Daily sums of the downward solar radiation, i.e., the global radiation, have been computed from imaging data of reflected solar and emitted infrared radiation which were measured from the geostationary satellites METEOSAT I and II during the periods 1-30 June 1979 and 1-30 April 1982, respectively, over Europe and northern Africa. The method assumes basically that clouds perturb the “clear sky” radiance fields. Thus, the satellite data of reflected solar radiation are used as indicators for the presence of clouds and of their transparency for radiation scattered upward from underneath. Radiative transfer calculations have been employed to determine clouds and atmospheric transmittance for solar radiation at different zenith angles. A comparison with simultaneous measurements of the European network of pyranometers yields rms differences of 0.25 kWh m−2 (June 1979) and 0.28 kWh m−2 (April 1982), respectively, for monthly averages of daily sums. This uncertainty corresponds to ∼5–6% of the availabl... Abstract Daily sums of the downward solar radiation, i.e., the global radiation, have been computed from imaging data of reflected solar and emitted infrared radiation which were measured from the geostationary satellites METEOSAT I and II during the periods 1-30 June 1979 and 1-30 April 1982, respectively, over Europe and northern Africa. The method assumes basically that clouds perturb the “clear sky” radiance fields. Thus, the satellite data of reflected solar radiation are used as indicators for the presence of clouds and of their transparency for radiation scattered upward from underneath. Radiative transfer calculations have been employed to determine clouds and atmospheric transmittance for solar radiation at different zenith angles. A comparison with simultaneous measurements of the European network of pyranometers yields rms differences of 0.25 kWh m−2 (June 1979) and 0.28 kWh m−2 (April 1982), respectively, for monthly averages of daily sums. This uncertainty corresponds to ∼5–6% of the availabl...