Recent Model Development Activities

Kate Evans, Oak Ridge National Laboratory

For simulations under the Ultra-High Resolution DOE BER project, check the official website for updates. Unfortuantely due to network security restrictions at ORNL, most of the sites below will not be viewable by those outside of ORNL. The official website should provide clear access however.


Project Members working on various runs below and in links: Rick Archibald, David Bader, Julie Caron, Kate Evans, Mat Maltrud, Julie McClean, Jim Hack, Salil Mahajan, Mark Taylor, Pat Worley











Animation of the T341 atmospheric column integrated instantaneous precipitable water from a preindustrial run of the coupled CCSM4. Precipitable water hourly for March (clicking the picture will download the 4.6MB movie):



CESM-CISM run atm diagnotics


45 years 1960-2005

ACME v0 test run diagnostics


ne30 F1850 CAM5: tuning run 4 years with settings as in 'final CAM tunings' doc

ne30 F1850 CAM5: tuning run 8 years with settings as in 'final CAM tunings' doc, but old seasalt

High Resolution Runs being evaluated


Link to untested production runs in the CCSM4

Link to present day/AMIP development in the CCSM4

Link to HOMME development in the CCSM4

Link to Rick's matrix of parameter space for physics tunings with T85, F compset


CAM4 T85 tuned model runs


1/25/11 3 year T85 preindustrial tuned run with 1 degree FV land, and 1 degree active ocean and sea ice, versus observations. For this run: rhminh=0.68, rhminl=0.905 as with the atm/land only run. This run is drifting a little from the F compset version, double ITCZ present as with FV and SE.

1/25/11 3 year T85 preindustrial tuned run with 1 degree FV land, and 1 degree active ocean and sea ice, versus tuned and balanced T85 atm/T85 lnd only preind simulation. For this run: rhminh=0.68, rhminl=0.905 as with the atm/land only run.

4 year T85 preindustrial tuned run versus CCSM4 released FV 1deg. Uses 2deg preind aerosol dataset, newer preind Hadley SST's, and subcycling. For this run: rhminh=0.68, rhminl=0.905.

Years 5-10 T85 preindustrial tuned run versus obs. Uses 2deg preind aerosol dataset, newer preind Hadley SST's, and subcycling. For this run: rhminh=0.68, rhminl=0.905.

5 year T85 preindustrial run to tune with subcycling and new spun up land file year 22 For this run: rhminh=0.68, rhminl=0.905.


CAM T341 Tuning Runs

Tuning of the CCSM4 configured for high resolution.
9/30/11 Comparison of T341 preindustrial tuning runs (1) 3 year with 3 months spin-up, rhminh=0.65, rhminl=0.903, dt=1800 compared to (2) 2 year with 1 year spin-up, rhminh=0.80, rhminl=0.905, dtime=900.

8/29/11 2 year T341 preindustrial tuning run with 1 year spin-up, rhminh=0.80, rhminl=0.905, dtime=900, compared to obs.

8/29/11 2 year T341 preindustrial tuning run same as above, compared to earlier t341 below to test recent tweaks to CESM.

3/28/11 1 year T341 preindustrial tuning run, 4 months spin-up, rhminh=0.97, rhminl=0.893, dtime=900, compared to tuned t85. This tuning range appears to be out of the normal range of operation, and tuning further in this line of high thresholds probably won't produce a better simulation. The cloud totals are higher, as we hope to match obs, but other variables such as LWCF and SWCF do not correspond as expected. To continue, an adj of the phsyics parameterization timescale would probably be the route with greatest cahnce for success.

3/11/11 1 year T341 preindustrial tuning run, only 1 month spin-up, rhminh=0.95, rhminl=0.89, dtime=900, compared to tuned t85. This run, with high RH thresholds is to compare to similar CAM-SE run that produced higher could levels.

1/29/11 1 year T341 preindustrial tuning run, 1 year spin-up, rhminh=0.80, rhminl=0.905, dtime=900, compared to tuned t85.

1 year T341 preindustrial tuning run, 1 year spin-up, rhminh=0.80, rhminl=0.905, dtime=900, compared to obs.

9/29/11 3 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.65, rhminl=0.903, dt=1800 compared to newer obs.

3 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.65, rhminl=0.903, dt1800. Model is pretty close to energy balance, here its tuned to be RESTOM close to 0 and close to RESSURF.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.65, rhminl=0.902. Model still needs help with E balance, here its tuned to be RESTOM close to 0, but ~0.4 Watt discrepancy.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.30, rhminl=0.80.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.45 (even though it says h50), rhminl=0.85.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.55, rhminl=0.85.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.65, rhminl=.905 with physics dt=900.

1 year T341 preindustrial tuning run, 3 months spin-up, rhminh=0.65, rhminl=.905

Initial 1 year T341 preindustrial tuning run versus same settings for T85 (tuned), rhminh=0.68, rhminl=.905


CAM FV Prognostic Aerosol File Runs - Rerun with new CCSM4 release and datasets


12/06/10 2001-5 (Cyclic 2000) with BAM chemistry and new datasets as below versus obs

11/27/10 2001-5 (Cyclic 2000) with BAM chemistry and new datasets versus same model preindustrial as below with 1 yr spin up.

Preindustrial T85 run with default 2 degree climatological aerosol dataset versus newer 1 degree resolution created from run below

1851-8 (Preind) BAM chemistry with new datasets at 1 degree versus same model with no active chemistry with 1 yr spin up.



Link to older runs, T85 tuning work and development using pre-CCSM4 release

Link to subcycling diagnostic plots, results published in Taylor et al., J Phys. Conf Series, 2010.



For Additional Information:

Kate Evans
1 Bethel Valley Road
P.O. Box 2008, MS 6016
Bldg. 5600, C315
Oak Ridge, Tennessee 37831-6015
Phone: (865) 576-6517
Fax: (865) 576-5491
Email: evanskj at ornl dot gov