Summary of the EUCLIPSE KICK-OFF
EUCLIPSE/GCSS-BLWG meeting 27-30 September 2010
Contents
1. Results at selected grid points (GCPI/CloudNet/ARM/AMMA)
2. Lagrangian
transitions
3. CGILS
4. New
proposals for EUCLIPSE WP3 activities
1. Results at selected
grid points (GCPI/CloudNet/ARM/AMMA)
This project aims at validating paramerizations
applied in ESMs with aid of observations from field campaigns and satellite
data. To this end ESMs operating in both a NWP and in a free climate mode will output
data for selected grid points (GCPI/CloudNet/ARM/AMMA). To make a connection to
the model intercomparison studies (CGILS and the transition cases) it is
imperative to use Single Column Model versions that are identical to their
respective parent ESMs.
1.1 Next
step
A breakout meeting will be scheduled in the spring
period, 2011.
1.2 Next
deliverable
D3.6:
Compilation of ESM results at selected grid points (GCPI/CloudNet/ARM/AMMA)
(Month 18).
2. Lagrangian
transitions
2.1 General
1. The observed variation in the modeled radiative
fluxes calls for an
intercomparison experiment that will be dedicated to results from radiative
transfer schemes used in the LES models and SCMs.
2. The sea
surface albedo will be a function of the solar zenith angle (as proposed by
Marat Khairoutdinov).
3.
Modelers should take case to use the right julian day for the simulations.
4. In each LES model the calculation of the
tendencies due to subsidence should be computed with an upwind scheme.
4. Calls to the radiation scheme used in the LES
models should be made at least once every 60 seconds.
5. The SCM code should be identical to its parent
ESM. Modeling results may be
dependent on the time step used.
2.2 Composite
cases
1.
Modelers should relax the vertical profiles towards the initial profiles in the
sponge layer.
2. Modelers should re-run the reference case first (after we make sure the LWD
and SWD are approximately the same after the first hour). As a next step the
two other composite cases will be simulated.
2.3 Lagrangian
transitions (Composite cases and ASTEX Lagrangian)
During the first half of the ASTEX Lagrangian the
mean state and turbulence structure is well represented by the LES models.
However, during the second half the boundary layer deepens too rapidly. The weakening of the large-scale
divergence is too strong (see also Sigg and Svensson 2004), and the longwave
radiative cooling across the cloud top appears too high in the models. Both
effects tend to enhance the entrainment rate (because the large-scale
divergence controls the vertical stability).
2.3.1 Refinement
of the ASTEX case
1.
Change the large-scale divergence rate. There is a possibility to
diagnose the large-scale divergence from new runs. For example, the UKMO used a
high resolution LAM to simulate the ASTEX area during the time of observations.
Another possibility is to make use of the fact that the temperature in the free
atmosphere was observed to be close to a steady-state. The large-scale
divergence can then be estimated with aid of the radiative flux profiles from
the LES models. The smaller weakening of the large-scale divergence is also
supported by Sigg and Svensson (2004).
2. During the transition the downwelling radiative
fluxes in the free atmosphere were observed to be time dependent. This effect
can be incorporated by implementing a time-varying free atmosphere state to get
a better agreement with the observations of radiative fluxes.
3. Change the geostrophic forcing to get a better
agreement with the observed weakening of the absolute value of the mean
horizontal wind velocity.
2.4 Next
steps
1 November 2010: Release of refined case
description.
1 February 2011: Submission of SCM and LES results
for ASTEX and the three composite cases.
LES results will include simulations using the
standard grid configuration on a 4x4 km^2 horizontal domain size. In addition, for
ASTEX large horizontal domain simulations are requested. Depending on the computational resources
one has available it is optional, yet desired, to use a finer resolution than
suggested on the website. For both the control as well as the large-domain simulations it is
important to deliver instantaneous 3D fields. This output will be used for a
detailed offline diagnosis of various PDFs and to calculate "tailor made"
statistics for validating assumptions made in SCM parameterizations (e.g.
massflux, eddy diffusivities etc.).
2.5 Next
deliverables
D3.1:
Description of the set-up for the following cases: ASTEX, the GPCI
stratocumulus and
shallow
cumulus atmospheric columns, and the SCM equilibrium state study (Month 12).
D3.2:
Storage in a public archive of instantaneous 3D LES fields and diagnostics from
LES fields that are key to parameterization schemes (Month 24).
D3.3:
LES and SCM results of the mean state, turbulence structure and microphysics for
the
ASTEX
case and the GPCI stratocumulus and shallow cumulus cases (Month 30).
3. CGILS
3.1 CGILS SCM
3.1.1
Summary
1. Evidences suggest that the CGILS results from
some models can be used to interpret GCM cloud feedbacks. For other models,
especially those displaying multiple equilibrium behavior, alternative
configuration of running the SCMs may be insightful. This was demonstrated by
the LMD group who added transient forcing to break up the multiple
equilibriums.
2. Two groups of models have been identified to show
the same sign or negligible cloud feedbacks at all three locations, displaying positive
and negative feedbacks. It would be interesting to see if the corresponding GCMs
show the same sign of cloud feedbacks.
3. The deciding physical process appears to be the
mixing at the cloud top, carried out either by explicit cloud-top entrainment
or shallow convection.
4. Comparisons with LES and with GCMs are yet to be
carried out. SCMers are waiting for LES and CFMIP results.
3.1.2 CGILS Future Plan
1. Use seasonal, interannual, and decadal variations to evaluate the models
2. Close coordination with CFMIP to conduct SCM&GCM sensitivity experiments
3. Use LES results to constrain the SCM parameterizations
3.1.3 Next Steps
1. Understanding of processes in each model.
2. Re-configure simulations, such as adding transience, if results are
suspicious.
3. Linking with GCMs
4. Use LES models to say something about the SCMs
Next CGILS meeting:
A telecon before the end of 2010
A CGILS session at the next EUCLIPSE general meeting
3.1.4. Next deliverable
D3.1:
Description of the set-up for the following cases: ASTEX, the GPCI stratocumulus
and
shallow
cumulus atmospheric columns, and the SCM equilibrium state study (Month 12).
3.2
CGILS LES
3.2.1 Summary
CGILS
is promising but challenging
Forcing
problems: Almost sorted out, except surface pressure and qv drying at S11/12
below 1.5 km
S6
(trade Cu): LES ~ agree at dz/dx = 100/40 m
control
PBL is deeper than climo,
+2K
cloud response is in the noise
S11
(decoupled Sc):
Some
LES make solid Sc with dz = 25m; others require finer dz to
do
so. Shallowing/FT drying feedback may hinder solid Sc.
+2K
cloud thinning in solid Sc models – working on why.
+2K
PBL deepening in all models
S12
(well-mixed Sc):
Some
LES collapse, some donŐt.
+2K
response not yet robust enough to take seriously
3.2.2 Plans
Maybe one more case
rerun, then write up results by early 2010.
3.3
Next deliverable
D3.3:
LES and SCM results of the mean state, turbulence structure and microphysics for
the
ASTEX
case and the GPCI stratocumulus and shallow cumulus cases (Month 30).
4. New proposals
for EUCLIPSE WP3 activities
The following proposed plans can help to provide
the following deliverables:
D3.4:
Identification and comparison of the key quantities used in ESM
parameterization schemes that control the cloud properties simulated in ESMs
with LES results and observations (Month 30).
D3.5:
Equilibrium solutions of SCMs, with an emphasis on the equilibrium cloud-top
height, cloud liquid water path, cloud fraction, and drizzle rate. Identification
of the key quantities that control these quantities (Month 30).
4.1 Radiation
intercomparison
The observed intermodel variation in the radiative
fluxes calls for an intercomparison experiment that will focus on the radiative
transfer schemes used in the LES models and SCMs.
Vertical atmospheric profiles representative for
the thermodynamic state off the coast of South-West Africa will be provided by
Pier Siebesma (KNMI). For this area satellite observations of the liquid water
path and upwelling radiation at the top of the atmosphere are available to
allow for a detailed comparison with modeling results.
In addition, modelers will be asked to apply their
radiative transfer schemes to initial vertical profiles of the ASTEX, composite
and CGILS cases. In addition, horizontally averaged instantaneous vertical
thermodynamic profiles from LES models can be provided to SCMs. This can yield
information about the inhomogeneity effect ("albedo bias") as LES
models compute the radiative fluxes in each subgrid column.
The vertical input profiles should be provided in
an uniform format. The models should run their radiation scheme for one time
step only, and all the other physical processes like convection and turbulent
transport should be switched off. This exercise can be repeated for different
solar zenith angles. In this way intermodel variations in the radiative
transfer can be identified.
4.2 "EUCLIPSE-CGILS
extended"
Obtain "fingerprints" of ESM cloud
parameterizations in terms of cloud liquid water path, cloud fraction, drizzle
and cloud base/top heights in an equilibrium state. To this end run SCMs to equilibrium
states for a wide range of SST, LTS and large-scale divergence rates. This can be done by increasing the
number of columns from the GPCI domain. Another approach could be to do a more
idealized set of experiments in which parameters like the SST, LTS and LS-div
are systematically varied. Such a study has been performed with a mixed layer
model by Stevens (2002). To set up such a set of idealized experiments for SCMs
a careful matching of the subsidence warming and radiative cooling in the free
atmosphere is required. The KNMI (Roel
Neggers and Sara dal Gesso) will explore the feasibility of these experiments.
Equilibrium
Liquid Water Path solutions as a function of the large-scale divergence and SST
obtained with a mixed layer model (Stevens 2002).
Report by
Stephan de Roode (TU Delft), 5 October 2010.