Authors: Amelia Bradshaw*, University of North Carolina - Wilmington, Narcisa G Pricope, University of North Carolina at Wilmington, Forrest Stevens, University of Louisville, Andrea Gaughan, University of Louisville, Steele Olsen, University of North Carolina at Wilmington
Topics: Remote Sensing, Africa
Keywords: KAZA, UAV, RapidEye, Southern Africa, Savanna, Drylands, Human-Environment, Vegetation Structure, Remote Sensing
Session Type: Interactive Short Paper
Start / End Time: 3:20 PM / 5:00 PM
Room: Balcony M, Marriott, River Tower Elevators, 4th Floor
Presentation File: No File Uploaded
The Kavango-Zambezi Transfrontier Conservation Area (KAZA) is a collaborative effort among five nations in Southern Africa to manage a vast expanse of drylands ecosystems, home to a wide range of wildlife and a substantial human population. Over the past few decades, this region has seen a shift toward a drier climate with higher spatial and temporal precipitation variability, which have affected the overall productivity and condition of the landscape. This work is part of a larger collaborative effort aimed at understanding the interdependencies between humans and their natural resources base and how shifts in biophysical parameters (such as vegetation changes) are influencing livelihood outcomes for smallholder communities in the region. To begin examining these shifts in vegetation at various spatial and temporal scales, we conducted over sixty unmanned aerial vehicle (UAV)-based ground surveys across different vegetation types in a community-based organization in Namibia during the summer of 2017. We collected hyper-fine resolution multispectral and vegetation structure data and used structure-from-motion data processing workflows to extract products such as orthophotos, digital terrain models (DTM) and digital surface models (DSM). In this paper, we present comparative results of structural vegetation classifications between UAV-collected field data and RapidEye imagery acquired over our study region on anniversary dates to assess the potential of extrapolating structural metrics between the two platforms.