Monitoring and modelling tropical deforestation: Introduction to the Special Issue

This Special Issue of the Singapore Journal of Tropical Geography derives from a set of papers on remote sensing and forest governance in Indonesia, presented at a workshop held at the Center for International Forestry Research (CIFOR) in Bogor in June 2004. The workshop assembled experts in forestry, geography and biodiversity conservation to examine the ways that imaging satellites such as Landsat are shedding new light on problems and processes associated with tropical deforestation. Although the workshop mainly focused on some of the major drivers of this process in Indonesia, the issues of illegal logging, fires and the establishment of industrial plantations are common to many other parts of the tropics where forests have been converted to other land uses, often at rates that merit global concern (Sidaway & Teo, 2005). Although alarm bells have been rung some time ago to warn against the consequences of tropical deforestation (e.g. loss of biodiversity and ecological services, air pollution from fires, climate change), the scope of the problem remains an area of active study and debate. As a tool for detecting change in forest cover, satellite imagery increasingly determines how we measure and monitor loss of tropical forest. More pertinently, the increased availability of inexpensive imagery coupled with increased computer power since the early 1990s has facilitated the diffusion of new information among a wide range of actors in civil society. This has the potential to revolutionize forest governance by improving transparency in the forestry sector in particular, and in land use policymaking more generally, by making available information that some central governments and large landowners have guarded quite closely. Although this Special Issue explores some of the limitations of existing satellites for mapping forests, with relatively new sources of satellite imagery (e.g. MODIS or Moderate Resolution Imaging Spectrometer), near real-time monitoring of tropical deforestation is starting to become as much a reality as the monitoring of hotspots from vegetation fires has been for over a decade. Thus, with an evolving constellation of orbital and geostationary satellites, we should expect operational monitoring of land use and land cover change (LULCC) in the tropics at more frequent intervals and at finer spatial resolutions. The improved ability to observe changes at finer temporal and spatial scales is likely to translate into major advances in predictive capability that will be helpful for understanding changes in land use dynamics. Over the past decade, a dominant approach to understanding LULCC dynamics is to link data available from satellite imagery to ancillary information in order to model the underlying drivers of landscape transformation. Efforts to quantitatively link land cover to theorized driving forces of change are not novel; for instance, earlier statistical models linked regional or national land cover metrics (e.g. per cent forest cover) to demographic and other factors (e.g. population density, rate of economic growth). The greater temporal and spatial resolution currently afforded by satellite imagery thus allows a finer spatial lens through which to view such dynamics. When corresponding fine-scale or disaggregate data on biophysical or socioeconomic drivers are available, LULCC models