Evaluation of low-altitude vertical aerial videography as a method for identifying and estimating abundance of residual trees

Graduation Date

2006

Document Type

Thesis

Program

Other

Program

Thesis (M.S.)--Humboldt State University, Environmental & Natural Resource Sciences: Natural Resources Planning and Interpretation, 2006

Committee Chair Name

Steven J. Steinberg

Committee Chair Affiliation

HSU Faculty or Staff

Keywords

Residuals, Aerial videography, Humboldt State University -- Theses -- Natural Resources Planning and Interpretation, Remote sensing, Legacy trees, Aerial video, Old-growth residuals

Abstract

Low-altitude color aerial video was acquired within the northern section of the Redwood Region in Humboldt and Del Norte Counties, northwestern California. Four interpreters viewed a sample of video and identified residual trees within one-third hectare circular plots. Each sample plot was ground-truthed and residuals were identified and mapped. Error matrices presented indicated that identification of residuals was not highly accurate for individual trees, nor consistent among interpreters. However, for three of four interpreters, linear regressions of number of interpreter-identified residuals per plot versus number of field-identified residuals per plot had significant slopes (p 0.005). Coefficients of determination were 0.23, 0.22, and 0.41 for the three interpreters. Interpreters were not very successful at identifying old-growth legacy trees in video, and clonal rings of redwood trees were often mistaken for residuals in video due to large crown diameter. It was concluded that low-altitude color aerial videography may not be accurate enough for identification of individual residuals, but could be used effectively to estimate abundance of residuals in an area of interest, e.g. a watershed. Double-sampling and training of interpreters based on lessons learned in this study could improve prediction intervals of future studies. Identification of legacy trees in aerial video needs further investigation.

https://scholarworks.calstate.edu/concern/theses/cn69m633f

Share

 
COinS