A multi-resolution satellite imagery approach for large area mapping of ericaceous shrubs in Northern Quebec, Canada Olivier R. van Lier a , Richard A. Fournier a, *, Robert L. Bradley b , Nelson Thiffault b,c a Centre d’Applications et de Recherche en Te ´le ´de ´tection, De ´partement de ge ´omatique applique ´e, Universite ´ de Sherbrooke, Sherbrooke, QC, Canada J1K 2R1 b Centre d’E ´ tude de la Foreˆt, De ´partement de biologie, Universite ´ de Sherbrooke, Sherbrooke, QC, Canada J1K 2R1 c Direction de la recherche forestie `re, Ministe `re des Ressources Naturelles et de la Faune du Que ´bec, Que ´bec, QC, Canada G1P 3W8 1. Introduction Forests cover nearly half the total geographic area of Canada (402.1 m ha) and play a vital role in its socio-economic develop- ment. Forest management is predicated, therefore, on the sustainability of this resource for future generations. In north- eastern Canada, foresters are increasingly preoccupied by the growth check of black spruce (Picea mariana (Mill.) BSP) seedlings induced by ericaceous shrubs such as Kalmia angustifolia L., Rhododendron groenlandicum (Oeder) Kron & Judd, and Vaccinium spp. that can quickly invade sites following harvesting or wildfires. These shrubs interfere with black spruce growth through direct competition for resources (e.g. Thiffault et al., 2004) and by modifying soil properties such as pH, litter decomposition and nutrient mineralization rates (e.g. Joanisse et al., 2007). The transformation of productive forest stands into heaths poses a threat to the sustainability of Canada’s forest sector; adequate monitoring tools must be developed to assess the magnitude of the problem. It is thus essential that we develop a method to map ericaceous swards, as this would give us an indication of conditions favoring their encroachment on disturbed forest sites. Remote sensing technologies offer various options which are compatible to detection, mapping, and monitoring of invasive species (Wang, 1994; Joshi et al., 2004). Mapping the presence of ericaceous shrubs on regenerating forest sites has already been achieved with 96% overall accuracy through the interpretation of high resolution Compact Airborne Spectrographic Imager (CASI, 2.5 m spatial resolution) images (Franklin et al., 1994a). Likewise, Franklin et al. (1997) were able to distinguish three different classes of ericaceous shrub cover (low, moderate, and heavy) on disturbed sites with high accuracies (87–99%) using high spatial resolution imagery (CASI). The use of very high spatial resolution imagery is, however, costly and time-consuming for large-scale detection purposes. K. angustifolia can be detected with lower resolution images (i.e. Landsat-TM), but with a concomitant lower overall accuracy of 82% (Franklin et al., 1994b). The studies from Franklin et al. were limited to the detection of these shrubs at a local scale mapped on one image and confined to disturbed sites. International Journal of Applied Earth Observation and Geoinformation 11 (2009) 334–343 ARTICLE INFO Article history: Received 8 August 2008 Accepted 22 May 2009 Keywords: Multi-resolution Object-oriented classification Ericaceous shrubs Forestry ABSTRACT Invasive ericaceous shrubs (e.g. Kalmia angustifolia, Rhododendron groenlandicum, Vaccinium spp.) may reduce the regeneration and early growth of black spruce (Picea mariana) seedlings, the most economically important boreal tree species in Quebec. Our study focused, therefore, on developing a method for mapping ericaceous shrubs from satellite images. The method integrates very high resolution satellite imagery (IKONOS) to guide classifiers applied to medium resolution satellite imagery (Landsat-TM). An object-oriented image classification approach was applied using Definiens eCognition software. An independent ground survey revealed 80% accuracy at the very high spatial resolution. We found that the partial use (70%) of classified polygons derived from the IKONOS images were an effective way to guide classification algorithms applied to the Landsat-TM imagery. The results of this latter classification (78.4% overall accuracy) were assessed by the remaining portion (30%) of unused very high resolution classified polygons. We further validated our method (65.5% overall accuracy) by assessing the correspondence of an ericaceous cover classification scheme done with a Landsat-TM image and results of our ground survey using an independent set of 275 sample plots. Discrimination of ericaceous shrub cover from other land cover types was achieved with precision at both spatial resolutions with producer accuracies of 87.7% and 79.4% from IKONOS and Landsat, respectively. The method is weaker for areas with sparse cover of ericaceous shrubs or dense tree cover. Our method is adapted, therefore, for mapping the spatial distribution of ericaceous shrubs and is compatible with existing forest stand maps. ß 2009 Elsevier B.V. All rights reserved. * Corresponding author. Tel.: +1 819 821 8000x63209; fax: +1 819 821 7944. E-mail address: Richard.Fournier@USherbrooke.ca (R.A. Fournier). Contents lists available at ScienceDirect International Journal of Applied Earth Observation and Geoinformation journal homepage: www.elsevier.com/locate/jag 0303-2434/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jag.2009.05.003