About KNMI Colloquia and workshops Staff Publications Contact Sitemap Homepage Research KODAC: KNMI Operational Data Centre
Introduction
Previous shallow cumulus intercomparison studies by the Boundary Layer Cloud working group of GCSS have shown that the LES community is successful in simulating the characteristic features of shallow cumulus convection, such as the thermodynamic mean state, cloud-base mass fluxes and cloud cover (Stevens et al., 2001; Brown et al., 2002, Siebesma et al., 2003). After simulating stratocumulus in the last two workshops, the latter with microphysics included, the time is ready for a precipitating cumulus case as well! The here proposed RICO Precipitating Shallow Cumulus Case is based on data obtained during the successful 'Rain In Cumulus over the Ocean' (RICO) measurement campaign, which took place in the vicinity of the Caribbean islands Antigua and Barbuda during Dec 2004 - Jan 2005.

Originally it was proposed to simulate a single day of the RICO campaign, the 11th of Jan, for which an almost complete suite of measurement platforms is available. However, the large scale meteorological forcings on this day turned out to be rather complicated, needing more time than available for analysis. Further, the soundings and dropsondes showed only a weak inversion with consequently high relative humidities (up to 70 %) in the inversion layer and just above. This makes a simulation very sensitive to specified forcings and initialization.

Instead, we turned towards a composite case, based on a three week rather undisturbed period with typical trade wind cumuli and a fair amount of precipitation. Simulating a composite case has the advantage of focusing first on general characteristics of precipitating cumulus (such as average rain rate and area coverage) before concentrating on details of a specific day. The chosen period is visible in Figure 1 in which a time series of the area rainrate, derived from the SPol radar observations, is displayed. A three week interval from December 16th to January 8th is easily distinguished since it is flanked by two deep(er) convection rainfall events, bringing much more rain to the RICO region. The area averaged amount of rain during the three week trade wind cumulus period is about 0.3 mm/day.

Both the LES case and the SCM case will be based on this period. The LES will perform two 24 hour simulations based on an average three week large scale forcing (one with and one without microphysics) and the SCM are asked to simulate the whole three week period.
For both cases, the two main questions are:

  • What will be the mean state to which the boundary layer develops?
  • Are the models able to reproduce the observed rainfall?

When the latter question is answered positively, the sensitivity of the case for variations in the microphysical situation can be studied in further detail. People are invited to perform as many sensitivity tests as they like to do!



Figure 1. Time series of area rainfall during RICO, derived from SPol radar observations.

© KNMI