paved, open space is lost, and cities face increasing costs of installing infrastructure, such as roads and sewers, as development sprawls ever outward.

The Seattle Metropolitan area has faced considerable development pressure as the population has increased in recent decades. From 1990 to 2000, the population increased by over half a million people (20%). Corresponding pressure on surrounding areas to develop has led to conversion of forested land and farmland to housing developments.

The purpose of this study is to attempt to quantify these changes in land use in the Seattle area from 1990 to 2000. Previous work has mapped changes in population density in Seattle, based on U.S. Census data, using ArcView 3.3 Geographic Information Systems (GIS) technology. This study analyses Landsat TM and ETM+ images from 1990 and 2000, respectively, to examine changes in land cover, and compares these results to the GIS population density data. The patterns of land use change are discussed, as well as the accuracy of remote sensing when compared to GIS data in determining population density patterns.

Stephanie Horn – Landcover change detection of Huntsville, AL between 1991 and 2001

Huntsville (also known as “The Rocket City”), AL is one of the fastest growing cities in the state of Alabama. This city was founded by pioneer John Hunt in the year of 1805. By 1819, Huntsville had grown tremendously and was the largest town in the Alabama Territory. During the 1840’s and 50’s, Huntsville was the cotton trading center of the Tennessee Valley when planters and merchants originally from Virginia and the Carolinas built impressive town homes. Huntsville was also Alabama’s first English-speaking city. Of the 180,000 city residents, more than 10 percent are natives of other countries and more than 100 languages and dialects are spoken there.

For over a decade (1991-2001) now, this city has experienced a dramatic landscape change due to the increasing population, new development of commercial zones, and different agricultural land uses. During this decade, the suburban areas, as well as, the urban areas have both increased in population due to new jobs, better school systems, and living preferences. The agricultural land use changes annually depending on the crop that’s being planted. To detect the increase or decrease of the different landcover classes of Huntsville, I’m going to use satellite imagery and some image processes such as supervised and unsupervised classification, area summary reports, and NDVI comparison.

Iona Hawken – Urban Jungle: Urbanization, land-cover change, and urban-rural dynamics in a development frontier of Amazônia

Two myths dominate both research and discourse on Amazônia in general: one, that the Amazon is dominated by “wild” and “pristine” nature untouched by humans; and two, that the only real human populations that belong to the Amazon are indigenous people. In much research on Amazônia, even in those studies that try to combine goals of both conservation and development, humans are ignored in studies of nature in the Amazon. Cities do not easily fit into an image of the Amazon.

Amazon cities are smaller, and less distinct from the surrounding environment due to their heavier reliance on local materials for construction, and less developed infrastructure (roads, buildings). There are thus challenges in the application of remote sensing technologies to the detection and differentiation of frontier rainforest cities from other deforested areas due to ranching, logging, and agriculture in the Amazon.

Using Landsat TM and ETM+ datasets, my project goals are to quantify and compare spatial and temporal urban growth and land-use change around the Amazon city of Iquitos, in Peru. To do this, I proposed to quantitatively, qualitatively, and semi-quantitatively measure urbanization patterns over time and space in an Amazonian city, detect changes in land use in the hinterlands connected to the city, and compare and evaluate methods for detecting urban growth and land use change in a rainforest city.

My research objectives for this project were to:

Compare spatial and temporal urbanization and land-use change over time;
Detect land-cover change in hinterlands of an urban area to attempt to quantify the effects of urbanization on connected rural areas;
Compare and evaluate different methods of detecting urbanization by applying different urban detection and classification methods to an Amazonian city.

Katherine Hansen – Habitat Characterization of the Lone Star Tick Amblyomma americanum

Ticks are known to be important vectors of several human diseases. The lone star tick, Amblyomma americanum (L.), is known to transmit three species of bacteria that can cause diseases in humans, including: Ehrlichia chaffeensis, Ehrlichia ewingii, and Borrelia lonestari. The most common and most widespread of these, E. chaffeensis, is known to cause human monocytic ehrlichiosis (HME). A. americanum is known to be the most aggressive and have the most non-discriminatory biting habits of the 40 species of North American ticks. This behavior, combined with A. americanum’s ability to transmit various species of bacteria, makes this tick an important vector for disease in the United States.

Review of the literature and recent studies have shown that the traditional range of A. americanum is expanding into regions that it has not been previously recognized. This expansion is correlated with a rise in HME cases across the eastern half of the United States. This expansion has largely been attributed to the population increase of the white-tailed deer (WTD) Odocoileus virginianus, because WTD have been identified as the principle host for A. americanum during all three of its developmental stages. The potential spread of A. americanum may be limited by the density and distribution of the white-tailed deer.

A. americanum’s principal habitat is deciduous forest with dense undergrowth. Because of this habitat preference, I am hypothesizing that A. americanum’s presence can be identified using vegetation indices obtained by satellite imagery. To test my hypothesis, I will obtain images that contain two different vegetation indices, including the normalized difference vegetation index (NDVI) and the enhanced vegetation index (EVI). Using statistics, I will determine which vegetation index is more significantly associated with the presence of A. americanum. I will then use the more significant index to determine the vegetation index ranges that are most significantly associated with the presence of A. americanum. The significant ranges will then be used to classify the vegetation index in order to create a model that will show the most appropriate habitat areas for A. americanum. This model will be able to identify the current range of A. americanum, as well as, identifying suitable areas of habitat that A. americanum may be able to expand to in the future.

Erika Schielke – Predicting locust plagues in western Australia using NDVI

Locust plagues can cause extensive damage to crops in Western Australia. The dependence of the locust life cycle on rainfall provides an opportunity to use climate data to predict the location of outbreaks, allowing for increased monitoring and early control efforts. I tested whether NDVI anomalies could be used to predict the location of locust outbreaks. This approach assumes that NDVI is a proxy for rainfall. Because fall rains are linked to outbreaks of locusts during the following spring, I used 16-day composite MODIS NDVI images to compare April NDVI values from 2004 to average NDVI values, and tested whether areas of increased NDVI in 2004 corresponded to locations of locust outbreaks between September and November of 2004. The technique had little predictive power, which is most likely explained by a complex relationship between NDVI and rainfall in Western Australia. A more thorough understanding of the relationship between NDVI and rainfall in Western Australia is necessary before NDVI anomalies could successfully be used to predict locust outbreaks in this region.

Laura Evans – Heavy industry decline and the return of vegetation to Allegheny

Allegheny County lies atop one of the world’s largest bituminous coal deposits and is home to the city of Pittsburgh and-at one time-the nation’s steel industry. At its center is the conjunction of three major rivers: the Ohio, the Allegheny and the Monongahela – all three of which are easily navigated by barge. Pennsylvania’s coal is higher in sulfur content than most other coals, which adds sulfur dioxide to the atmosphere. Sulfur dioxide not only affected the general visibility in the region and the buildings of Allegheny County (many of the buildings’ facades still remain black, but vegetation as well. Sulfur dioxide is not only bad for a plant’s “breathing,” but also dissuades growth when it falls as acid raid-penetrating the soil. When sulfur dioxide penetrates the soil, it makes it difficult for seeds to take hold and for tree roots to grow. This is probably why, in photos or images of the hillsides in Allegheny County where there tends to be less soil, you see fewer, if any, trees than you would today.

I hope that by comparing two satellite images I will be able to see a change in the amount of urban areas and vegetation in Allegheny County. Ideally, there would be a decrease in the urban regions and an increase in vegetation or perhaps suburban areas. For the purposes of this study, I distinguish urban municipalities as towns, boroughs or cities with heavy industry (steel mills, coke plants or heavy industry of some sort)The urban municipalities of Allegheny County show that from 1970-2000 there is a decrease in population. Suburban municipalities were areas without heavy industry. They could have areas that appear to be heavy industry on a satellite image, but these were actually strip malls, airports, or shopping centers. I intended to classify the image based on these distinctions and then compare the statistics from the classification to see if there was noticeable change in the numbers between 1976 and 1999.

Andrea Johnson – Monitoring gas and oil pipelines with remote sensing in a tropical forest

The vast Camisea gasfields lie in the Lower Urubamba River Valley of the department of Cuzco, in southeastern Peru. In 2000 the Peruvian government granted licenses for exploration and exploitation to two multinational consortiums: Argentina-based Pluspetrol (responsible for the extraction and processing) and a group called “TGP”, which contracted to build and operate the two buried pipelines that carry natural gas and gas liquids from the Urubamba (714 in total length). The Camisea project has been controversial due to its delicate location and its grand scale. The pipeline begins in pristine rainforest within the Nahua Kugapakori State Reserve for indigenous peoples, traverses the Andes, and emerges at a bay in the buffer zone of the world-renowned Paracas Marine Reserve.

There is significant interest in establishing a consistent and independent system for monitoring the operations and impacts of the Camisea pipeline over the 40-year duration of TGP’s concession and beyond. Several of the most serious potential consequences of a pipeline built in tropical rainforest - erosion, colonization, and deforestation - are, in theory, amenable to monitoring through the use of satellite imaging technologies. Moreover, the company that built and operates the pipeline is required - through its own EIA, as well as Peruvian government regulations and loan conditions imposed by the Inter-American Development Bank - to comply with a number of measures to minimize, mitigate, and rehabilitate any negative environmental impacts. In this project I investigate the potential for using Quickbird satellite imagery to monitor environmental impacts and operational compliance along a buried pipeline passing through tropical rainforest.

Russell Schimmer – Identifying copper mine tailings using remote sensing

The purpose of this project was to assess whether Cu mine production tailings have a specific electromagnetic reflectance spectral, or range of spectral signatures, that can be identified utilizing satellite imagery and remote sensing technology. The study site of the Phelps Dodge Bagdad Cu Mine in Bagdad, Arizona was chosen due to its relatively intense mining and production of Cu concentrate-processing approximately 85 thousand tons of porphyry Cu per twenty-four hour cycle-the bulk of the waste which is disposed of as Cu tailings. Since 1997, the on-site Lower Mammoth tailings pond has been utilized for this deposition. A June 13, 2000 Landsat 7 ETM+ (Enhanced Thematic Mapper Plus) image was chosen for the project in order to utilize the spectral bands at the .45-.515 microns and 2.09-2.35 microns wavelengths. These two bands covered two specific spectral signatures distinctive to Cu bearing minerals: a sharp reflectance peak at ~.48-.50 microns visible range and a sharp absorbent dip at ~ 2.35-2.40 microns.

Neither the reflectance peaks nor valleys register high on the reflectance intensity scale (~.15 and .05 respectively; the units are based on an arbitrary scale of 0 to 1 designating 100 percent absorbance and reflectance respectively) . By applying thresholds on these two bands to differentiate them from the surrounding surface features and topography, the intent was to isolate these two spectral signatures. Although the eventual process of image manipulation and application using Landsat 7 ETM+ spectral bands 1, 4, and 6 did successfully distinguish areas specific to the locations of known low percentage H₂O saturated to dry Cu tailings, it is yet unclear whether these findings are the direct result of distinct Cu spectral signatures or of a specific mixture of other constituents and influences inherent to the Cu tailings’ matrix.

Yukiko Tonoike – Exploring how snow cover and vegetation nomadic populations in NW Iran

This paper stems from my dissertation project on Dalma ceramic assemblages from western Iran dating to the 5th millennium BC. In order to better understand and evaluate some of the theories associated with this ceramic assemblage, I will examine and consider the question of seasonal migration in the area. Using ethnographic data and remote sensing, I will examine the phenomenon of seasonal, vertical transhumance.

The questions that will be addressed are:

why is seasonal migration desirable or necessary
what are the possible routes of migration
when do seasonal migrations take place and
what factors affect the timing of seasonal migration.

Center for Earth Observation

...supporting the remote sensing needs of the University