Use of Unmanned Aircraft Systems to Delineate Fine-Scale Wetland Vegetation Communities

Remote sensing of wetlands has primarily focused on delineating wetlands within a non-wetland matrix. However, within-wetland changes are arguably just as important as loss of wetland area, particularly in a time of accelerated climate change. Remote sensing is a critical source of data for ecological models that explain and predict landscape changes, but data specifications, including spatial and temporal resolution, must be appropriate for applications. Unmanned Aircraft Systems (UASs) can be used to collect fine spatial resolution data with a temporal resolution more tailored to application need, instead of satellite orbital times or flight schedules. We used data collected from an UAS to acquire true color data within a wetland landscape and tested our ability to automatically classify plant communities from fine-resolution data. Classification accuracies were low for certain classes when nine vegetation communities were mapped, but the overall accuracy was on par with other remote sensing analyses. We demonstrate that classification data derived from UAS fine-resolution imagery is reasonably accurate and discuss the benefits and challenges of using UAS for wetland mapping.