Joint Centre for Mesoscale Meteorology (JCMM)
JCMM INTERNAL REPORT NO.59
Investigation of different representations of rain in a 1-D evaporation model
by Hugh Swann
The bulk-microphysics scheme previously used in the UK Met. Office Large Eddy Model has single-moment bulk-water representation of precipitating particles, in common with most other cloud resolving models (eg Flateau et al 1989, Lin et al 1983). In such schemes, precipitating particles are represented by just one variable (q, the mass-mixing ratio) so the number concentration and the mean drop diameter both increase monotonically as a function of q. More recently, however, a double-moment bulk-water scheme has been implemented, which uses two variables to represent precipitating particles (the mass-mixing ratio and n, the particle number concentration). This reduces the number of adjustable coefficients and allows some processes to be modelled in a more realistic way than is possible with the simpler schemes because the number concentration (and hence the mean diameter) is allowed to vary independently of the mass-mixing ratio.
One feature of cloud simulations that the double-moment scheme is expected to model more accurately is the reduction of the number of small drops compared to larger drops in the evaporation region of rain. The single-moment schemes, especially when an exponential drop size distribution is assumed, implies generation of small drops at a rate faster than they would be generated through the evaporation of large drops, as found by Clough and Franks (1991). A simple 1-D evaporation model with a 'passive' atmosphere is used to model rain falling through subsaturated air to the surface, causing rainfall rates of up to 10mm/h.
Three different representations of rain are compared, a single-moment scheme, a double-moment scheme and an explicit scheme. The shape of the initial drop size distribution is assumed to be either an exponential or a generalised gamma distribution (the explicit scheme is able to deviate from this initial shape, whereas the bulk water schemes can not). The explicit scheme uses a large number of variables each representing a size bin of rain drops. this method is too computationally expensive to be used in a 3-D model, but can be used in this 1-D model to produce an exact solution to the precipitation and evaporating equations, for evaluating the bulk-water approximations.