LININTERP WARNING on the WACCM3.19 control run

[jyangdas@gmail_com]

I successfully configure and make WACCM3 on my system. After I submit the job with 16 CPUs, the model stops during the initialization process. The error message is:
***************************
LININTERP WARNING: 43.9393939393939377 % of output grid will have to be extrapolated
LININTERP WARNING: 43.9393939393939377 % of output grid will have to be extrapolated
...
interval was not found for col i 2
ENDRUN:CFINT2
rhtrunc < 0.0
ENDRUN:RADCSWMX
...
findsp not converging at point i, k 2 60
t, q, p, enin 27.3249650714363170 600.006794593493737 60652.0541162984009 1701013063.35417676
tsp, qsp, enout 2194146.96087572258 1.00000000000000000 -2831703968.78208637
ENDRUN:FINDSP
ENDRUN:PBUF_GET_FLD_IDX: index not found for û~Aÿ
...
g_tot > 1.0
ENDRUN:RADCSWMX
...
prob with neg rliq1 -0.258159222904476964E-09
ENDRUN:PCOND
***************************

I have changed nothing of the codes. What's wrong? Thanks very much.

 

[mmills@ucar_edu]

From your other posts, I gather you are running WACCM4. What is your reason for wanting to run 3.1.9 as well? It is a significantly different model and should not be used as a control for WACCM4 runs.

We have limited resources at NCAR for supporting WACCM. Prior to the release of WACCM4, our official policy was to support running WACCM3.1.9 only on NCAR's bluefire machine. Users may encounter significant challenges porting to other machines. With the release of WACCM4, we no longer have the resources to officially support 3.1.9 at all.

 

[jyangdas@gmail_com]

First, thanks very much for your reply. I have run WACCM4. But I find that the up-limit of CO2 level is 3000 ppmv in WACCM4, and the WACCM3.19 has not this limit. I need set CO2 level to be as high as 200,000 ppmv, therefore, I try to run WACCM3.19. I will try other way to solve my problems.

By the way, according to my understanding, in most of coupled photochemical and radiative-convective model, the eddy diffusion coefficient profiles are specified using fixed values, close to observation data. However, in WACCM, how to treat the species vertical transport, especially near the tropopause? Because I focus on the water vapor concentration above the tropopause in a extreme warm climate, I guess some key part of the WACCM is not suitable for this warm climate, which I should change. Could anybody, who is familiar this WACCM, give me some lights on the issue? Thanks very much.

 

[mmills@ucar_edu]

I am not aware of any such hard-coded upper limit in WACCM4, though it may be somewhere in the coupled components or in CAMRT. Regardless of whether you use 4 or 3.1.9, raising CO2 to be 20% of the atmosphere is going to cause lots of things to go haywire, such as the CO2 cooling in the MLT region, for a start. What planet are modeling?

 

[jyangdas@gmail_com]

(1) The CO2 limit is set in the lnd model: clm_varctl.F90
if ( (co2_ppmv 3000.0_r8) ) &
call shr_sys_abort( subname//' ERROR: co2_ppmv is out of a reasonable range' )

(2) I focus on a paleoclimate issue: aftermath of snowball Earth, ~600 Ma. First, the CO2 might be very low and the Earth entered into global freezing. After that, the CO2 concentration would increase due to absence of surface weathering process, and much high-level CO2 (~20%) must be accumulated to melt the snowball.

(3) What is CO2 cooling in the MLT region? Does it mean "the CO2 emission treatment above 50 km is very crude"? Can you give me more description? Or, can you give some paper on this?

Thanks very much for your guidance.

 

[mmills@ucar_edu]

You should talk with Jeff Kiehl, who developed the radiation code. It may be more appropriate for you to use CAM or CAM-Chem, rather than WACCM, for this problem. There is a paleoclimate working group of the CESM, and Jeff Kiehl is a co-chair.

See: http://www.cesm.ucar.edu/working_groups/Paleo/