CAM5.2

CAM5.2 development


Quick links ("clicking" brings you to the corresponding section)

1. Bugfixes (B) and new Physics (P)

2. CAM5.2-SE ne30

2.1. CAM Standalone tuning runs (SE)
2.2. Coupled simulations (SE-C)
2.3. Aerosol Indirect Effect (AIE)
2.4. Timings (T)

3. CAM5.2-FV 2 degree

3.1. CAM Standalone tuning runs (FV)
3.2. Coupled simulations (FV-C)

4. CAM5.1-FV 1 degree + bugfixes (doesn't include new activation)

4.1. Coupled simulations to test bugfixes on CAM5 (FV-B)

Key simulations are shown in PINK

 

1. Bugfixes (B) and new Physics (P)

Bugfixes (B)

Number
Casename
Description
Diagnostics
Comments
B1 f.e10.FC5.f09_f09.uwshfix.002 Bugfix to fix the unreasonable concentration of tracers with WACCM. w/ control Climate conserving for CAM5. Solves the problem with WACCM.
B2 F1850C5_5_1_32_mawu bug fix for water uptake in the modal aerosol w/ control  
B3 F1850C5_5_1_32_obklen Fix for obklen variable w/ control  
B4 F1850C5_5_1_32_zm Deep convection bug fix w/ control  
B5 cesm1_0_rel07_scalingdustfix Bug fix for dust from fractional land cells w/ control This run was done with an older version of the model
B7 f2000_cam5_beta15_modal Bug fix for ndrop w/ control  

New Physics (P)

Number
Casename
Description
Diagnostics
Comments
P1 f.e10.FC5.f09_f09.cam5_1_32-activat.002 Activation w/ control  

 


2. CAM-SE ne30

2.1. CAM standalone tuning runs: CAM-SE ne30

Number
Casename
Description
Diagnostics
Comments
SE1 TOPO.ne30_g16.FAMIPC5.EXP_05_ext

16x topography + no damping


energy_fixer =-1
nu =10.e14
nu_q =10.e14
nu_p = 0.e14
nu_div=-1
hypervis_subcycle =1
hypervis_subcycle_q=1
topo=USGS-gtopo30_ne30np4_16xdel2-PFC-consistentSGH.nc

w/ obs

This run was done by Pat on hopper.

For more info on previous runs see Pat's webpage:
From a cdg machine:
cd /project/amp/patc/Topography_EXP
firefox CAM-SE_Topography_Experiments.html

 

SE2 f.e10.FC5.ne30_ne30.topo.003

16x topography + 25x damping

energy_fixer=-1
nu =10.e14
nu_q =10.e14
nu_p = 0.e14
nu_div=25.e14
hypervis_subcycle=2
hypervis_subcycle_q=1

topo=USGS-gtopo30_ne30np4_16xdel2-PFC-consistentSGH.nc

w/ obs  
SE3 f.e10.FC5.ne30_ne30.topo.004

Same as SE2 + ftype =1

Change from:
se_ftype=0 (tendencies are applied equally over all dynamics steps)
to
se_ftype=1

w/ obs
w/ control (S2)
w/ FV-1d

 
SE4 f.e10.FC5.ne30_ne30.topo.010

Same as SE3
+ activation change
+ changes in tuning parameters (rpen and a2l) to improve stratocumulus

Change of tuning parameters
rpen=10->5
a2l=30->20

w/ obs
w/ control (S3)
w/ FV-1d

 
SE5 f.e10.FC5.ne30_ne30.topo.011

Same as SE4
+ fix for the land scaling for dust

w/ obs
w/ control
w/ FV-1d
 
SE6 f.e10.FC5.ne30_ne30.topo.012

Same as SE4
+ retuning of dust
:
dust_emis_fact = 0.35->0.50

w/ control  
SE7 f.e10.FC5.ne30_ne30.topo.013

Same as SE4
+ retuning of dust
:
dust_emis_fact = 0.35->0.65

w/ control
 
SE8 f.e10.FC5.ne30_ne30.topo.014

Same as SE4
+ 6 bugfixes
- Fix for the unreasonable concentration of tracers with WACCM.
- Fix for water uptake in the modal aerosol
- Fix for obklen variable
- Fix in the deep convection
- Fix for dust from fractional land cells
- Fix for ndrop

w/ obs
w/ control
w/ CAM5.1(FV-1d)
 
SE9 f.e11.FC5.ne30_ne30.tuning.015

Same as SE8
+ retuning of dust + increase penetrative entrainment

dust_emis_fact = 0.35->0.55
rpen=5->3

w/ obs
w/ control
 
SE10 f.e11.FC5.ne30_ne30.tuning.016

Same as SE9
+
fix to flatten oceans near the coast

w/ obs
w/ control
w/ CAM5.1(FV-1d)

 
SE11 f.e11.FC5.ne30_ne30.tuning.017

Same as SE10
+ a2l: 20->30

w/ obs
w/ control
w/ CAM5.1 (FV-1d)

 
SE12 f.e11.FC5.ne30_ne30.tuning.018

Same as SE11
+ Back off the flattening of the ocean

w/ obs
w/ control
w/ CAM5.1 (FV-1d)

 

 

2.2. Coupled simulations: CAM-SE ne30

Number
Casename
Description
Diagnostics
Comments
SE-C1 b.e11.B1850C5CN.ne30_g16.tuning.001

Same as SE12 in the coupled framework
(This run is using a cold start - mistake in the initialization)

w/ obs
w/ CESM-CAM5.1 (FV 1d)

w/ CAM standalone

w/ CAM standalone with pre-industrial emissions

Problem in initialization (Start up run instead of hybrid run)

This is fixed in SE-C1-fix

SE-C1-fix b.e11.B1850C5CN.ne30_g16.tuning.001

Same as SE-C1 but fix the problem in initialization

w/ obs
w/ CESM-CAM5.1 (FV 1d)

Timeseries(ANN)

 

SE-C2 b.e11.B1850C5CN.ne30_g16.tuning.002

Same as SE-C2
+ rpen =3->5
(This run is using a cold start - mistake in the initialization)

w/ obs

Problem in initialization (Start up run instead of hybrid run)

SE-C3 b.e11.B1850C5CN.ne30_g16.tuning.003

CAM5.1 + 6 bugfixes (without activation change)
- Fix for the unreasonable concentration of tracers with WACCM.
- Fix for water uptake in the modal aerosol
- Fix for obklen variable
- Fix in the deep convection
- Fix for dust from fractional land cells
- Fix for ndrop

w/ obs
w/ CESM-CAM5.1 (FV 1d)

Timeseries(ANN)

 

SE-C4 b.e11.B1850C5CN.ne30_g16.tuning.004

Same as SE-C2
+ rpen =10->5


w/ obs
w/ CESM-CAM5.1 (FV 1d)
w/ control

 

 

2.3. Aerosol indirect effects: CAM-SE ne30

Number
Casename
Description
Diagnostics
Comments
AIE1 f.e11.FC5.ne30_ne30.AIE.018

Same as S12
+ use pre-industrial aerosol values

w/ control

AIE = -1.0 W/m2
d(SWCF) (pd-pi) = -1.3
d(LWCF) = 0.3
d(FSNTC) = -0.47
d(LWP) = 3.3

For comparison in CAM5.1 we had:
AIE = -1.0 W/m2
d(SWCF) (pd-pi) = -1.58
d(LWCF) = 0.56
d(FSNTC) = -0.2
d(LWP) = 3.4

2.4. Timings: CAM-SE ne30

Number
Casename
Description
Number of years per day
T1 f.e10.FC5.ne30_ne30.topo.006

8 nodes
phys_loadbalance=0
pure mpi

2.61
T2 f.e10.FC5.ne30_ne30.topo.005

8 nodes
phys_loadbalance=2
pure mpi

2.87
T3 f.e10.FC5.ne30_ne30.topo.007

32 nodes
phys_loadbalance=2
pure mpi

8.47
T4 f.e10.FC5.ne30_ne30.topo.008

8 nodes
phys_loadbalance=2
4 threads

3.06
T5 f.e10.FC5.ne30_ne30.topo.009

8 nodes
phys_loadbalance=2
2 threads

3.34
T6 f.e11.FC5.ne30_ne30.tuning.018

12 nodes
phys_loadbalance=2
2 threads

4.8

 

 


 

3. CAM-FV 2 degree

3.1. CAM stanalone tuning runs (CAM-FV 2 degree)

Number
Casename
Description
Diagnostics
Comments
FV1 f.e11.FC5.f19_f19.tuning.001

FV 2d + 6 bugfixes + activation change
- Fix for the unreasonable concentration of tracers with WACCM.
- Fix for water uptake in the modal aerosol
- Fix for obklen variable
- Fix in the deep convection
- Fix for dust from fractional land cells
- Fix for ndrop

- Activation change in microphysics

w/ obs
w/ CAM5.1 (FV-2d)
w/ CAM5.1 (FV-1d)

 

 

FV2 f.e11.FC5.f19_f19.tuning.002

same as FV1
+ rpen =3

w/ obs
w/ CAM5.1 (FV-2d)
w/ control

We have tried to use the same tuning parameters for CAM-SE and CAM-Fv but this pushes us too far away in the SWCF.
We back off from this change and use dp1 as a tuning parameter to decrease SWCF and try to improve SWCF

FV3 f.e11.FC5.f19_f19.tuning.003

same as FV1
+ cldfrc_dp1 = 0.1->0.05D0

w/ obs
w/ CAM5.1 (FV-2d)
w/ control

 

 

 

3.2. Coupled simulation with (CAM-FV 2 degree)

Number
Casename
Description
Diagnostics
Comments
FV-C1 f.e11.FC5.f19_f19.tuning.001

Same as FV3 in coupled framework
Use: cldfrc_dp1 = 0.05D0

w/ obs
w/ CAM standalone
w/ CESM-CAM5.1 (FV2 deg)

 

FV-C2 f.e11.FC5.f19_f19.tuning.003

CAM5.1 + 6 bugfixes (without activation change)
- Fix for the unreasonable concentration of tracers with WACCM.
- Fix for water uptake in the modal aerosol
- Fix for obklen variable
- Fix in the deep convection
- Fix for dust from fractional land cells
- Fix for ndrop

w/ obs
w/ CESM-CAM5.1 (FV2deg)

Timeseries (ANN)

 

 


 

4. CAM5.1-FV 1 degree

4.1. Coupled simulations to test CAM5.1 + bugfixes (doesn't include activation)

Number
Casename
Description
Diagnostics
Comments
FV-B1 b.e10.B1850C5CN.f09_g16.bugfix.001

CAM5.1 + 6 bugfixes (without activation change)
- Fix for the unreasonable concentration of tracers with WACCM.
- Fix for water uptake in the modal aerosol
- Fix for obklen variable
- Fix in the deep convection
- Fix for dust from fractional land cells
- Fix for ndrop

See wiki

run done on hopper