Comparing, matching, detecting, and predicting drought-induced tree mortality in the Sierra Nevada, California

Abstract

In the last decade, technology and computing capabilities have evolved to allow the detection of individual trees with remote sensing. This innovation is invaluable to researchers and managers who seek precision methods to measure forest condition in a rapidly changing world. To apply remotely sensed methods to evaluate population trends, a framework is required to enable the application of these methods to detect and predict the trends of tree populations. Such a framework requires assessing the uncertainty and bias associated with applying remotely sensed methods to reflect the population trends of individual trees. The goal of my dissertation was to establish a framework for defining the population trends of remotely sensed trees in the context of an extensive tree mortality event. Further, I endeavored to measure the uncertainty and bias associated with applying remotely sensed trees to represent drought-induced mortality trends.