Abstract
The cloud diabatic forcing (CDF) of the atmosphere, defined as the difference between the diabatic heating in average and in clear-sky conditions is estimated from time series of simultaneous observations of diabatic heating and cloud amount. The beating is evaluated by the residual method using four-time-daily ECMWF(European Centre for Medium-Range Weather Forecasts) initialized analyses for three January and three July months. The cloud data are obtained from the International Satellite Cloud Climatology Project (ISCCP).
The CDF is dominated by release of latent heat in the storm track regions and the intertropical convergence zone (ITCZ). Over the summer continents the CDF is weakly negative, due to the decrease of sensible heating with increasing cloud amount. The error in the CDF due to random errors in the daily heating varies from less than 10 W m−2 over subtropical continents to more than 30 W m−2 over storm track regions and parts of the ITCZ. In the tropical high atmosphere (above 440 hPa) the estimated CDF is most probably too small.
The accuracy of the time-average heating is determined by comparing the distributions of the net source of atmospheric total energy (Qtot ≡ diabatic heating plus net source of latent energy) with simultaneous observations of the net radiation at the top of the atmosphere (R), obtained from the Earth Radiation Budget Experiment (ERBE). Since for time-mean conditions over land the flux of heat into the earth's surface, represented by R − Q, is very small, over land the difference between R and the estimated Qtot is a measure for the error in R and Qtot. The value of the three-monthly average of R − Qtot over large continental areas varies from 4 W m−2 over Europe in July to −56 W m−2 over Canada in January. Locally over land the value of the three-monthly average R − Qtot is generally between 25 and 50 W m−2.
