One objective of the Surface Heat Budget of the Arctic (SHEBA) program is to develop improvements to Arctic global climate model simulations by thoroughly measuring atmospheric radiative, dynamic, and microphysical properties at a well-instrumented ice camp over the course of a year. To use such a suite of measurements for GCM validation and improvement, it is necessary to have the ability to relate the measurements made at a single point on the Arctic Ocean surface to larger geographical areas comparable to a GCM grid cell. Satellite data offers the most comprehensive resource for the GCM scaling problem in that NOAA polar orbiter images covering the SHEBA area could be collected for nearly every day of the field program. Using the TeraScan (SeaSpace, Inc.) satellite tracking system aboard the Canadian Coast Guard Ship Des Groseilliers, between ten and fifteen NOAA polar orbiter overpasses were tracked and archived every day throughout most of the duration of SHEBA. Here we analyze a time series of Advanced Very High Resolution Radiometer (AVHRR) images covering SHEBA during May 1998. A discrete-ordinates radiative transfer model is used to retrieve cloud optical depth and particle effective radius for each pixel in these images, and from these retrievals the surface shortwave and longwave radiaton budget is mapped over a geographic area comparable to a GCM grid cell. Area-averaged surface radiative fluxes thus derived are compared with the SHEBA surface radiation measurements, to demonstrate the importance of spatial inhomogeneity in cloud and surface properties, even when stratiform cloud cover appears uniform in texture