Reproduced, with permission, from:
Reproduced, with permission, from:
Frank G.Sadowski, F. G., N. Roller, and J. E. Colwell. 1993. The benefits of the Landsat Pathfinder Program for social science needs in global change research. In Proceedings of the 25th International Symposium on Remote Sensing and Global Environmental Change, 285-93. Vol. 2. Ann Arbor, MI: Environmental Research Institute of Michigan.
Consortium for International Earth Science Information Network (CIESIN)
ABSTRACT
Successful implementation of studies addressing the human dimensions of global change will pose major conceptual and methodological challenges for the social sciences. The Landsat Pathfinder Program will offer significant benefits for meeting the challenges facing social scientists studying global change. The benefits include: (1) assisting the development and implementation of a global perspective for research; (2) aiding the coordination of interdisciplinary research; and (3) aiding the longitudinal studies of global environmental change.
INTRODUCTION
There is increasing awareness that the collective actions of humanity are capable of affecting the Earth system and its processes. Humanity, in turn, is capable of being affected by changes in the Earth system induced by these effects. The nature of humanity's actions at different levels in society (local, national, and international) and in the fabric of society (for example, its social, economic, and political structures and institutions) comprise what are commonly referred to as the human dimensions of global change.
Although nearly all major global change research programs have a human dimensions component, until recently, less emphasis has been placed on this area of global change research than on the study of the physical processes of global change. This is changing, however, and major new initiatives with respect to the study of the human dimensions of global change are emerging. One of these new initiatives is the development of international programs of research such as the Human Dimensions of Global Environmental Change Program (HDP) of the International Social Science Council (ISSC). This Program is focusing the attention of social scientists on understanding the driving forces of human actions within the context of specific issues associated with global change. Another initiative is represented by the recent establishment of the Consortium for International Earth Science Information Network (CIESIN), which has begun to institutionalize the development of access to, use, and understanding of data and information required to study the human dimensions of global change.
These new initiatives are stressing the integration of the study of the human dimensions with the physical processes of global change to achieve a comprehensive understanding of the Earth as a system. Such an understanding will require that studies of impacts caused by changes in physical processes be placed into the context of the societal structures and activities in which they occur in order to properly assess the cultural effects and the mitigation potential associated with these changes. Correspondingly, the results of studying the human dimensions driving forces need to be extended to assessments of the magnitude of their effects on the physical processes involved in order to determine their impact on global change.
CHALLENGES OF GLOBAL CHANGE RESEARCH FOR THE SOCIAL SCIENCES
Integrating the study of the human dimensions and the physical processes of global change will pose conceptual and methodological challenges to researchers. In particular, social scientists face significant challenges as they begin to address the human dimensions of global change. The challenges, presented by Jacobson and Price (1990), and expanded on here, include:
1) Adopting a global rather than a national or local perspective for research. Most methodologies for research in the social sciences have evolved for use at the national level or smaller areas. Most current studies of the driving forces of human actions are conducted on regional and local levels, and the applicability of the results beyond the study areas frequently is unclear.
CIESIN has found that studies of global change will require research at all geographic scales. In a sponsored set of pilot projects, CIESIN identified the need for appropriate development and use of investigative approaches (such as global context information, global hypotheses, common survey questions, and generation of comparable data sets for different sites) and scientific techniques (such as stratification, sampling and scaling) to develop coherent connections between local and global studies (CIESIN, 1993; Thomas, et. al., 1993).
2) Developing interdisciplinary coordination for defining variables and testing relationships between them. Much of the communications between social scientists and natural scientists, and even among the various social science disciplines, have been limited by a lack of shared concepts, terms, and information.
CIESIN has determined that close collaboration among scientists from diverse disciplines will be necessary for the design and establishment of inductive research strategies, such as case studies, or the development and evaluation of deductive strategies, such as modeling (Thomas, et. al., 1993).
3) Developing the necessary methodologies for collecting and evaluating data to provide longitudinal studies of the past with the present, and projections into the future.
CIESIN has gained insight into this based on its findings from a sponsored pilot project program (Thomas and Roller, 1993).
LANDSAT PATHFINDER PROGRAM
The U.S. National Aeronautics and Space Administration (NASA) has initiated a Pathfinder Program as one element of its Earth Observing System (EOS) Program which contributes to the U.S. Global Change Research Program. The objectives of the NASA Pathfinder Program are to: (a) provide EOS investigators, and other researchers, access to data sets applicable to global change research prior to the availability of EOS data, and (b) test the ability of the developing EOS Data and Information System (EOSDIS) to archive and deliver digital data (and derived products) from satellite-borne sensors to investigators.
The Landsat Pathfinder Program was initiated in 1991 as a component data development effort of the NASA Pathfinder Program. Landsat's global coverage with consistent data characteristics and format over a 20-year period offer the opportunity to develop consistently processed, multi-date, medium to high resolution Earth image data bases of sites selected for their local, regional, and global importance in global change research. The intent of the Landsat Pathfinder effort is to re-configure selected portions of the Landsat data archive to optimize the utility of Landsat data for global change research. The 20-year historical archive of Landsat MSS data at a spatial resolution of 80-meters, and 10-year archive of 30-meter Landsat Thematic Mapper (TM) data will provide multi-date, medium to high resolution data sets most effective for local studies of global change processes and for comparing analyses of global change processes from local to regional and/or global scales.
The Landsat Pathfinder Program currently includes two approved projects which are underway. The first is the Humid Tropical Forest Inventory Project which is associated with developing a complete three-epoch Landsat data base for the humid, tropical forests of the Amazon Basin, Central Africa, and Southeast Asia. The second is the North American Land Characterization Project which will bring together three epochs of historical Landsat data in a standard configuration for the land surface of the North American continent. The three epochs for both projects generally are the mid-1970s, mid-1980s, and the early to mid-1990s. A third project yet to be approved, the Intensive Test Site Project, proposes to develop multi-epoch Landsat data for a selected set of test sites distributed around the globe to be used for validating land cover data bases generated from continental and global scale coarse resolution satellite image data. It remains for additional projects to be defined and included as part of the Landsat Pathfinder Program.
Satellite image data such as Landsat data have been employed in research and applications by the physical and natural science communities for 20 years. Unlike the physical and natural sciences, the social sciences have made little use of satellite image data, having been constrained for reasons to include data costs, lack of training, and the costs and technological sophistication of the data processing procedures. However, the concept of the Landsat Pathfinder Program offers the potential to make satellite image data available in a form such that cost and technology are not limiting factors to its use. Development of standard, multi-date, geo-referenced image data products at reasonable cost as part of the Landsat Pathfinder Program could do much to make it economically feasible and more technologically attractive for social scientists to use the data, in much the same way that production of standard, demographic data sets (such as the Topologically Integrated Geographic Encoding and Referencing [Tiger] System of the U.S. Census Bureau) has facilitated their use.
The most appealing aspect of the development of Landsat Pathfinder Data Bases is the potential utility of the data for aiding the technical efforts of global change science. A sponsored set of pilot projects by CIESIN (1993) has indicated that an increased use of satellite image data such as Landsat data will offer significant benefits for meeting the previously described challenges facing social scientists studying global change. The following sections provide additional discussion of these benefits.
DEVELOPING A GLOBAL PERSPECTIVE FOR RESEARCH
As social scientists begin to address global change, it is likely that many of the studies will typically focus on a local or regional manifestation of a global change in order that cause-effect relationships be completely understood. However, to be most effective, such studies will need to be performed within the overarching framework of a global context. This will then enable global researchers to compare and contrast the dynamics of separate local and regional studies, and assess the cumulative effects of many local and regional changes.
The globally consistent Earth image data base that Landsat data provide would assist developing a global research context for social scientists. The data furnish a consistent image base for aiding the selection of study areas for global change research. For example, it could be used to identify similar environments, some of which show evidence of human activity causing changes, and some which don't; or to identify different environments being subjected to similar types of human activity. Use of a consistently prepared Landsat Pathfinder Data Base in various local and regional studies would serve as an integrating medium for the social sciences in global change research by facilitating development of uniform research methods and the production of information on global change that can be compared worldwide.
Direct observation of some forms of human presence and activity of interest to social scientists may require finer spatial resolution than Landsat MSS data provide. However, Landsat data may still have utility for those scientists through its capabilities for detecting and stratifying related phenomena of interest. For example, the location and extent of some features, such as urban and industrial complexes, and some lines of transportation, can be detected. Additional characterization of such features will require other sources of data or other data collection activities (including field work). Various types of stratification may also be achieved with Landsat data, thereby facilitating more efficient field data collection. For example, urban areas can frequently be stratified by factors which relate to population density and socioeconomic status. Such stratification would make sample-based censuses and other activities more efficient.
A Landsat Pathfinder Data Base will hold enormous potential for contributing globally consistent information on the dominant human-caused land use/cover conditions that are of consequence on a global scale, and studied in various locales and regions. The SSRC Working Group on Land-Use Change (1991) has identified several high priority types of land use for study; with priority being determined by the magnitude of their global extent and their significance to the major kinds of global environmental change, such as carbon cycling, methane emissions, and biodiversity. The types include cultivated lands, livestock lands, forest lands, settlements, wetlands, and surface water, all of which are generally amenable to interpretation and analysis procedures on Landsat MSS data.
Interpretation procedures include stratification by major land use type and, within major types, subsequent stratification by other meaningful characteristics such as vegetation type and density, field size, cropping patterns, or various types and densities of disturbances (such as cutting, burning, and harvesting in forested regions). Analysis procedures can include enumeration of various cultural features such as hamlets or agricultural fields, rank ordering according to characteristics such as relative size or proximity, and area determination through direct measurement and/or sampling and estimation techniques.
AIDING THE COORDINATION OF INTERDISCIPLINARY RESEARCH
A Landsat Pathfinder Data Base will be helpful in facilitating the interdisciplinary research that is essential for understanding and managing global change primarily because satellite data is one of the few sources of data that is not discipline-specific. Satellite data record many types of phenomena, some of which relate to the human sciences, and some to the physical sciences. Frequently, the various types of phenomena are of interest to both groups of sciences. Satellite data thus provide a unique frame of reference for integrating the disciplines.
A Landsat Pathfinder Data Base could help facilitate the joint acquisition and analysis of physical process data and social science data by providing a common frame of reference for the stratification of study areas, and the selection of sample units on which to obtain information about human activities and their effects on the environment. Because the data are gathered on the same units, they are inherently integrated in space and time. Also, during field work, joint examination of Landsat images and comparison of social and natural science data by the multidisciplinary scientists involved in field data collection can lead to new insights and better interdisciplinary understanding (CIESIN, 1993).
Two brief examples may be sufficient for suggesting the possibilities. A physical scientist might use Landsat data to monitor the areas of logging and the types of timber being removed from forested regions in order to estimate the effect on the global carbon cycle. But the same Landsat data can assist the social scientist studying the social dynamics of logging by enabling him to relate logging activity to the proximity of population centers, lines of transportation, and processing and distribution centers (such as mills and shipping points), many of which are discernible by analysis of the data. Landsat data may show many aspects of forest exploitation which are of interest to both the physical and social scientists, such as the size and arrangement of logging patches, the progression of the logging activity, the state of forest regeneration or the new land use/land cover that develops after logging. Equally important, Landsat data will readily show where logging is not occurring.
Another example might be the investigation of new irrigated agriculture, which can be readily detected by Landsat. Irrigation activities have human causes and effects on the socioeconomic conditions in the area, including food security. They also have effects on a variety of physical phenomena, including albedo, evapotranspiration, and other factors affecting climate.
Most social scientists may require that images be overlain with human dimensions strata such as political or census boundaries, or other meaningful cultural information in order for Landsat data to be used most effectively to extract social science information. Such combinations of data will enhance the interdisciplinary nature of the data. As an example, the merging of census boundaries with Landsat data will enable population data to be evaluated relative to elements of the physical environment (such as area occupied, and density and distribution with respect to land cover features), a type of analysis not readily feasible with much of the available population data.
AIDING LONGITUDINAL STUDIES
Land use change has been identified as a priority human dimensions issue because of the increased awareness that change in the earth's land surface due to human activity is a major driver of environmental change (SSRC Working Group on Land-Use Change, 1991). A recommended approach for studying global land use/land cover change is being developed by an Ad Hoc Committee of members of the IGBP and HDP programs (ICSU/ISSC Ad Hoc Committee, 1992). They state the need for land use information requirements that include the forms of land cover change, the patterns of change, and the rates of change as influenced by the driving forces of human actions. They note that current global aggregations of land use/land cover change are subject to debate for lack of precise measurement and/or estimation. They also note that global aggregate relationships to driving forces generally mask the increased complexities of human actions at the sub-global scale. To better understand the linkages between human activity and land use/land cover change, research programs are to be proposed to conduct systematic field studies of regional "situations" of human activity and land cover relationships, and to use the relationships for modeling and projecting land cover change.
Landsat MSS data provide the longest duration archive (20 years) of moderately high spatial resolution satellite image data for monitoring the types and rates of land surface change imposed by human activity. The consistent data characteristics of spatial resolution and digital format are readily conducive to the production of "change" images on which the spatial and temporal dimensions of land use/cover change can be detected and evaluated. The production of such images generally consists of co-registering two or more images of the same area acquired at different points in time, adjusting the radiometric properties of the images to normalize for varying atmospheric conditions, implementing a change detection metric on the combined data sets, and producing an output product that can effectively convey land surface change on an image, graphic, or statistical basis. (While the consistent data characteristics of Landsat MSS data enable ready production of change images, it should be noted that the procedures of change image production can also be implemented on multiple data sets of non-similar data characteristics, thus allowing combining Landsat MSS data with data from other sensors, such as Landsat Thematic Mapper (TM) and SPOT.)
For purposes of a Landsat Pathfinder Data Base, a careful selection of consistent time intervals from the repetitive coverage that is generally available for most regions of the world will enable consistent longitudinal studies of the forms, spatial extent, and rates of land use/cover change over the whole globe. Acquisition of such information on regional scales will begin to provide the more precise measurement/estimation of land use/cover change which can contribute to better global aggregations.
Direct evaluation of land use/cover change on Landsat data will also provide opportunities for identifying or inferring many of the proximate sources of land use change. Proximate sources have been defined as the end or near-end products of human action that comprise the immediate agents of change, usually the application of specific technologies (or practices) that change or maintain land cover in order that land use can be changed or maintained (Turner, 1989). High priority proximate sources of land use change were identified by the SSRC Working Group on Land-Use Change (1991) as those activities closely linked to changes in cultivation, livestock rearing, and forests. They were noted to include biomass burning, cutting, and planting (or permitting regrowth); fertilizer and biocide use; species and cultivar replacement; mechanical operations; and the build-up of physical infrastructure. Major additional sources of land use change are caused by irrigation, drainage, and energy production.
Many of the proximate sources of land use change listed above will be readily evident as sequential observation of the spatial patterns and temporal changes in land use/cover on Landsat data are evaluated in the context of other existing physical and socioeconomic data and information. For those regions where such ancillary information is abundant, the Landsat observations will enable consistent monitoring of known activities. Where and when the spatial and temporal changes are noted to depart from past trends, and for regions where little ancillary information is available, the Landsat observations can serve to guide and focus study of the activities by virtue of identifying locations where activities are occurring, which then will suggest types of additional data required and methods of acquisition.
The development of more extensive information linking the actual conditions of land cover with the proximate sources of land use change will ultimately enable more informed evaluations of the interactions among the various driving forces which motivate the proximate sources of change. Lack of use of Landsat data by the social science community in the past has limited the applicability of knowledge gained about the driving forces of human actions to the locales from where the land cover conditions and proximate sources of change were studied in close proximity. Extrapolation of such knowledge to larger regions involved conjecture and assumption about the interactions of the driving forces.
CONCLUSIONS
Landsat Pathfinder is a program of enormous potential value for social scientists studying global change. Landsat's global coverage with consistent data characteristics and format offer the opportunity to develop standardized, globally consistent earth image data bases. Such data bases will promote integrated research among the social sciences, and with the physical sciences, by facilitating development of uniform research methods and the production of information on global change that can be compared worldwide. Such data bases will enable interpreting and quantifying land surface conditions in the pre-EOS era. Such data bases will be important for detecting and evaluating changes in land use/cover over time.
Both social scientists and physical scientists can use satellite data in complementary ways. Satellite data helps promote the concept of connectedness between social and physical studies by providing a common source of information for integrated studies, and enabling geographic and temporal integration of their studies. This is a major, unique contribution that a Landsat Pathfinder Data Base could provide.
The goal of Landsat Pathfinder is to provide affordable remote sensing data, ready for analysis. Indeed, by selecting from the archive cloud-free scenes for multi-year epochs, by processing that data to remove geometric distortions (to be compatible with GIS), merging these data with appropriate socioeconomic data, and making that data available at low cost, Landsat Pathfinder can ensure that social scientists are able to use satellite data. The benefits of this to studying the human dimensions of global change will be quite significant.
Consortium for International Earth Science Information Network (CIESIN), 1993. CIESIN 1991 Science Pilot Project Program, Final Report, Saginaw, Michigan.
ICSU/ISSC Ad Hoc Committee of the International Geosphere-Biosphere Program and Human Dimensions of Global Environmental Change Program, 1992. The Case for the Study of Global Land-Use and -Cover Change. Draft Report, June.
Jacobson, H.K. and M.F. Price, 1990. A Framework for Research on the Human Dimensions of Global Environmental Change. International Social Science Council (ISSC)/United Nations Educational, Scientific, and Cultural Organization (UNESCO) Series: 3.
SSRC Working Group on Land-Use Change, 1991. Untitled Report. Committee for Research on Global Environmental Change, Social Science Research Council, May.
Thomas, M., N. Roller, J. Colwell and C. Kottak, 1993. Integrating Regional Studies of Deforestation into a Global Change Context. Proceedings of the 25th International Symposium, Remote Sensing and Global Environmental Change, Graz, Austria, 4-8 April.
Thomas, M. and N. Roller, 1993. CIESIN's Experiences with Integrated Global Change Research. Proceedings of the 25th International Symposium, Remote Sensing and Global Environmental Change, Graz, Austria, 4-8 April.
Turner, B.L., 1989. The Human Causes of Global Environmental Change. In R.S. DeFries and T. Malone, eds. Global Change and Our Common Future. National Academy Press, Washington, D.C.