Visibility parameterization from microphysical observations for warm fog conditions and its application to the Canadian MC2 model

The droplet number concentrations in operational forecast models are presently fixed or obtained from an assumed particle size spectra. In this work, in-situ observations collected within the low level clouds during Radiation and Aerosols Cloud Experiment (RACE) over Eastern Canada during summer and fall of 1995 were used to parameterize visibility (Vis) as a function of both droplet number concentration (Nd) and liquid water content (LWC). The main measurements used in the analysis were LWC, Nd, and temperature (T) obtained from hot-wire probes, FSSP spectra, and Rosemount probe, correspondingly. The Vis is calculated from the FSSP spectra. The results showed that Vis is nonlinearly related to both LWC and Nd. Comparisons between newly derived parameterizations and the ones already in use as a function of LWC suggested that if models can obtain total Nd (diagnostically) and LWC at each time step from the bulk physical parameterizations, Vis can then be obtained for warm fog conditions at T>0°C. Using outputs from the Canadian GEM forecast model, being tested with a new multi-moment bulk microphysical scheme, the new Vis parameterization will be tested and the results will be summarized. Preliminary results suggested that better forecast of the fog visibility is strongly related to model based LWC and Nd, and that, because Nd is not a prognostic variable, Vis calculations can only be obtained using parameterized equations e.g. Nd versus T, or assumed size distributions.