No abstract

No abstract

Radiometric cross calibration of Earth observation sensors is a crucial need to guarantee or quantify the consistency of measurements from different sensors. Twenty desert sites, historically selected, are revisited, and their radiometric profiles are described for the visible to the near-infrared spectral domain. Therefore, acquisitions by various sensors over these desert sites are collected into a dedicated database, Structure d'Accueil des Données d'Etalonnage, defined to manage operational calibrations and the required SI traceability. The cross-calibration method over desert sites is detailed. Surface reflectances are derived from measurements by a reference sensor and spectrally interpolated to derive the surface and then top-of-atmosphere reflectances for spectral bands of the sensor to calibrate. The comparison with reflectances really measured provides an estimation of the cross calibration between the two sensors. Results illustrate the efficiency of the method for various pairs of sensors among AQUA-Moderate Resolution Imaging Spectroradiometer (MODIS), Environmental Satellite-Medium Resolution Imaging Spectrometer (MERIS), Polarization and Anisotropy of Reflectance for Atmospheric Sciences Couples With Observations From a Lidar (PARASOL)-Polarization and Directionality of the Earth Reflectances (POLDER), and Satellite pour l'Observation de la Terre 5 (SPOT5)-VEGETATION. MERIS and MODIS cali-brations are found to be very consistent, with a discrepancy of 1%, which is close to the accuracy of the method. A larger bias of 3% was identified between VEGETATION-PARASOL on one hand and MERIS-MODIS on the other hand. A good consistency was found between sites, with a standard deviation of 2% for red to near-infrared bands, increasing to 4% and 6% for green and blue bands, respectively. The accuracy of the method, which is close to 1%, may also depend on the spectral bands of both sensor to calibrate and reference sensor (up to 5% in the worst case) and their corresponding geometrical matching.

No abstract

No abstract