Date of Award
Program or Major
Doctor of Philosophy
Quantification of the direct impact of land use in the tropics on net biotic carbon flux requires estimates of rates of deforestation, pre- and post-disturbance biomass, and fate of the cleared land. Synoptic observations of the extent, persistence, rates of secondary succession, and structure or biomass of regrowing forests would also help constrain estimates of net carbon flux due to tropical land use. While remote sensing applications can provide estimates of the rates of deforestation and the fate of the cleared land (pasture, croplands, or secondary vegetation), techniques for estimating persistence, rates of succession, and biomass of secondary vegetation are needed.
We documented the spatial and inter-annual variability in the rates of forest clearing, formation rates and persistence of secondary vegetation for 3 sites in Amazonia and 4 sites in Southeast Asia using Landsat TM data from mid-1980s to late-1990s. Secondary vegetation was a large, rapidly changing pool. Variability in the observed annual rates of deforestation and secondary vegetation formation was high. The transition probabilities of both the formation and clearing of secondary vegetation decreased with age. Persistence of the secondary vegetation pool was also highly variable, likely indicating two distinct land use trajectories: rotational agriculture/pasture maintenance versus abandonment.
We also evaluated the spatial, temporal, and noise constraints of JERS SAR data for mapping and monitoring biomass of secondary vegetation in Rondonia, Brazil. Results indicate that quantitative estimates of biomass using single date JERS-1 imagery is problematic because of temporal variability in backscatter due to intrinsic texture, system noise, and environmental effects. However, JERS-1 data are still useful for distinguishing of secondary vegetation stands at different stages of development. Multi-temporal analysis significantly improves biomass estimates to the point where it is possible to map changes in biomass. Slight reductions in the variability in estimates of normalized radar cross-section greatly improve biomass estimation. Merging JERS-1 SAR data with Landsat TM derived age estimates improved characterization of clearings and secondary vegetation in Rondonia by providing information on the relative differences in secondary vegetation development and residual slash with age.
Salas, William Albert, "Optical and radar remote sensing of land use and land cover change in the tropics: An assessment of deforestation and secondary vegetation" (2001). Doctoral Dissertations. 27.