Modeling historic, current, and available aboveground forest biomass along the Missouri River corridor

Abstract

This research presents the culmination of statistical, landscape, and geospatial analyses that examine the geographic dynamics of aboveground forest biomass (AFB) within the Missouri River corridor, Missouri USA. The Missouri River corridor is a region specifically within Missouri that encompasses 106,000 km², and is regarded as a processing region for improving the viability of Missouri's biomass/biofuel industry. Current and historic forest inventory data coupled with remote sensing, edaphic, physiographic, and climate variables were integrated into an ensemble regression tree method, Random Forest (RF), to estimate AFB, determine external driving forces of AFB, and visualize geographic locations where the greatest deviations exist between current and historic AFB values. The applicability of constructing a hybrid modeling framework using RF was initially tested in Chapter 2 by estimating current (observed data derived from Forest Inventory and Analysis) and theoretical (based on 20% of AFB found within Missouri) AFB, and calculating the percent change to determine percent changes in AFB across the landscape. The third chapter extended the RF modeling procedure to include historical information derived from General Land Office (GLO) data to estimate a baseline measure of AFB. Current AFB was again estimated and then compared to historic values where an additional synthesis was performed to investigate the top predictors of AFB. The fourth chapter examined a fuzzy logic approach for developing a suitability index based on available AFB. Available AFB was determined by applying physical constraints onto estimated AFB from the RF model, which included forest transitions and distance to rivers. The model results failed to reject our null hypothesis that there were no differences between observed and predicted AFB, and x model accuracy was very low for all AFB estimate. Results from these investigations indicated that 1) the greatest potential for increasing AFB may be along the floodplains of the Missouri an