Conservation impact assessment and SAR forest cover mapping in the Colombian Andes
OA Version
Citation
Abstract
This thesis addresses research questions from two distinct yet complementary fields—conservation science and remote sensing—through a case study in the Colombian Andes. In Chapter One, we explore the impact on forest cover of the largest and longest-running public land acquisition (PLA) program in the tropics between 2000 and 2021. Using matching and Difference-in-Differences with multiple time periods we find that as of 2021 there has been a 3.5% increase in forest cover on protected parcels and that impact increases for at least 10 to 12 years post-treatment. We also find that impact varies significantly by factors like slope, accessibility, and department. In Chapter Two, we attempt to improve the forest cover data used in Chapter One by integrating synthetic aperture radar (SAR) observations from Sentinel-1 for areas where persistent cloud cover precludes the use of optical data. We find encouraging evidence to suggest that SAR data can be used with the continuous change detection and classification algorithm to detect forest change in topographically-complex regions, but conclude that accuracy improvements and widespread workflow adoption are dependent on the accessibility of high resolution digital elevation models and improved radiometric terrain correction for Google Earth Engine.
Description
2024
License
Attribution-NonCommercial 4.0 International