A DYNAMIC-STOCHASTIC SCHEME FOR PREDICTING LARGE-SCALE PRECIPITATION AND CLOUDS

S. V. Kostrykin and I. N. Ezau

The proposed dynamic-stochastic scheme generalizes the well-known Smagorinsky scheme by taking into account heterogeneity of cloud variables inside a mesh cell. Based on simple physical assumptions, diagnostic formulas are derived for calculating the cloud amount and water content in a mesh cell as a function of total water content. A probability function is defined from empirical data. Experiments with a global medium-range numerical weather prediction model have been conducted to forecast precipitation over Russian territory. On the basis of quantitative criteria it is shown that this scheme has an advantage over the standard scheme. The model response to a change in precipitation parametrization has been studied using a climate model developed at the Institute of Numerical Mathematics, Russian Academy of Sciences (the INM RAS model). The analysis of the results obtained and verification against reanalysis data are presented.

Joomla templates by a4joomla