Date of Award
Fall 2000
Project Type
Dissertation
Program or Major
Natural Resources
Degree Name
Doctor of Philosophy
First Advisor
Barrett Rock
Abstract
The current study assesses the potential of hyperspectral data for monitoring the initial stages of damage in Norway spruce forests characterized by subtle changes in foliar chlorophyll and chemistry. Both field and airborne high spectral resolution reflectance measurements were obtained for selected study sites in the Krusne hory, Czech Republic.
High spectral resolution airborne canopy data and field foliar samples were acquired simultaneously in August 1998 for a total of 51 study sites within the Krusne hory. The sites were selected to represent a full range of damage conditions in even-aged Norway spruce (Picea abies (L.) Karst) stands located between 820--920m elevation.
Reflectance, foliar pigments, nitrogen and chemical constituents were determined for first-, second- and third-year needles. A strong correlation to damage was established for the foliar chemistry. A significant increase in polar compounds (such as tannins, sugars and starch) and a reduced needle lignification occurs with increasing damage.
Foliar chemical constituents appear to be effective indicators of long-term environmental conditions. The strong relationship between damage level and polar compounds suggests high potential for use of these constituents as bio-indicators of stress.
Both field and airborne high spectral resolution data separate the initial forest damage classes. Based on field reflectance measurements for third-year needles, derivative indices from the red edge region were most strongly correlated to damage level, followed by indices ratioing damage-sensitive and damage-insensitive bands and a parameter describing the fit of an Inverted Gaussian curve. Red/red edge spectral data from the Airborne Solid State Array Spectrometer (ASAS) had the highest potential for separation of initial levels of damage, which corresponds with the region suggested as most sensitive to damage as seen in conducting the field reflectance measurements.
Both optical and derivative indices incorporating reflectance from the red edge region present the most potential as indicators of initial damage in both leaf and canopy spectra.
Recommended Citation
Entcheva, Petya K., "Remote sensing of forest damage in the Czech Republic using hyperspectral methods" (2000). Doctoral Dissertations. 2139.
https://scholars.unh.edu/dissertation/2139