Date of Award

Winter 1997

Project Type

Dissertation

Program or Major

Natural Resources

Degree Name

Doctor of Philosophy

First Advisor

Russell Congalton

Abstract

The 1983 Forest Inventory and Analysis (FIA) data of the states of Maine, New Hampshire and Vermont (the study area) contain large amounts of field-measurements of many ecologically important variables. Despite the vast potential usefulness of the FIA data for scientific research, the data were until now, literally unused except for a few administrative purposes, because of problems in the way the data were organized, summarized, and coded for storage. The primary objective of this research was to solve the problems that had thus precluded these FIA data from use in scientific applications, and present the data in a form that is readily accessible and usable for research. This objective was achieved by adapting the un-summarized data in a relational database management system (RDMS) and geographic information systems (GIS). RDMS-GIS technologies would make these data amenable to more types and multiple spatial scales of analyses than previously possible, thus providing the scientific community with an unusually large, high-quality, and spatially referenced data set.

The FIA data also contain field and laboratory measurements of soil properties made at the geo-referenced FIA plot locations. These soil data also provided the basis for other studies in this dissertation. These studies included analyzing the spatial variability of selected soil attributes in the study area; evaluating the nature of the differences in specific soil properties among the ecological land classification map (ECOMAP) section and subsection units; and assessing the variability of specific soil properties in the NRCS-State Soil Geographic Database (STATSGO) of the study area. Both the ECOMAP and the STATSGO studies involved the use of GIS techniques and multivariate statistical methods for map unit analyses.

This dissertation also included more theoretical investigations relating to applied statistics and soil science. One of these addressed the unanswered question of whether or not it is necessary to use non-linear transformations prior to computing variability statistics from non-normally distributed soil data, and explored the use of coefficient of variation as a semi quantitative index of nonnormality in soil variables. Another study looked at why and how error matrices and related statistics can be used as an effective, comprehensive quantitative method of evaluating soil classification and soil map quality.

Download

Share

COinS