Estimation of trees height and vertical structure using SAR interferometry in uneven-aged and mixed forests

Area of Science:

• Forestry
• Remote Sensing
• Geospatial Analysis

Background:

• Accurate forest height estimation is crucial for biomass, succession dynamics, and management decisions.
• Assessing tree height in complex, multi-storied forests presents significant inventory and accuracy challenges.
• Synthetic Aperture Radar (SAR) interferometry is explored as a technique to overcome these challenges.

Purpose of the Study:

• To estimate tree height and determine the vertical structure of mixed, uneven-aged forests using SAR interferometry.
• To evaluate the potential of Envisat ASAR data for characterizing forest vertical structure.
• To assess the accuracy of SAR-derived forest parameters against ground truth data.

Main Methods:

• Utilized SAR interferometry with Envisat ASAR data to generate a Digital Surface Model (DSM).
• Estimated tree height and forest vertical structure by comparing the DSM with a Digital Elevation Model (DEM).
• Validated results using ground truth data from 11 sample plots (100x100m).

Main Results:

• The study estimated a mean tree height of 7.69m with a standard deviation of 22m relative to reality.
• The vertical structure was accurately identified as three-storied across all plots, matching ground truth.
• Discrepancies were noted in the percentage share of trees within the understory and middle story compared to ground data.

Conclusions:

• SAR interferometry, using Envisat ASAR data, provides an acceptable accuracy for determining forest tree height.
• The technique is effective for characterizing the vertical structure of complex forest stands.
• This method shows promise for improved forest inventory and management applications.