Methods for improving timber inventory projections in Alabama

A new regional timber inventory projection system (DPSupply) is presented in this dissertation. This new system uses dynamic programming to identify the optimal decision rule for determining available harvest based on stumpage net present value over a range of specified price levels. The final harvest quantity decision is made based on the market demand for stumpage products and cutting priorities are based on highest net present value. Once a schedule of available harvests is determined, harvests are simulated and a residual inventory estimate is produced for each future period in the projection. A comparison of the model's projection of inventory with FIA's 1990 estimate of inventory for Alabama is provided and indicates that model estimates fall within a 95% confidence interval of the FIA estimates. An example projection of Alabama's inventory through 2010 is also provided.; The primary data for this system include data from the U.S.D.A. Forest Inventory Analysis (FIA) work unit, the Alabama Department of Revenue (severance tax) and Timber-Mart South (stumpage prices). The two most recent FIA surveys are used to develop sets of dbh (diameter at breast height) and volume growth models by forest type and size class, and to build sets of product proportion functions. The Timber-Mart South data were used to construct price equations for stumpage valuation. Timber removal levels are estimated based on the severance tax information.; The DPSupply system has the ability to project inventory at the product level (sawtimber or pulpwood). A multinomial logit model was used to estimate sets of product proportion functions to determine product volume in each stand for each forest type and size class. This technique provides a powerful tool for inventory projection research. All product proportion function models contain the same variables, and the models are used in the MANAGE module of the DPSupply system to calculate net present value for each plot based on the plot's mix of products. They are also used in the HARVEST module of the program to determine the set of harvestable acres that should be used to meet individual product level demands. The product proportion function method appears robust enough to have broad application, including adapting it to projection systems that are age based (rather than dbh based).; Severance tax data were used with the FIA removals data to determine actual annual removal levels. Using an adjustment procedure developed through this research, a comparison of South Alabama adjusted removal levels with other specifications of removals highlights the importance of jointly using the FIA data and severance tax data to provide more accurate inventory projections.