Proceedings - C-SRNWP Project
Proceedings - C-SRNWP Project
Proceedings - C-SRNWP Project
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
A simple parameterization for detrainment in shallow<br />
cumulus.<br />
Wim de Rooy and Pier Siebesma<br />
Royal Netherlands Meteorological Institute, The Netherlands<br />
Recently there has been a regained<br />
interest in the parameterization for<br />
entrainment in cumulus convection.<br />
Unfortunately little attention has been paid<br />
to the parameterization of the detrainment<br />
process although this counterpart of the<br />
cloud mixing process is, as this study<br />
shows, even more important for obtaining<br />
realistic mass flux profiles in cumulus<br />
convection.<br />
A new simple but flexible<br />
parameterization for the detrainment<br />
process in shallow convection is presented.<br />
From ensemble considerations and<br />
confirmed by LES results it is shown that<br />
the magnitude of the fractional detrainment<br />
for shallow convection is decreasing with<br />
increasing depth of the cloud layer. A<br />
simple detrainment formulation is<br />
presented that takes this dependency into<br />
account by considering the mass flux<br />
profile in a non-dimensionalized way. The<br />
only free parameter in this detrainment<br />
formulation is the percentage of the mass<br />
flux at cloud base that is still present halfway<br />
the cloud layer (called m * ). This free<br />
parameter turns out to depend on the<br />
environmental conditions as well as<br />
updraft properties and can be expressed<br />
with a parameter coming from the<br />
buoyancy sorting concept, χ<br />
c<br />
, the fraction<br />
of environmental air necessary to make the<br />
updraft air just neutrally buoyant and<br />
averaged over the lowest half of the cloud<br />
layer. LES results show a distinct relation<br />
___________________<br />
* Corresponding author address: Wim<br />
de Rooy, KNMI, P.O. Box 201, 3730<br />
AE, De Bilt, The Netherlands. E-mail:<br />
rooyde@knmi.nl<br />
between m * and χ<br />
c<br />
which can be used to<br />
close our parameterization.<br />
z [m]<br />
2500<br />
2250<br />
2000<br />
1750<br />
1500<br />
1250<br />
1000<br />
750<br />
500<br />
250<br />
0<br />
initial profile (0:00 UTC)<br />
LES (24:00 UTC)<br />
fixed ε and δ (24:00 UTC)<br />
new δ (24:00 UTC)<br />
4 5 6 7 8 9 10 11 12 13 14 15 16 17<br />
q t<br />
[gr kg -1 ]<br />
Fig. 2. Total specific humidity profiles<br />
after 24 hrs of simulation during the RICO<br />
case for the LES model (black squares)<br />
and the SCM using fixed ε and δ (blue<br />
squares) or the new detrainment<br />
parameterization (red triangles). The black<br />
solid line represents the initial profile.<br />
Results of the new parameterization in<br />
a Hirlam single column model (SCM) for a<br />
wide range of shallow cumulus convection<br />
cases (BOMEX, ARM and RICO)<br />
illustrate the strength of this new<br />
detrainment parameterization. As an<br />
example, we show in Fig. 2 the total<br />
specific humidity profile of the<br />
independent RICO case after 24 hours of<br />
simulation. Presented are the results for an<br />
LES model, a SCM with fixed fractional<br />
entrainment (ε) and detrainment (δ)<br />
coefficients as known from literature, and<br />
finally a SCM with the new detrainment<br />
parameterization. The new<br />
parameterization gives an almost perfect<br />
match with the LES humidity profile.<br />
There is also a large improvement on the<br />
114