Evaluating burning in the Cairngorms National Park in relation to the Muirburn Code by using GIS
MetadataShow full item record
While heather burning is seen as a necessary environmental practice in heathland management, there are concerns that it’s over use is having an adverse effect on threatened rare ecosystems. This study aims to assess the successfulness of GIS techniques in detecting and measuring the extent of heather burning undertaken in the Cairngorms National Park, in conjunction with the Muirburn Code. An initial sensor test compared the use of two sets of low resolution multispectral sensor data from Sentinel and Landsat, plus a non-multispectral high resolution aerial image. This resulted in the Sentinel data, although not most accurate overall, being the best at identifying areas of burn. Further testing of different supervised classification algorithms using the same training data, identified ML as the most appropriate algorithm to use in the study from the algorithms tested. The findings led to Sentinel data and the ML algorithm being used for the main research of the study, identifying burning undertaken in relation to altitudes, slopes and soil parameters highlighted in the Muirburn Code as non-burn areas, between the years 2016-2018. This identified an overall increase in burning within the park but a decrease in the number of burns undertaken outside the parameters of the Muirburn Code, indicating an increase in the adoption of the Code. Whilst the data collected showed a relatively high burn accuracy of above 0.7 in the spring and above 0.8 in the summer this could be further improved upon by higher resolution data and the use of more infra-red bands.