Invasive plant species spread quickly across river catchments, where they threaten biodiversity, ecosystem functioning, and landscape character. Despite worldwide eradication efforts impact grows. New techniques are needed to provide information on spatial structure of invasions, needed for monitoring and control. Remote sensing permits rapid mapping of large, remote areas, e.g. up river catchments where access is limited. Accurate maps are essential to effective conservation management (Lyons et al., 2011), but traditional methods rely heavily on expensive time-consuming ground sampling (Lyons et al., 2011; Valle et al., 2015). New high-resolution multi-spectral satellite platforms offer novel opportunities, while Unmanned Aerial Vehicles (UAV’s) offer ultra-high resolution and are a cost effective method of acquiring such images over large areas (Colomina and Molina, 2014; Cano et al., 2017).
The project aims to assess detectability of three priority invasive species using three remotely sensed datasets with differing spectral, spatial and temporal resolutions, permitting assessment of a) feasibility and b) effectiveness of detecting each species, at multiple scales to inform prioritisation of appropriate technologies into a cost-effective, multi-scale monitoring plan. The Wear Invasive Non-Native Species (WINNS) ground-truth data (Atkinson RIPPLE2019) will be used in combination with purchased satellite data (1m); government aerial data (25cm), and multi-spectral drone data (sub-10cm). Species classification and mapping using these data will be piloted using pixel and object-based approaches, and accuracy, feasibility, appropriate scale-of-use and cost assessed. This will provide well-tested remote-sensing approaches to mapping for three proposed invasive species (Heracleum mantegazzianum; Reynoutria japonica; and Impatiens glandulifera, providing the basis for a multi-scale remote-sensing strategy to direct management, while optimising costs. Remote sensing skills are required for this PhD, with an understanding of new cost-effective platforms, software and analytical techniques, but a formal development programme will be devised (e.g. OBIA, image interpretation).
This project is part of the ONE Planet DTP. Find out more here: View Website