BALTEX BRIDGE cloud liquid water network project: CLIWANET

1
GCSS Modelling strategy and the use of new
satellite products A. Pier Siebesma, Joao
Teixeira and Christian Jakob siebesma_at_knmi.nl chai
r GCCS Climate Research Department Royal
Netherlands Meteorological Institute (KNMI) De
Bilt, The Netherlands KNMI
Photo courtesy Bjorn Stevens
2
Objective of GEWEX Cloud System Studies (GCSS)
See http//www.gewex.org/gcss.html
3
(Simplified) Working Strategy of GCSS
Large Eddy Simulation (LES) Models Cloud
Resolving Models (CRM)
Single Column Model Versions of Climate Models
3d-Climate Models NWPs
Development
Testing
Evaluation
4
Website http//gcss-dime.giss.nasa.gov/
5
Example from a GCSS Case Study
Large Eddy Simulation (LES) Models Cloud
Resolving Models (CRM)
Single Column Model Versions of Climate Models
3d-Climate Models NWPs
Development
Testing
Evaluation
6
The RICO field study (B. Rauber, L. di Girolamo,
H. Gerber, L. Nuijens, B. Stevens
Precipitating trade wind cumulus
7
12 Large Eddy Simulation results of the RICO
intercomparison (1)
Good agreement of moisture and temperature
between LES model results and the radiosonde
observations (after 24 hours) But
8
Large Eddy Simulation results of the RICO
intercomparison (2)
Liquid water
Rain water
Precipitation flux
obs
Large(r) disagreement for the cloud liquid water
and.an even larger disagreement for the rain
water (also rain flux)
WHY??
Courtesy Steve Abel , UK Met Office
9
A Matter of Magnitudes!!

• total specific humidity qt 10 g/kg
• clouds appear in those parts of the atmosphere
• where qt exceeds the saturation specific humidity
  qs
• ql qt-qs O(10-2) qt
• rain water forms out of cloud liquid water and
  is a result
• complicated microphysical (subgrid) clustering
  processes.
• qr O(10-1) qt

Tip of the iceberg
10
Combining satellite data with groundbased
observations!
Jan. 15, 2005
MISR cloud height
SPOL radar precip rate

• Clouds as shallow as 500m do precipitate
• linear relationship cloud depth
• versus precipitation rate

Courtesy Larry Di Girolamo (U. of Illinois)
11
Example from a GCSS evaluation study
Large Eddy Simulation (LES) Models Cloud
Resolving Models (CRM)
Single Column Model Versions of Climate Models
3d-Climate Models NWPs
Development
Testing
Evaluation
12
GCSS/WGNE Pacific Cross-section Intercomparison
(GPCI) GCSS Contact person Joao Teixeira
ISCCP total cloud cover

• GPCI is a working group of the GEWEX Cloud System
  Study (GCSS)
• Models and observations are analyzed along a
  Pacific Cross section from stratocumulus, to
  cumulus and to deep convection for a 1998 JJA
  period
• Models GFDL, NCAR, UKMO, JMA, MF, KNMI, DWD,
  NCEP, ECMWF, BMRC, NASA/GISS, UCSD, UQM, LMD,
  CMC, CSU, GKSS
• GPCI is an extension of a FP6 EUROCS project
  results www.knmi.nl/samenw/eurocs and are
  published Siebesma et al. QJRMS 2003

13
present observational Data sets hosted on the
DIME site http//gcss-dime.giss.nasa.gov/
14
TYPICAL EXAMPLE
Total cloud cover (JJA98)
TOA net shortwave radiation (JJA98)
15

• But how about vertical structure of cloud related
  fields.?

16
GPCI mean relative humidity JJA 2003
17
Relative humidity differences model- AIRS (JJA03)
UKMO PBL is deeper and moister
Models are much drier above Sc AIRS sampling ?
18
(No Transcript)
19
CloudSats vertical view along the GPCI
Cross-Section (july-august 2006) (courtesy
Cristian Mitrescu NLR).
NIGHTTIME
DAYTIME

• All 20 GCM/RCMs will deliver data for JJA 2006.
• Cloudsat/Calipso cloud cover and liquid water/ice
  water profile products for this transect will be
  extremely desirable for further model
  evaluation!!!!!!

20
GPCI Summary

• Status
• Results from 20 models
• 1st workshop, NASA/GISS, Sep. 2006, together with
  BLCWG and RICO
• Lots of interesting results
• Lack of Sc cloud (getting better) SW biases
• Deep convective regime optically to thick. SW
  biases
• Too much precip in tropics
• How dry is sub-tropical upper troposphere? Need
  for observations
• Next steps(NASA funding for 1 person (MAP), next
  3 years)
• Organize model data on the web gt connection to
  DIME
• Satellite observations already on DIME
• A careful look at AIRS data (too dry above the
  Scu fields)
• Rerun all models for JJA 2006 and evaluate with
  cloudsat/calipso derived data

21
PAN-GCSS Meeting

• Planned
• June 2-6, 2008
• Meteo-France, Toulouse France
• Themes
• New Experiments and Observations A-train, AMMA,
  VOCALS
• Tropical Convection (together with SPARC
  troposphere-stratosphere interaction
• Operational (very) high resolution modelling
  (NWP/Climate)
• Cloud Climate Feedback.

• Working Group breakout sessions
• BL-clouds, deep convection, polar clouds, cirrus
• GPCI
• Metrics
• Microphysics

Excellent opportunity to have a
A-Train/Earthcare/GCSS encounter!! Thank You
22
Model RH only for TCC lt 70 model- AIRS (JJA03)
Larger bias in tropical upper troposphere
Slightly smaller bias above Sc
23
GPCI Motivation

• To study important physical regimes and
  transitions stratocumulus, shallow cumulus and
  deep convection
• To evaluate models and observations in the
  tropics and sub-tropics in terms of the
  atmospheric hydrologic cycle
• To utilize a new generation of satellite datasets
  (e.g. AIRS, CloudSat, GPS)
• To create a database of models and observations
  for continuous studies of the tropics and
  sub-tropics (in construction on the DIME site)
• To try to answer specific scientific questions
  regarding diurnal cycle, humidity structure,