A Semi-Lagrangian Advection Algorithm for Falling Raindrops in Atwo-Moment Microphysics Schemes

A semi-Lagrangian algorithm (SLA) is implemented in NOAA's Global Forecast System (GFS) forsimulating raindrop sedimentation in a double-moment microphysics schemes. This SLA includesa significant improvement to its predecessor for single-moment raindrop sedimentation. It isnumerically stable and mass-conserving when used to sediment raindrops in double-momentmicrophysics schemes. Numerical results from an idealized single-column model show that theSLA overcomes an issue of mass accumulation at the cloud bottom in the case of the Eulerianalgorithm for raindrop sedimentation, which is due to the assumption of constant terminalvelocity within a time step of sedimentation. The results from the single-column model also showthat the time step in the SLA can be 10 times greater than that in the Eulerian algorithm forsedimentation. Further numerical experiments using NOAA's GFS show that using the SLAmitigates the numerical instability problem associated with a newly-implemented double-momentmicrophysics scheme in the GFS.