Conference Contribution Details
Mandatory Fields
Scarrott, R.G., O' Connor, B., Dwyer, N., Cawkwell, F.
ENVIRON 2011 - Towards 2020: Environmental challenges and opportunities for the next decade
Characterising landcover using vegetation seasonality profiles determined from satellite imagery
University College Cork, Cork, Ireland
Oral Presentation
Optional Fields

The pan-European CORINE landcover classification system has limitations for national-scale applications in Ireland. CORINE is derived from satellite images acquired a maximum of twice within any one year, thereby  failing to fully indicate seasonal changes in vegetation. This has been shown to provide useful information to improve landcover characterisation. A lack of information on seasonal variability within a single class, for example different pasture management practices, can limit the use of the data as well as contribute to potential misclassification. This research explored an alternative approach to deriving information about Irish landcovers using the seasonal cycle of vegetation growth retrieved from medium resolution satellite imagery. 

 The study used the Enhanced Vegetation Index (EVI) product for 2006 from the MODIS-TERRA satellite sensor. A time-series of 250m spatial resolution, 16-day composite EVI images was processed, using time series analysis methods. The seasonal profile per pixel was first modelled to reduce the effect of cloud interference. Modelled pixels were then clustered using a divergence-guided ISODATA clustering algorithm, grouping pixels of similar seasonality together. A final Jeffries-Mathusita threshold analysis determined cluster distinctiveness.  

 Clear seasonal patterns in vegetation seasonality were detected across the island, with 5 clusters each having a distinct seasonality pattern, and 5 cluster groupings containing up to 20 similar patterns. Certain areas, predominantly within the CORINE “pastures” class, exhibited two seasonal peaks in growth, with a marked decrease in vegetation activity in June/July. Visual analysis of these areas on higher spatial resolution satellite imagery also acquired during 2006, confirmed their location within pasture areas. Land use such as silage cutting, practiced in intensively-managed grasslands, is suggested as the most likely cause of the detected double-season cycle. This finding has important implications for carbon accounting under Ireland’s commitments to the Kyoto protocol and can help improve our knowledge of grassland management in Ireland.

Environmental Protection Agency under the Climate Change Impacts on Phenology programme.