An object-oriented approach to detecting disturbance on peat soils.
John Connolly, Nicholas Holden
Bioresources Research Centre, Biosystems Engineering UCD School of Agriculture, Food Science and Veterinary Medicine, University College Dublin; firstname.lastname@example.org
Peat soils are an important part of the global climate cycle. They only cover 3-6% of land surface area but contain one third of total soil organic carbon (SOC) (Gorham 1991). Peat soil is spatially extensive covering between 13.8 and 20% of the land area (Hammond 1979, Connolly et al. 2007, Connolly and Holden In Press) Recent studies estimated the peat SOC stock to be 1071 Mt C and 1503 Mt C or 53% and 62% of the total national SOC stock (Tomlinson 2005, Eaton et al. 2008). Disturbance, both natural and anthropogenic, may however enhance C loss and impact on the ability of the resource to sequester and store C (Turetsky et al. 2002). Peat soil disturbances in Ireland include drainage, extraction, forestry and burning and others. Each of these disturbances may be characterised by geometric patterns, straight lines and right angles. For example drainage ditches, on both the industrial and non-industrial peat landscape, can be clearly seen in high resolution imagery (McGovern et al. 2000). In an object-oriented approach these geometric patterns can be clearly demarcated. This work is using an object-oriented approach (Feature Analyst in ArcGIS) to classify and determine the extent and severity of peatland disturbance in Ireland
Access presentation here (5 meg)
Return to Presentation Programme Page