Comparing AWIFS and MERIS images for land use-land cover mapping in Spain

As global land use - land cover mapping has become of great importance, an evaluation of the benefits and disadvantages of using satellite data with either increased spatial or spectral resolution would be adequate for the improvement of methodologies. This paper describes the comparison of AWiFS (Advanced Wide Field Sensor) and MERIS-FR (MEdium Resolution Imaging Spectrometer - Full Resolution) - based classifications of the Spanish province of Madrid. Maximum Likelihood Supervised Classification was performed using training areas based on data coming from the Spanish National Forest Inventory and from the CORINE Land Cover 2000 data bases. The classification process with AWiFS was more laborious than with MERIS- FR but the overall accuracy could be increased by 10%. For some surfaces such as deciduous forests, the high spectral resolution of MERIS- FR might be an advantage. Researchers will have to decide if the better results obtained with AWiFS compensate for the higher cost of the images and the more effortful processing.