A method to estimate monthly cloud conditions (total fractional cloud cover) from multispectral satellite data is described. The operational cloud classification scheme SCANDIA (the SMHI Cloud ANalysis model using DIgital AVHRR data), based on high resolution imagery from the polar orbiting NOAA satellites, is used to produce monthly cloud frequencies for all months of 1993. The annual mean is computed and the diurnal variation of cloudiness is investigated for June and December. Cloud analyses are made for an area covering the Nordic countries with a horizontal resolution of four km
Comparisons with existing surface observations show very good agreement, especially in the summer half of the year. some problems are indicated in the winter season when a minor underestimation of cloudiness is noticed. The underestimation is mainly due to the non-separability of low-level water clouds from cloud-free areas at very low sun elevations. Despite these problems, general cloud patterns are well described also in cold winter situations. Improvements of the method are discussed and an enlargement of the analysis area is envisaged.
The method is proposed as a valuable tool for local and regional monitoring of the cloud climatology. Comparisons with forecasted cloudiness from atmospheric models are suggested as well as special studies of cloud conditions in the Polar areas.