The main objective of this study is to define relevant data to improve land-cover maps of the Arcachon bay. With this aim, the authors introduce the use of radar data by assessing their potential for mapping intertidal flats and salt-marshes. In addition, the potential of using images acquired at different seasons is evaluated in order to derive the wavebands useful to test a multi-temporal classification strategy. For this purpose, hyperspectral field measurements carried out seasonally are analyzed. Then, the most discriminating SAR and optical wavebands are concatenated and the resulting muti-sensor multi-temporal composite images are classified. The validation of the resulting maps is performed on a regularly surveyed site. Image data selection, map production and validation are detailed in the present poster.