Cold Pools and the Organization of Tropical Convection in Global Cloud-System Resolving Simulations

Cold pools produced by tropical oceanic precipitating convection can affect existing and
subsequent convection in many ways. Cold pools are usually detrimental to isolated convection,
but may be beneHcial when convection is organized, and appear to play an important role in the
transition from shallow to deep convection. The DYAMOND (DYnamics of the Atmospheric general
circulation Modeled On Non-hydrostatic Domains) project produced simulations of forty days,
beginning 1 August 2016, using global models with a cloud-system resolving grid spacing of 5 km
or less. Nine different groups submitted the results of such simulations. How might we use this
set of simulations to learn more about the effects of cold pools on convection?

We have developed a global (tropical ocean) cold-pool climatology from one of the simulations.
We found that a simple cold pool detection algorithm seems to perform quite well for the tropical
oceans. In the oceanic ITCZs, simulated cold pool occurrence (fraction of time a grid box is part of
a cold pool) ranges from 0.1 to 0.25. The western Atlantic ITCZ and the western Indian ITCZ are
the most active. Simulated gust front event frequency ranges from 1 to 2 per day (see graphic).
The western Atlantic, eastern PaciHc, and the western Indian ITCZs are the most active by this
measure. The average cold pool duration (at a location) ranges from 3 to 5 hours. There is little
large-scale geographical variability within the oceanic ITCZ. The gustiness speed (difference
between mean wind speed and vector mean wind speed) pattern closely matches that of the gust
front frequency. We did not Hnd an obvious correlation between cold pool activity and shear of the
mean zonal wind.

Our next steps will include (1) applying the same cold-pool detection algorithm to tropical buoy
data in order to evaluate the simulated cold pools against observations, (2) calculating a measure
of clustering or organization, the radial distribution function, for the simulated convective updrafts
to determine if there is a relationship between cold pools and convection organization in these
simulations, and (3) extending the cold pool analysis to tropical land areas. We plan to present
results from these efforts at the meeting.