Detecting Moorland Wildfire Scars and their Persistence in the Landscape using Synthetic Aperture Radar (SAR) in the Peak District National Park, UK

The overall aim of this research is to assess the ability of SAR to detect moorland wildfire scars and their persistence in the landscape using the Peak District National Park (PDNP) in the UK as a case study. Spatially-robust data to monitor wildfire scar size and severity in UK moorlands is currently rare. Fires can burn deep into peat soils and contribute to the release of carbon dioxide into the atmosphere and may also affect the water quality of nearby streams. Information on fire extent would be useful for conservation organisations such as Moors For The Future who are trying to preserve the delicate peatland environment. Knowing the size and location of fire scars would help the Fire and Rescue Service to plan future response to moorland fires. Fire scar boundaries can be mapped in the field using Global Positioning Systems (GPS), however this is labour intensive. Indeed in the PDNP wildfire scar mapping is conducted by park rangers which provides essential ground truth data for assessing against the SAR data. Therefore this particular area provides a unique opportunity for testing an alternative SAR technique for monitoring wildfire scars in the moorland landscape. Previous research shows that SAR has been successfully applied for wildfire scar detection in other types of environments such as boreal (Bourgeau-Chavez et al, 1997) and the tropics (Huang and Siegert, 2004). This research presents some of the first results of the project which tests the capability of ERS 2; ASAR (C-band) and PALSAR (L-band) data to detect several wildfire scars from 2003 - 2008 of various spatial scales and fire severity. Some of the key areas of interest the paper will explore are at Bleaklow and the Kinder plateau. The Bleaklow peat fire of 18th April 2003 was larger (7km2) and more severe than at Kinder, which burned between 26-29th May 2008 and covered an area of 10 ha. All the wildfire scars were GPS, mapped just after the fire event. Archival time-series SAR imagery was supplied by the Landmap Service, originally sourced from the European Space Agency (ESA). C-band ASAR and ERS 2 data and L-band PALSAR data were pre-processed in SARscape 4.2. The results of this research will report on (a) the affect of land cover on backscatter and coherence results (b) the affect of precipitation on the backscatter signal (c) whether fire intensity and size affects the detectability of wildfire scars using SAR and InSAR (d) the degree that SAR and InSAR can be used to assess the persistence of wildfire scars in a UK moorland landscape.http://www.sed.manchester.ac.uk/http://mimas.ac.uk/