The Medium Resolution Imaging Spectrometer, MERIS, fulfils the information gap between the current high and low spatial resolution sensors. In this respect, the use of MERIS full resolution data (300 m pixel size) has a great potential for regional and global land cover mapping. However, the spectral and temporal resolutions of MERIS (15 narrow bands and a revisit time of 2-3 days, respectively) could be further exploited in order to get land cover information at a more detailed scale than the original 300m. This study evaluates the possibilities of MERIS to retrieve sub-pixel land cover information for the main 9 land cover classes in The Netherlands. Linear spectral unmixing, with an iterative optimization of the number of endmembers, was used to classify 2 MERIS full resolution scenes. In this study, the Dutch land use database, LGN5, was used as a reference. This geographical database was first aggregated from its original 25 m pixel size to 300m and then used to identify homogeneous areas to select the endmembers. Additionally, the fractions of the different land cover types present in each MERIS pixel were recorded during the spatial aggregation so that sub-pixel classification accuracy could be assessed. Results pointed out that MERIS has a great potential to provide sub-pixel land cover information because the classification accuracies of the sub-pixel fractions were above 60%.
展开▼
