UAV to Inform Restoration: A Case Study From a California Tidal Marsh

• Main Authors: John Haskins, Charlie Endris, Alexandra S. Thomsen, Fuller Gerbl, Monique C. Fountain, Kerstin Wasson
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2021-04-01
• Series: Frontiers in Environmental Science
• Subjects: remote sensing; tidal marsh; restoration; topography; digital elevation model; image classification
• Online Access: https://www.frontiersin.org/articles/10.3389/fenvs.2021.642906/full

Monitoring of environmental restoration is essential to communicate progress and improve outcomes of current and future projects, but is typically done in a very limited capacity due to budget and personnel constraints. Unoccupied aerial vehicles (UAVs) have been used in a variety of natural and human-influenced environments and have been found to be time- and cost-efficient, but have not yet been widely applied to restoration contexts. In this study, we evaluated the utility of UAVs as an innovative tool for monitoring tidal marsh restoration. We first optimized methods for creating high-resolution orthomosaics and Structure from Motion digital elevation models from UAV imagery by conducting experiments to determine an optimal density of ground control points (GCPs) and flight altitude for UAV monitoring of topography and new vegetation. We used elevation models and raw and classified orthomosaics before, during, and after construction of the restoration site to communicate with various audiences and inform adaptive management. We found that we could achieve 1.1 cm vertical accuracy in our elevation models using 2.1 GCPs per hectare at a flight altitude of 50 m. A lower flight altitude of 30 m was more ideal for capturing patchy early plant cover while still being efficient enough to cover the entire 25-hectare site. UAV products were valuable for several monitoring applications, including calculating the volume of soil moved during construction, tracking whether elevation targets were achieved, quantifying and examining the patterns of vegetation development, and monitoring topographic change including subsidence, erosion, and creek development. We found UAV monitoring advantageous for the ability to survey areas difficult to access on foot, capture spatial variation, tailor timing of data collection to research needs, and collect a large amount of accurate data rapidly at relatively low cost, though with some compromise in detail compared with field monitoring. In summary, we found that UAV data informed the planning, implementation and monitoring phases of a major landscape restoration project and could be valuable for restoration in many habitats.