Dr Isla Duporge
Isla’s research is focused on advancing methods using geospatial technologies to monitor wildlife and analyse movement in relation to anthropogenic risk, this is being carried out using several methods:
- Automating detection of African Elephants in very-high-resolution satellite imagery (Worldview 3 & 4) using a convolutional neural network (CNN). Testing the model against human performance and whether the algorithm can generalise to detect elephant populations in different locations.
- Conducting a wildlife census survey in southern Zambia as part of an UN-FAO project using a long-range fixed-wing drone – WingtraOne. Isla is now using audiometry – a measurement of the range and sensitivity of hearing generated for different species, cross-referencing these with sound measurements from several UAVs to find the recommended flight altitude to minimise disturbance.
- Using high-resolution satellite imagery to analyse elephant movement in relation to pastoral presence via mobile livestock enclosures. Assessing changes between 2011 & 2019 in relation to the abandonment and establishment of new settlements. This work relies on sixteen images from the Worldview-2 satellite (50cm resolution) and six years of elephant tracking data in Samburu, Kenya.
- Using the Environmental Evidence Protocol Isla conducted a Systematic Evidence Map to gain a better understanding of illegal hunting hotspots and analyse correlations with several covariates, including, proximity to roads, water bodies, human settlements and different land ownership areas.
Isla is jointly supervised by Professor David Macdonald & Dr Tiejun Wang at the Faculty of Geo-Information Science and Earth Observation, University of Twente. National Geographic are supporting her fieldwork research as an ‘Early Career Explorer’. Isla was formally a Visiting Research Scholar at the United Nations Office of Drugs and Crime, researching the use of geospatial technologies to combat wildlife crime.