Transcript CAM 4 Update • The plan Phil Rasch – What is happening over the next 2.5 days – Review of progress over last year –

CAM 4 Update
• The plan
Phil Rasch
– What is happening over the next 2.5 days
– Review of progress over last year
– What remains to be done before CAM4
– Last minute tweaks to Agenda? Comments?
Questions?
Agenda Outline
• Tuesday PM & Wednesday joint with Chemistry and
Climate WG
• Tuesday PM –
– Overview
– Update on Aerosols/Clouds for CAM4
– Reception
• Wednesday AM
– Aerosol/Clouds Applications
– Radiation
• Wednesday PM
– Convection
Agenda Outline – Pt II
• Thursday AM
– WACCM
– General Interest Presentations
• Thursday PM
– General Interest Presentations
– Updates From other WGs
– General Discussion & Closing
• Friday AM
– Extra-Terrestrial applications of CAM
April 2007 SSC - CAM 4 Update
Phil Rasch
• Defined CAM3.5, an interim model for use in
BGC development.
– Revised deep convection
• 3 candidate schemes (Neale/Richter, Zhang, Wu/Zhang)
• extremely careful evaluation procedure developed
• Neale Richter scheme selected for 3.5, but this does not
imply that it will be used in CAM4
– Change to location of call to cloud fraction
calculation (better physical justification)
– Freeze dried clouds
As of April 2007 SSC
Next Steps (nearly in the bag)
•
Move forward in aerosol formulation:
– merger of CGD & ACD bulk aerosol formulation
(these aerosols will be used for transient simulations with CCSM3.5)
(DONE, predicted or prescribed fuctionality. now the preferred
dataset, coupled 0.5x0.5 also done)
– integrate with new microphysics & Ghan droplet activation(DONE)
•
Complete the Assessment of UW PBL and Shallow convection schemes
(mostly DONE, papers in prep, 1st fully coupled simulations in place)
•
Complete the Assessment of Morrison/Gettelman microphysics (mostly
•
Revised gravity wave and orographic wave breaking formulations
•
Complete the Assessment of Hybrid Isentropic coordinates (Papers in
DONE, papers submitted, indirect effects being assessed,
coupling with UW scheme in process [some surprises])
(works, but not yet exploited)
prep, On Hold)
(As of April 2007)
Next Steps
(likely changes, but final outcome uncertain)
• Explore horizontal and vertical resolution sensitivities (need to
increase level of interaction with the DOE effort) (working, time
for more focus on vertical)
• Ghan aerosol formulation desirable, interim aerosol solution (by
Rasch, Mahowald, Hess, Barth, Bond) is the backup (Progress, but
not there yet)
• Radiation
– AER RRTM (Progress, but not there yet)
– Assessment of change in representation still taking place
– new radiative transfer API in development
– Connections with microphysics, aerosols
– Sub-column generator
– BUGSrad (current implementation not ready)
– ??
(Comments following April 2007 SSC meeting)
Next Steps (very uncertain)
• subgridscale water variability (Little Progress)
– sub-column generator
– radiation --> heating & photolysis
– microphysics
• alternate deep convection schemes
– Zhang?
– Bretherton?
– Emanuel?
Time Schedules
• First Milestone for CAM4 completion was Jan 2008
• First delay --> March 2008
• Second delay --> October 2008
• Loss of Collins, Hack in the Climate Modeling Section (CMS).
• Peter Lauritzen (numerical methods expert) joins CMS from
ASP
• Rasch starts Sabbatical March 2008
– BUT I WILL CONTINUE BEING ENGAGED!
• CMS Attempting to refill Collins’ Position
• Hack on leave.
– No immediate opportunity to refill position
• Redirection of Software Engineering group resources
Unresolved issues
• PBL developments
– Visit last summer by Svensson, Holtslag & Park to work with
Rasch and Large on PBL
– Improvements to UW scheme, & sfc exchange parameterizations
– Revised Holtslag and Boville scheme
• Improved entrainment formulation for top of PBL
• 20m thick surface layer (31 layer model)
• Coupled simulation looks very interesting. Little improvement in
standalone CAM, but interesting coupled simulations!
• Opportunity to choose the best of both schemes
Low resolution CAM
-Preliminary results
Peter Hjort Lauritzen
NCAR
In idealized test cases (adiabatic baroclinic wave and aquaplanet)
-EUL-T85 is equivalent to FV-1.0x1.25
-EUL-T42 is equivalent to FV-2.0x2.50
-EUL-T21 is NOT equivalent to FV-4.0x5.0
Day 9 baroclinic wave test case
Williamson (2007), submitted to TELLUS.
Jablonowski and Williamson, Q. J. R. Meteorol. Soc. (2006)
EUL-T31 is used for paleoclimate applications
Is there a FV resolution resembling EUL-T31?
Day 9 baroclinic wave test case
Aqua-planet simulations (Neale & Hoskins, 2000)
Zonal averages
SURFACE PRESSURE
Global average
ZONAL SURFACE STRESS
Meridional eddy statistics
Efficiency
Standalone CAM (not idealized)
•
FV-2.7x3.3 is approximately 55% more expensive than EUL-T31
•
FV-2.7x3.3 is approximately 54% less expensive than FV1.9x2.5.
Settings
FV-2.7x3.3 : dt = 30 min. (10 min. dynamics dt)
FV-1.9x2.5 : dt = 30 min. (7.5 min. dynamics dt)
EUL-T31
: dt = 30 min.
Default IO settings, 30 day run on one node on Bluevista
Phil Rasch Post-doc Advert
• I have an opening for a post-doc to work on
two projects:
– Mimicing domain averaged Cloud Resolving Model
behavior using Statistical Learning Theory
– Exploring Climate Change mitigation through geoengineering
• Starts around now
• Contact me: Phil Rasch, <pjr at ucar.edu>,
303-497-1368