Coincident and collocated measurements of solar radiation from the Scanner for Radiation Budget (ScaRaB) on Meteor-3 and from towers in a boreal forest region during the Boreal Ecosystems and Atmosphere Study (BOREAS) period are used to evaluate an algorithm of Li et al. <1993a> which was developed to derive the net surface solar radiation flux from satellite measurements. The analysis shows that after correcting for mismatching between the footprints of the tower measurements and the satellite pixels, there is a substantial bias that is due to the presence of absorbing aerosols. Application of the aerosol correction term of Masuda et al. <1995> reduces the mean bias to about 5 W m-2. Cloud radiation forcing ratio R has been used in many studies to address the issue of a cloud absorption anomaly. Three methods, each with a different approach to the consideration of the effect on radiative fluxes by absorbing aerosols, are used to calculate R. Methods that account inadequately for aerosol effects when applied to the coincident and collocated tower and satellite measurements give large values of R (1.28--1.42) under conditions of heavy aerosol loading, which could be interpreted as an indication of a cloud absorption anomaly. However, application of the method that accounts best for aerosol effects gives values of R within the range 1.08~1.18 and so does not support the idea of anomalous cloud absorption.