Wetland monitoring using classification trees and SPOT - 5 seasonal time series

1 Multi-season reflectance data from radiometrically and geometrically corrected 2 multispectral SPOT-5 images of 10-m resolution were combined with thorough field campaigns 3 and land cover digitizing using a binary classification tree algorithm to estimate the area of 4 marshes covered with common reeds (Phragmites australis) and submerged macrophytes 5 (Potamogeton pectinatus, P. pusillus, Myriophyllum spicatum, Ruppia maritima, Chara sp.) over 6 an area of 145 000 ha. Accuracy of these models was estimated by cross-validation and by 7 calculating the percentage of correctly classified pixels on the resulting maps. Robustness of this 8 approach was assessed by applying these models to an independent set of images using 9 independent field data for validation. Biophysical parameters of both habitat types were used to 10 interpret the misclassifications. The resulting trees provided a cross-validation accuracy of 98.7% 11 for common reed and 97.4% for submerged macrophytes. Variables discriminating reed marshes 12 from other land covers were the difference in the near-infrared band between March and June, the 13 Optimized Soil Adjusted Vegetation Index of December, and the Normalized Difference Water 14 Index (NDWI) of September. Submerged macrophyte beds were discriminated with the 15 shortwave-infrared band of December, the NDWI of September, the red band of September and 16 the Simple Ratio index of March. Mapping validations provided accuracies of 98.6% (2005) and 17 98.1% (2006) for common reed, and 86.7% (2005) and 85.9% (2006) for submerged 18 macrophytes. The combination of multispectral and multisesasonal satellite data thus 19 discriminated these wetland vegetation types efficiently. Misclassifications were partly explained 20 by digitizing inaccuracies, and were not related to biophysical parameters for reedbeds. The 21 classification accuracy of submerged macrophytes was influenced by the proportion of plants 22 showing on the water surface, percent cover of submerged species, water turbidity, and salinity. 23 Classification trees applied to time series of SPOT-5 images appear as a powerful and reliable 24 ha l-0 06 92 53 7, v er si on 1 16 M ay 2 01 2