Global cloud cover: comparison of observations, validation of general circulation models and simulation of new observation technologies.

Summary Cloud cover is one of the major parameters in the study of the climate system. The lack of knowledge about cloud cover represents one of the main uncertainties in climate modeling and climate change prediction. In particular, the vertical distribution of clouds in the atmosphere is a determining factor of the earth's radiation balance. In order to study this vertical structure, we use two sets of data, surface observations (warren et al. data) and satellite observations (isccp data) that we start by comparing. We show the complementarity of these two data sets. If a good coherence is found for the detection of the total cloud cover, significant differences appear when we focus on a particular type of cloud (low clouds for example). The validation of the cloud cover simulated by a general circulation model (GCM) then becomes delicate. Considering the great heterogeneity of the cloud cover, it is essential to compare not only the simulated and observed mean radiation balance but also the cloud characteristics (top pressure, optical thickness) determined at a smaller spatial scale than the model and to compare them with the same parameters derived from the observation. A method of comparing the cloud cover simulated by a mcg with isccp satellite data has been developed. We use this method to validate physical parameterizations, wet convection and boundary layer scheme, in a mcg. However, this type of comparison does not allow a complete validation of the vertical distribution of the cloud cover. Indeed, the isccp climatology provides a description in terms of radiatively equivalent monolayer clouds at the top of the atmosphere to the actual cloud distribution. This representation, in the case of multilayered clouds or clouds with strong vertical development, is not representative of the true vertical structure of the cloud layers. It is therefore important to improve in the future the measurement techniques and to consider other instruments to be embarked in space to solve in particular the problem of detection of cloud systems organized in multilayers. It is in this perspective that we have tried to show the interest of embarking a backscattering lidar in space by coupling it to a passive radiometry instrument and to measure its limits. From the cloud cover simulated by the mcg, we have performed simulations of the restitution of this cloud cover from lidar and radiometric measurements. We have tested different hypotheses on some parameters such as the signal to noise ratio and sampling in order to determine the optimal conditions for the use of lidar. The lidar will allow a better study of the vertical distribution of clouds in the atmosphere. It will allow to improve the restitution of the altitude of the top of semi-transparent clouds such as cirrus and to highlight the multi-layer systems including cirrus as well as the systems of split cloud cover.