Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.

Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Simple, novel approaches to investigating biophysical characteristics of individual mid-latitude deciduous trees

Humphrey Wafula Kalibo, University of Nebraska - Lincoln

Abstract

Forests play a critical role in the functioning of the biosphere and support the livelihoods of millions of people. With increasing anthropogenic influences and looming effects associated with climatic variability, it is crucial that the research community and policy makers take advantage of the capabilities afforded by remote sensing technologies to generate reliable and timely data to support management decisions. Set in the species-rich woodland of Prairie Pines in Lincoln, Nebraska, this research addresses three distinct objectives that could contribute towards forest research and management. First, three supervised classification algorithms were applied to two hyperspectral AISA-Eagle images to evaluate their capability for spectrally identifying selected tree species. The findings show that each algorithm had low to moderate overall classification accuracies (46%-62%), probably due to mixed pixels resulting from pronounced heterogeneity in tree diversity; however, the algorithms could be a rapid means to assess species composition. The second objective is an investigation into how twelve individual morphologically different deciduous trees transmit incoming photosynthetically active radiation (PAR) over the course of the growing season. It was found that more diffuse light was transmitted than direct light, dictated by seasonality, vegetation fraction (VF), and leaf size. In the final objective, VF derived from upward-looking hemispherical photographs of twelve deciduous tree canopies and eight spectral vegetation indices (VIs) calculated from in situ single leaf-level reflectance data were used to investigate whether the VIs could mimic and estimate the temporal patterns of measured VF of each tree over the growing season. The findings show that all the indices accurately depicted the temporal patterns of the photo-derived VF. NDVI and SAVI had the highest correlations (R 2 > 0.7; RMSE < 20%; d > 0.7; E > 0.8) and closely mirrored the temporal patterns of VF for nine of the twelve trees, while SR worked well for low-foliage trees. The amount of foliage cover affected the accuracy with which the VIs estimated the VF of the various trees. Overall, this study has suggested a simple methodology that combines two distinct sets of data of different spatial scales to examine a vegetation biophysical parameter over a growing season. The findings provide preliminary insights into how individual mid-latitude deciduous trees partition incoming solar radiation at each phenological stage and over the growing season. The research should stimulate more robust inquiry into the use of both remote and in situ sensing tools in the study of vegetation biophysical parameters. ^

Subject Area

Geography|Environmental management|Remote sensing

Recommended Citation

Kalibo, Humphrey Wafula, "Simple, novel approaches to investigating biophysical characteristics of individual mid-latitude deciduous trees" (2016). ETD collection for University of Nebraska - Lincoln. AAI10104391.
http://digitalcommons.unl.edu/dissertations/AAI10104391

Share

COinS