The Kain-Fritsch convection scheme and ECMWF prognostic cloud scheme: Their Jacobians and linearization properties

Luc Fillion

Meteorological Service of Canada

In order to efficiently assimilate satellite information over cloudy regions (e.g. rain, liquid/ice water, cloud distribution), it is especially useful to examine the sensitivities of forecast model representations of these quantities. Moist-convective parameterization schemes and prognostic cloud schemes thus need to be carefully examined in terms of their sensitivities (Jacobians). This is particularly important in the context of variational data assimilation. We present results of such studies with the Kain-Fritsch (KF) moist convective parameterization (planned to be operational at the Canadian Meteorological Center (CMC) in the coming year). The degree of validity of the tangent-linear approximation of the KF scheme will be presented. Jacobians of the prognostic cloud scheme currently in operations at the European Centre for Medium Range Weather Forecasts (ECMWF) will also be discussed in details.