Time-Sampling Errors of Earth Radiation From Satellites: Theory for Monthly Mean Albedo

The Earth Radiation Budget Experiment wide-field-of-view (WFOV) radiometers aboard the Earth Radiation Budget Satellite (ERBS) provided a 15-year record of high-quality measurements for research into the radiant energy balance of the Earth. Monthly mean maps of RSR and outgoing longwave radiation (OLR) are primary data products from these measurements. The ERBS orbit had an inclination of 57° so as to precess through all local times every 72 days. Because of limited temporal sampling, some regions were not measured sufficiently often by the WFOV radiometers to produce accurate radiation flux values for these maps. The temporal sampling of any one region is very irregular; therefore, it is necessary to consider each region in detail for each month. An analysis of the errors, which result from computing the average value of the albedo of a region over a day or month based on limited sampling, is presented. It is necessary to take into account synoptic variations and their time correlations and differences of the regions' diurnal cycle from that assumed by the time-averaging algorithms. An expression is derived for the variance of the error of the computed daily and monthly mean albedo. Temporal correlation and variability of the albedo field are specified a priori. This analysis has been used for quality assurance to evaluate the temporal sampling errors of monthly mean RSR maps computed from the measurements by the WFOV radiometers aboard the ERBS and to delete those values for which the error variance is excessive.