Recently, the CERES geostationary calibration group at NASA Langley Research Center has developed a vicarious approach of calibrating geostationary (GEO) and low earth orbit (LEO) visible sensors using invariant desert sites. This approach uses multiple years of clear-sky TOA observations from the Tropical Rainfall Measuring Mission (TRMM) Visible and Infrared Scanner (VIRS) imager to construct bidirectional reflectance distribution function (BRDF) models of the deserts. Because the TRMM satellite is in a precessing orbit with an inclination of 35°, the VIRS imager views the same sub-tropical desert site at nearly all available solar zenith angles over the course of 46 days with a variety of viewing and relative azimuth angles, thus providing an unique opportunity of evaluating the angular dependencies of the desert TOA reflectance. The VIRS sensor calibration and stability are evaluated by inter-comparing the VIRS and MODIS TOA reflectance from the same desert site under near-identical viewing and solar zenith angular conditions. The VIRS-based BRDF model is then used to predict TOA reflectance for any target sensor that is to be calibrated. In order to account for the spectral band differences between the target sensor and the VIRS reference radiance, a spectral band adjustment factor (SBAF) is derived using the reflected solar desert spectra obtained from SCIAMACHY hyper-spectral radiances. Results from this calibration technique will be presented at the conference using the Sonoran and Saharan desert sites with GEO and polar-orbiting satellites. This technique can be easily applied to calibrate the future GOES-R and JPSS sensors.