Simultaneous surface and top of atmosphere radiative flux measurements and their applicability to evaluation of climate simulations R. P. Allan, A. Slingo, N. Clerbaux (RMIB) The evaluation of climate models requires not only measurements of radiative fluxes at the top of the atmosphere but also at the surface. In the present study, surface radiation measurements at Carpentras, France, are combined with top of atmosphere fluxes derived from Meteosat for April 1999. This enables the calculation of atmospheric flux divergence and the adequate representation of the diurnal cycle. Times of clear-sky are determined using the shortwave and longwave radiation measurements, based on previous analysis of surface radiation budget data, and evaluated using Meteosat cloud amount during the day-time. Monthly-mean diurnal cycles of the radiation budget are constructed taking into account of the following: grid-box to point-measure inconsistencies, specification of surface albedo, variation of the solar declination throughout the month. The sensitivity of the monthly diurnal clear-sky fluxes to the clear-sky identification method are assessed. The inferred atmospheric anomalous solar absorption by clouds is found to be highly sensitive to the specification of surface albedo. Applicability of the monthly processing to future Geostationary Earth Radiation Budget (GERB) data and its use in evaluation of climate models is discussed. [4.2 - ATMOSPHERIC RADIATION, Earth's surface radiation from satellites, surface observations and models; Conv: Atsumu Ohmura] ------------------------------------------------------- A diagnostic analysis of atmospheric moisture and clear-sky radiative feedback in the Hadley Centre and GFDL climate models R. P. Allan, V. Ramaswamy, A. Slingo The interannual variability of water vapour and temperature profiles are diagnosed from the Hadley Centre and Geophysical Fluid Dynamics Laboratory (GFDL) climate models, both of which are forced by observed sea surface temperatures. Both models produce a similar sensitivity of clear-sky outgoing longwave radiation to surface temperature of about 2~$Wm^{-2}K^{-1}$, indicating a consistent and positive clear-sky radiative feedback. However, differences between changes in the temperature lapse-rate and the height dependence of moisture fluctuations suggest contrasting causal mechanisms. The GFDL model appears to give a weaker water vapour feedback (i.e. changes in specific humidity). This is counteracted by a smaller upper tropospheric temperature response to surface warming which implies a compensating positive lapse-rate feedback. The diagnostic analysis is enhanced by comparing the moisture variations with those resulting from satellite observations and the ECMWF re-analysis (ERA). Moisture variations in ERA are larger than those measured by satellite and simulated by the models, most especially in the lower troposphere. This appears to be attributable in part to spurious lower tropospheric moisture trends in ERA, which reduces the reliability of this particular reanalysis for the validation of feedbacks. It is essential to test the validation product (analyses/observations) on a continuous basis in order to ensure the reliability of the long time-series. A method is suggested to check and improve upon re-analysis results with regards to moisture profile changes using observed column integrated water vapour. [4.3 ATMOSPHERIC RADIATION - Radiative Forcing and Climate Feedbacks, Conv: Graeme Stephens] ------------------------------------------------------- Changes in tropical OLR - a missing mode of variability in Climate Models? R. P. Allan, A. Slingo, B. Wielicki & Co-authors. Top of atmosphere measurements of radiative fluxes are presented from scanning instruments (ERBS, ScaRaB and CERES) and the non-scanning wide-field-of-view ERBS instrument (WFOV). A marked increase in low-latitude outgoing longwave radiation (OLR) of order 3 Wm-2 during the early 1990's in the WFOV data is also apparent in the independent scanner data. Analysis of callibration issues have previously further indicated the reliability of such trends in the radiation budget. Also, the stability of clear-sky OLR in the scanner record indicates that the apparent increase in OLR is due to decreases in cloud amount. There is no strong supporting evidence from cloud datasets, although ISCCP cloud amount is shown to decrease during the early 1990's and SAGE II data indicates a reduction in high cloud altitude during this period. Corresponding time-series from the Hadley Centre Climate model (HadAM3) forced by observed sea surface temperatures do not show such an increase in OLR, even when forced by an array of known climate forcings (e.g. greenhouse gas, aerosol, ozone and natural). Interannual variations in Longwave Cloud Radiative Forcing in the model are small indicating that, if the satellite observations are reliable, simulated tropic-wide cloud amount shows an unrealistically low variability. EOF analysis of tropical ocean OLR shows that second to the dominating El Nino pattern, captured both by observations and models, is the increase in OLR, captured only by the observations, which appears to dominate in the subtropics. Links with changes in temperature lapse-rate and climate feedbacks are discussed. [2.2 CLIMATE VARIABILITY, Modes of Current Climate Variability; Conv: Michael MacCracken... or in Atmos. Radiation Session (4.1 or 4.3)]