Fulltext

Contribution of multi-source remote sensing data to predictive mapping of plant-indicator gradients within Swiss mire habitats = Eignung von Fernerkundungsdaten aus mehreren Quellen zur räumlichen Modellierung floristischer Zeigerwertgradienten in Schweizer Moorhabitaten

Ecker, Klaus ; Waser, Lars ; Küchler, Meinrad

In: Botanica Helvetica, 2010, vol. 120, no. 1, p. 29-42

Ajouter à la liste personnelle
    Titre
    • Contribution of multi-source remote sensing data to predictive mapping of plant-indicator gradients within Swiss mire habitats
    Titre traduit ()
    • Eignung von Fernerkundungsdaten aus mehreren Quellen zur räumlichen Modellierung floristischer Zeigerwertgradienten in Schweizer Moorhabitaten
    Auteur
    • Ecker, Klaus. WSL, Swiss Federal Research Institute, Zürcherstrasse 111, 8903, Birmensdorf, Switzerland
    • Waser, Lars. WSL, Swiss Federal Research Institute, Zürcherstrasse 111, 8903, Birmensdorf, Switzerland
    • Küchler, Meinrad. WSL, Swiss Federal Research Institute, Zürcherstrasse 111, 8903, Birmensdorf, Switzerland
    Type de document
    • Postprint
    Langue
    • Anglais
    Publié dans
    • Botanica Helvetica, 2010, vol. 120, no. 1, p. 29-42. SP Birkhäuser Verlag Basel
    Autre version électronique
    Identifiant OAI-PMH
    • oai:doc.rero.ch:319116
    Summary
    Remote-sensing plays an important role in wetland monitoring on the regional and global scale. In this study we assessed the potential of different optical sensors to map floristic indicator gradients across complex mire habitats at the stand level. We compared traditional CIR photographs from RC30 cameras with modern digital ADS40 data and SPOT5 satellite images as well as fine-scale topo-structure derived from LIDAR data. We derived about 70 spectral and 30 topo-structural variables and evaluated their ability to predict the mean ecological indicator values of the vegetation across a sample of 7 mire objects. The airborne images (RC30, ADS40) and the LIDAR data were found to have a high potential for use in vegetation mapping; they explained on average 50% of the variation in observed ecological indicator values. The RC30 data slightly outperformed the less optimally collected ADS40 data. The LIDAR topo-structural variables showed equal overall predictive power as the airborne images, but they performed clearly better in predicting soil moisture, soil dispersion and light. Combining both airborne images and topo-structural data improved the predictions of all indicator values considerably. The combined use of these data sources is therefore recommended for use in fine-scale monitoring of priority habitats in nature conservation