Date of Award

Spring 8-1-2008

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Geography, Geology, and Anthropology

First Advisor

Ryan R. Jensen

Second Advisor

Susan M. Berta

Third Advisor

Paul W. Mausel

Abstract

The ever-increasing concentration of anthropogenic greenhouse gases (C02, CH4, and CFCs) has been identified as a likely (greater than 90% confidence) cause of the observed increase of global mean temperatures since the mid-20th century (IPCC, 2007). The effect of human-induced climate change could be unprecedented and far-reaching. Carbon sequestration into trees and forests is an effective and inexpensive way for mitigating the C02 level in the atmosphere. Hence, accurate measurement of biomass will be of great importance to global carbon cycle and climate change. This study performed a wavelet-based forest aboveground biomass estimation approach in a temperate deciduous forest, the Hoosier National Forest, in Indiana. Wavelet analysis, specifically two-dimensional discrete wavelet transform (DWT) was applied to ASTER images to obtain wavelet coefficients (WCs), which were correlated with forest inventory data using multiple linear regression analysis to investigate the relationship. Different mother wavelets and level of decomposition were tested. Moreover, vegetation indices, RATIO, normalized difference vegetation index (NDVI), and principal component analyses (PCA) were computed and correlated with field biomass measurements. The results indicate that wavelet coefficients correlate better with field biomass data than vegetation indices. For level one decomposition, the correlation coefficients are 0.3 to 0.5, while 0.1-0.3 for vegetation indices; for level two decomposition, the overall R value increased by 0.2, and for level three, the R value can be increased to 0.6-0.7. Meanwhile, tree per acre and basal area were also examined and correlated with field measurements. This study demonstrates that wavelet-based biomass estimation could be a very promising approach for solving the uncertainty between reflectance value from satellite images and forest biomass and therefore providing better biomass estimation; however, further research is needed for identifying robust wavelet coefficients and optimizing procedures.

Download

Share

COinS