The main scientific objective of DAISEX (Digital Airborne Spectrometer Experiment) was to demonstrate the retrieval of geo/biophysical variables from imaging spectrometer data. Target variables included surface temperature, Leaf Area Index (LAI), canopy biomass, leaf water content, canopy height, canopy structure and soil properties. The imaging spectrometers used for DAISEX were the DAIS-7915, HyMap and POLDER. The campaign took place during the summers of 1998, 1999 and 2000 in Barrax (Spain) and Colmar (France). A new algorithm is under development for the atmospheric correction of the hyperspectral and multiangular data acquired during this campaign. This algorithm is intended to improve the current atmospheric correction by taking into account the coupling between atmosphere and surface (including a non-Lambertian treatment of the latter). Moreover, the hyperspectral data allow to characterise in detail the absorption and the multiangular characteristics of the data allow to describe accurately the aerosol scattering. The method consists in identifying some pixels on an image with an a priori information about its BRDF and assume that the atmosphere is the same over the whole image. Applying a radiative transfer code we can reproduce the reflectance measured by the sensor by modifying the parameters describing the surface and the aerosols through an iterative process. Once the atmosphere is known the system atmosphere-surface is uncoupled and the reflectance for the whole image can be obtained.