Cloudiness parameterization and verification in a large-scale atmospheric model

Abstract

This paper addresses the problem of predicting cloud cover and its radiative impact in a large-scale atmospheric model. A convective and stratiform condensation scheme including cloud water content as a predictive variable is implemented in the Canadian global spectral model. An important aspect of the scheme is that the cloud amount estimation is a part of the condensation scheme and is a key element in the sub-grid scale stratiform condensation parameterization. The cloud cover from the scheme is verified quantitatively using satellite data. The dependence of the grid-scale relative humidity threshold on the horizontal and vertical resolutions is examined. The possibility of parameterizing stratiform clouds as vertically sub-grid clouds and its verification is investigated. It is shown that the total cloud cover is better estimated as the sum of separate estimates of convective and stratiform cloudiness within the framework of the condensation processes parameterized in the model. The convective cloud cover is found to be very important to the radiative budget. An improvement in the model forecast, hydrological balance and cloudiness prediction is noticed when the stratiform relative humidity threshold decreases with height. The study also presents a new 3-dimensional view of the cloudiness estimated by the original scheme and provides a simple vertical and horizontal sub-grid scale cloud cover parameterization. Vertically sub-grid stratiform clouds combined with horizontally sub-grid convective clouds provide a remarkable improvement in the estimation of total cloud cover.