Abstract
The usefulness of the Moon as a light source for on-orbit calibration derives in large part from the exceptional stability of the lunar surface reflectance, along with the absence of intervening atmosphere between the lunar source and a space-based sensor. These properties also are key to using the Moon for inter-calibration of sensors that have viewed it, as the need for near-simultaneous observations is removed. However, since no two lunar observations will have exactly the same illumination and viewing geometry, a lunar reflectance model is required to account for BRDF-related differences in the lunar brightness for different observations. The level of uncertainty achievable for lunar inter-calibration depends on the relative accuracy of the lunar model, i.e. the capability to accurately predict the variations in the lunar brightness for different observation geometries, mostly irrespective of the model's absolute radiometric accuracy. An example will be given of two Moon observations acquired by Aqua-MODIS and SNPP-VIIRS having nearly the same geometry, but taken more than 11 years apart. Cross-comparison is achieved by referencing both against the lunar spectral irradiance model developed by the U.S. Geological Survey's Lunar Calibration project. Applying lunar inter-calibration to meteorological imagers in geostationary orbit has the potential to place onto a common, stable radiometric scale the extensive collections of Earth observation data taken from around the world. This has significant implications for developing long-term climate records that extend decades into the past.
Using the Moon as a Common Reference for Inter-calibration
The usefulness of the Moon as a light source for on-orbit calibration derives in large part from the exceptional stability of the lunar surface reflectance, along with the absence of intervening atmosphere between the lunar source and a space-based sensor. These properties also are key to using the Moon for inter-calibration of sensors that have viewed it, as the need for near-simultaneous observations is removed. However, since no two lunar observations will have exactly the same illumination and viewing geometry, a lunar reflectance model is required to account for BRDF-related differences in the lunar brightness for different observations. The level of uncertainty achievable for lunar inter-calibration depends on the relative accuracy of the lunar model, i.e. the capability to accurately predict the variations in the lunar brightness for different observation geometries, mostly irrespective of the model's absolute radiometric accuracy. An example will be given of two Moon observations acquired by Aqua-MODIS and SNPP-VIIRS having nearly the same geometry, but taken more than 11 years apart. Cross-comparison is achieved by referencing both against the lunar spectral irradiance model developed by the U.S. Geological Survey's Lunar Calibration project. Applying lunar inter-calibration to meteorological imagers in geostationary orbit has the potential to place onto a common, stable radiometric scale the extensive collections of Earth observation data taken from around the world. This has significant implications for developing long-term climate records that extend decades into the past.