Radiative Closure Experiments at a Cloud-Free Desert Site, Nevada, as Part of MISR Algorithm Validation

Radiative closure experiments involving a comparison between surface-measured spectral irradiance anti the surface irradiance calculated according to a radiative transfer code at a desert site in Nevada under clear skies, yield the result that agreement between the two requires presence of an absorbing aerosol component with an imaginary refractive index equal to 0.03 and a 50:50 mix by optical depth of small and large particles with log-normal size distributions. The mode radius of the small particle distribution is 0.03 ~m and that of the large 0.5 Lm. The same aerosol model can be used for both day-to-day fits in one campaign year, and also between campaigns in different years. The high imaginary index required for the fits suggests presence of urban-type particles in the aerosol, but an alternative under study is to rely on absorbing iron oxide components in a dust fraction to account for some of the absorption. INTRODUCTION As part of the algorithm validation phase of preflight activities for the Multiangle Imaging SpectroRadiometer (MISR) and in conjunction with development of a so-called vicarious method of on-orbit absolute instrument radiometric calibration, a series of radiative closure experiments have been carried out under cloud-free conditions at a dry lake site (Lunar Lake) in central Nevada. Vicarious calibration methods are entirely independent of the pre-launch laboratory and postlaunch onboard methods, except for tracing of field instrument calibrations and calibration panels to NIST standards, and are consequently expected to play an important role in determination and maintenance of the MISR radiometric calibration over time. Vicarious calibration of satellite instruments involves the use of homogeneous naturally occurring ground targets that are large enough to encompass a number of satellite pixels. Dry lakes at moderate altitude in remote places of the southwestern United States have often been used for this purpose (see for example [ l]), to minimize atmospheric scattering and absorption and water vapor interference. Cloud-free stable atmospheric conditions are also sought. The top-of-atmosphere (TOA) radiance at the sensor is calculated with a radiative transfer code (RTC) utilizing measured target reflectance over the area of several pixels together with atmospheric spectral optical depth, single scattering phase function and water vapor absorption. The TOA radiance so-obtained is assigned to the digital numbers measured by the satellite to obtain a reevaluation of the radiometric calibration coefficients for the sensor. Up to and including the present time, never in the field do direct observations of the composition or size distribution of the atmospheric aerosol burden accompany the other field measure-