WSM 6-class graupel scheme

WSM6 is a single-moment microphysics scheme in WRF carrying six hydrometeor classes: water vapour, cloud water, rain, cloud ice, snow, and graupel. It predicts one moment — mass mixing ratio — for each, with the particle size distribution’s shape fixed by assumption.

It is probably the most common microphysics choice in regional modelling, and the reason is that it sits at a defensible balance point rather than being best at anything.

Why graupel earns its class

Graupel is the class that distinguishes WSM6 from the five-class WSM5, and for convective work it matters. Graupel forms when supercooled water freezes onto ice particles in an updraft, and it falls fast — much faster than snow. In deep convection this is a substantial part of how heavy surface precipitation is produced: mass carried aloft, rimed, and returned quickly rather than drifting.

Without a graupel class, a scheme has to route that mass through snow, which falls slowly and spreads out. The result is precipitation that is too light and too widespread, with too little intensity in the core — precisely the error that matters for anything hazard-related.

For tropical maritime convection, which is what the Galapagos work involves, that argues for including graupel. The tropical setup uses WSM6 for this reason.

The single-moment limitation

Predicting mass but not number concentration means the particle size distribution’s shape is prescribed. Since fall speed and evaporation rate both depend strongly on particle size, a fixed distribution fixes those too, and the assumed distribution came from observations of particular cloud types.

The practical consequence is that WSM6 cannot represent situations where the size distribution genuinely differs from its assumption — heavily polluted conditions producing many small droplets, for instance. Double-moment schemes can, at roughly twice the cost and with additional parameters that are themselves poorly constrained.

The choice is a real judgment rather than an obvious upgrade. Double-moment buys physical realism it may not be able to exploit; single-moment buys runtime that lets you run more configurations, which for a sensitivity study is often the better trade.

See also: microphysics schemes, convective rain, the tropical WRF setup, and the physics map.