Experiences and perspective in compiling long-term remote sensing data sets on landscapes and biospheric processes

Abstract

Remote sensing has been employed to study the Earth's landscapes for more than a century. Over the last thirty years major changes in all aspects of the technology, as well as scientific understanding of the measurements, have revolutionized use of these data for terrestrial research. The emergence of biophysical interpretations of land remotely sensed data has created great interest in the potential of these data to study the dynamics of landscapes and biospheric processes as contributors to global change. However, the evolutionary history of land remote sensing, particularly in the Landsat era, has in many ways worked against this use of the data. Until recently, there was little appreciation of the need to compile long-term historical records of these observations. There are currently no formal scientific institutional systems dedicated to preserving and distributing the long-term record of Earth observations acquired by Landsat, AVHRR or other land observing systems. As the International Geosphere-Biosphere Program begins, this issue should be one of the first priorities of the participants.

Seasonal Land-Cover Regions of the United States

“…These data are routinely used for land-cover assessment, although several practical issues have limited their usefulness for land-cover mapping over subcontinental or larger areas (Goward 1990). The large volume of data (number of scenes and number of pixels) required to cover even a single continent and the complexity of data acquisition and analysis have made such analyses prohibitively expensive (see Woodwell et al 1984).…”

“…This process tends to eliminate clouds except in areas where there are no cloud-free pixels during the 14-day period. A time series of these composite "greenness" images depicts phenological events, most notably the annual progression from Spring greenup ("green wave") when the northern hemisphere's deciduous trees develop leaves and crops emerge and develop to the ensuing Fall's retrogression ("brown wave") when trees drop their leaves and crops reach senescence and are harvested (Goward et al 1985;Goward 1989). With data such as these we are able to define regions having distinctive seasonal characteristics.…”