The Atmospheric Radiation Measurement (ARM) Program is a major research program initiated by the Department of Energy to improve our understanding of radiative and cloud processes critical to predicting the Earth's climate and its changes. Central to this concept is the use of four to six intensively instrumented sites for long-term study and characterization of the processes of interest. The instrumentation suites will include ground-based, high-accuracy radiometers for measuring the short and longwave surface flux, as well as an extensive set of ground- and air-based instrumentation for characterizing the intervening atmospheric column. Satellite-based measurements are expected to play a very important role in providing top-of-the-atmosphere (TOA) measurements.
In an earlier study, we assessed the present and planned satellite measurement capabilities as they relate to ARM. That study focused on the ability of satellites to provide such derived physical quantities as temperature and water-vapor profiles; cloud coverage, top height and temperature; and earth radiation budgets. While such derived products can be very useful, they often involve significant processing of the direct satellite data, e.g., inversion of radiance to obtain temperature and water-vapor profiles or the conversion of radiances to fluxes to obtain radiation budgets. This processing in turn often requires both modeling and auxiliary data sets (e.g., initial-guess profiles for the inversion, etc.). Thus, unless great care is taken, there is a very real danger that such "derived" satellite data would not be an independent data set against which to compare ARM outputs.
In this study, we examine the possibility of comparing ARM outputs directly with satellite meas-urements, thereby ensuring the independence of these two important data sets. Thus we have focused on four questions: 1) What do satellites really measure and how well do they measure it? 2) What can we do about the general lack of adequate visible channel calibration? 3) What is the best way for ARM to obtain near-real-time access to this unprocessed data? 4) What is the optimum way for ARM to make use of satellite data? The answer to each of these questions is summarized below and detailed in the papers that follow.