This report describes a doctoral dissertation research project on integrating localized knowledge of livestock herding in Central Tibet with satellite imagery of grasslands. Participatory mapping of land use/cover was carried-out in Gyama township in the Tibetan Autonomous Region of China during parts of 1987 and 1988. Prior to 1959, most tracts of cultivated land and grazing areas in this locale were controlled by the estate of a noble family (TPP 1987). An area of approximately 292 sq km was controlled.

Building upon approaches in public participatory geographic information systems (PPGIS), agrarian change studies, and geohistory, I use geographic information science (GIS) to develop a spatial model for the study of population-environment interactions in Tibet. The purpose of this work is to develop a theoretically informed and inductively designed spatial model for linking macro-scale physical and biological processes observed and modeled at a regional scale with micro-scale human organizational and decision-making processes. Participatory mapping efforts may contribute to better understanding GIS/remote sensing representations of Tibetan society and environment, and how China’s political and economic reforms in Tibet have altered long-term patterns of pastoral land use.

Livestock herding is a key basis of Tibet’s economy, and has been for millennia. China’s post-1959 reforms in Tibet significantly changed long-term Tibetan demographic processes resulting in rapid population growth, and the allocation of land and animals to households on a per capita basis (Ryavec 1999). Studies on agrarian change in the Third World, and historical transformations of societies from feudalism to capitalism, have not made sufficient use of spatial models to better understand how these transformations have affected population-environment interactions. The "Brenner debate" focuses on how social and economic differentiation gave rise to capitalism, but ignores how corresponding changes in land use and land cover interacted with these processes. According to Brenner, while the English lords maintained their control over the land, it was the market in leases which provided the basic condition for differentiation among the tenants, and resulted in the rise of capitalist tenant farmers (Glennie 1987). The older generation living in Central Tibet today grew-up within indigenous institutional land tenure arrangements between households and monastic and noble estates that defined most economic processes and forms of land use. Studies on the social and environmental change that occurred when Western peasant economies turned into capitalist economies must, in contrast, take place centuries after the events occurred. Tibet’s special place in world history merits detailed study. Research to date on Tibetan herding systems and change in land tenure has been based exclusively on fieldwork and small sample plots (Goldstein et al 1990, Cincotta et al 1991) making it difficult to integrate within a GIS framework. At the same time, an initial GIS database of population-environment interactions in Central Tibet I developed is based solely on digital land cover and census data (Ryavec and Veregin 1998) and needs to be integrated with localized knowledges of people.

In light of the inherent complexity of developing spatial models of population-environment interactions over time and space, I draw upon several theoretical perspectives. The Annales school of history is linked by interest in environmental and social topics. Perhaps the most well-known member of the school was Fernand Braudel who developed a distinctive geohistory focusing on a ‘total history’ integrating long-run changes in ways of life with short-term political events with which, he argued, historians were wrongly preoccupied. This perspective has some similarities with the classic cultural ecology school of geography pioneered by Sauer (1925) that seeks the impact and role of culture in the physical form of landscapes. The key difference between these schools lies in the geographic scale of their approach to space. The geohistorical approach considers geography as mainly a macro-scale physical territory which confines, sustains and, in part, explains the history of people and civilizations (Braudel 1963). Cultural ecologists, on the other hand, emphasize the micro-scale role of land use practices in shaping distinctive cultural landscapes.

Recent advances in GIS, and the development of global and regional land cover data sets, place the concerns of cultural ecology and geohistory in a new light. At the local level, PPGIS facilitates linking micro-scale human organizational and decision-making processes, including the role of cultural interests and localized knowledges, with regional scale land cover data sets. In turn, these regional data sets, when integrated with localized knowledges of people, open the door to new geohistorical perspectives on how regions partly shape societies.

My research project constitutes a form of de facto PPGIS as I interviewed local families and participated in some of their livestock grazing activities, often while recording locations with a GPS receiver. This assistance was invaluable in helping me to understand how and why they use the land in the ways they do. In this sense, a limited form of PPGIS took place. There is a difference, though, in the de facto nature of this PPGIS compared with the de jure PPGIS work usually discussed in the PPGIS literature. The major difference here is that local Tibetans did not carry-out the PPGIS themselves, nor have they benefited in any way from it. Even if political barriers to open, community based PPGIS efforts in Tibet were to disappear, the almost total lack of electricity and computers in the countryside would make it extremely difficult for geographic information systems (GISs) to be made available to Tibetans for addressing their needs. The ‘crisis of representation’ in academia today is a debate over who should speak for whom. If people outside of Tibet, including the Tibetan exile community, wish to gain firsthand knowledge of what is happening in Tibet in terms of changes in land tenure, land use and impacts on the environment, my GIS data sets and derived findings may be of significant value to them.

A collection of studies on how GISs are being used by indigenous peoples to better respond to new pressures on their resources appeared in the winter 1995 issue of Cultural Survival Quarterly. The insights provided on these projects offer important information for better designing GISs to address the needs of people in regard to their land use practices. I have taken these concerns into consideration when developing my GIS data sets of grazing patterns, grassland ecology, and socioeconomic factors in Tibet.

An important point raised by Kemp and Brooke (1995) concerning the development of digital maps of ecology and land use by the Nunavut Inuit of Cananda states that the process should be data driven rather than technology driven. Methodological decisions concerning the information people wish to store and analyze, and the type of data base they wish to develop, should be of paramount importance. Ideally, only after these decisions have been made should software and hardware be acquired. These concerns help to inform my efforts. When difficulty arises in linking disparate GIS data with fieldwork findings a new opportunity presents itself to improve upon our understanding of existing remote sensing and digital cartographic data sets of Tibet.

Some ways to integrate participatory mapping of land use practices with GIS data sets are outlined by Smith (1995), and include seasonal calendars and local histories. The seasonal dynamics of Tibet’s grassland ecology impacts greatly on herding patterns, leading to distinct seasonal pastoral land use regimes. Vegetation variations in time and space define movements and foraging behavior of livestock and influence the manner in which they affect the rangeland ecosystem, and also impact on the lives of the people who herd them. These seasonal dynamics indicate a need for multi-temporal satellite imagery, something monthly Advanced Very High Radiometer (AVHRR) images of vegetation biomass provide. In fact, the use of multi-temporal remote sensing imagery to study local social and environmental history is not very well explored and documented in the PPGIS literature. My research makes a special contribution to PPGIS in this regard.

My research on Tibetan geography represents a pioneering effort to develop a spatial model of Tibet’s grassland ecosystem and patterns of pastoral land use during both premodern and contemporary periods. Four objectives guide my efforts:

1. To develop a series of satellite images of key seasonal changes in grassland biomass for the entire Tibetan Plateau at a macro-scale resolution of 1 sq km.
2. To integrate local-scale herding patterns, including peoples’ perceptions of historical changes in land tenure as well as spatial and temporal variation in grassland type and biomass, derived from a Public Participatory GIS conducted in Gyama township of the Tibetan Autonomous Region of China during parts of 1987 and 1988, with macro-scale land cover data.
3. To map the arable land and grazing areas controlled by a Tibetan estate in Gyama township prior to China’s post-1959 reforms in Tibet, and to document how its control over the land affected ways of life and land use.
4. To construct a spatial model of pastoral land use on the Tibetan Plateau in relation to macro-scale physical and biological processes observed and modeled with remote sensing imagery and digital land cover maps. Local historical and contemporary data on land use, land cover and land tenure obtained from participatory mapping will be integrated within the model as a micro-scale case study.

GPS technology was employed to survey specific grassland areas grazed by people’s yaks, goats and sheep to facilitate linking with a spatial model. Interviews and participant observation of ten herding households elicited information about how past herding patterns were affected by indigenous rules and norms under Tibet’s premodern estate system as compared to limitations imposed by the physical environment. Attention also focused on how China’s household responsibility system implemented during the early 1980s affects pastoral land use and ways of life in the locale today.

A digital remote sensing database of grassland areas on the entire Tibetan Plateau has been created. This database is based on monthly normalized difference vegetation index (NDVI) data obtained by the National Oceanic and Atmospheric Administration’s Advanced Very High Resolution Radiometer (AVHRR) satellite sensor at a resolution of one sq km. NDVI data derive from mean monthly vegetation biomass values for 1992 and 1995. Following the methodology of Reed et al (1994), an additional eight images of vegetation metrics linked to key phenological events of grasslands have been created:

1. Time of onset of greenness (beginning of measureable photosynthesis)
2. Time of end of greenness (cessation of measureable photosynthesis)
3. Duration of greenness (duration of photosynthetic activity)
4. Time of Maximum NDVI (time of maximum measureable photosynthesis)
5. Value at onset of greenness (level of photosynthetic activity at beginning of growing season)
6. Value at end of greenness (level of photosynthetic activity at end of growing season)
7. Value of maximum greenness (maximum measureable level of photosynthetic activity)
8. Range of NDVI (range of measureable photosynthetic activity)

I am integrating socioeconomic data obtained from participatory mapping with these eight key digital satellite images in order to construct a spatial model of historical and contemporary human use of grasslands in relation to both the impact of land tenure arrangements, and spatial and temporal variation in grassland biomass.

Braudel, F. 1963. A History of Civilizations. London: The Penguin Press.