Data of the broadband albedo averaged at the MODIS pixel, which was derived from the measurements by either pyrometer (shortwave, in the range of 200-3600 nm) or albedometer (visible in 400-700nm, near infrared in 700-2100, and shortwave in 400-2100) during the ACLOUD and AFLUX campaigns. This data is intended to validate the RTM-SciML (short for radiative transfer model - scientific machine learning model) algorithm, which uses top-of-atmosphere radiance data from MODIS as input to perform albedo retrieval for sea-ice surface. In addition to the satellite and measured data, the pixel type and surface classification retrieved using scientific machine learning models (https://tc.copernicus.org/preprints/tc-2021-397/) are also included.