2 columns in cesm.log file.

[shiannder]

Hi,

I am curious to know about 2 columns within the cesm.log.23852.241015-223628 file I have. I think the first column is the tracer value, but what is the second column referring to (bolded)?

ecosys_driver:ecosys_driver_set_interior: NaN in dtracer_module, (i,j,k)=( 63 , 89 , 42 )
(lon,lat)=( 202.65888757817021 , 73.856890825333423 )
PO4 0.72219526767730713 7.4775732650251343E-012
NO3 13.473614692687988 2.1696642862581128E-010
diatChl 1.6691545923208650E-008 0.0000000000000000
diatC 0.0000000000000000 0.0000000000000000
diatP 3.0643426685017516E-010 0.0000000000000000
diatFe 2.3702857710437841E-013 0.0000000000000000
diatSi 5.4102211599627026E-009 0.0000000000000000
diazChl 0.0000000000000000 0.0000000000000000
diazC 0.0000000000000000 0.0000000000000000
diazP 0.0000000000000000 0.0000000000000000
diazFe 0.0000000000000000 0.0000000000000000
DI13C 2337.5002441406250 NaN
DO13Ctot 37.968872070312500 -7.5249064700486141E-011
DO13C 18.984436035156250 NaN
DI14C 2075.7087402343750 NaN
DO14Ctot 19.265087127685547 -1.1207917372472025E-010
DO14C 9.6325435638427734 NaN
DO14Cr 9.6325435638427734 -5.6039586862360127E-011
zootot13C 0.0000000000000000 0.0000000000000000
zootot14C 0.0000000000000000 0.0000000000000000
sp13C 0.0000000000000000 0.0000000000000000
sp14C 0.0000000000000000 0.0000000000000000
spCa13CO3 0.0000000000000000 0.0000000000000000
spCa14CO3 0.0000000000000000 0.0000000000000000
diat13C 0.0000000000000000 0.0000000000000000
diat14C 0.0000000000000000 0.0000000000000000
diaz13C 0.0000000000000000 0.0000000000000000
diaz14C 0.0000000000000000 0.0000000000000000

Thank you,
James

 

[jcampbell2548]

Hi,

I am curious to know about 2 columns within the cesm.log.23852.241015-223628 file I have. I think the first column is the tracer value, but what is the second column referring to (bolded)?




Thank you,
James

Hi, did you solve this? I now have the same error.

 

[shiannder]

Hi,

I do not get this error anymore, but I have not gotten my model to run either.

I added ./xmlchange DEBUG=TRUE to my start code to see more error details.

I made edits to my marbl file code to add 4 new tracers and I get other code errors in my cesm.log now.

Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.

Backtrace for this error:
(Task 140, block 2) Message from (lon, lat) ( 22.739, 63.824), which is global (i,j) (21, 90). Level: 1
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) it = 1
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x1,f = 0.1000000E-008-0.1651638E-003
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x2,f = 0.1000000E-005-0.8328159E-003
(Task 140, block 2) Message from (lon, lat) ( 22.739, 63.824), which is global (i,j) (21, 90). Level: 2
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) it = 1
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x1,f = 0.1000000E-008-0.1760819E-003
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x2,f = 0.1000000E-005-0.8459581E-003
(Task 140, block 2) Message from (lon, lat) ( 22.739, 63.824), which is global (i,j) (21, 90). Level: 3
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) it = 1
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x1,f = 0.1000000E-008-0.2085469E-003
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x2,f = 0.1000000E-005-0.8834217E-003
(Task 140, block 2) Message from (lon, lat) ( 22.739, 63.824), which is global (i,j) (21, 90). Level: 4
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) it = 1
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x1,f = 0.1000000E-008-0.2125728E-003
(Task 140, block 2) MARBL WARNING (marbl_co2calc_mod:drtsafe): (marbl_co2calc_mod:drtsafe) x2,f = 0.1000000E-005-0.8985626E-003
(Task 140, block 2) MARBL WARNING (marbl_interior_tendency_mod:compute_carbonate_chemistry): Warning reported from marbl_co2calc_interior() with dic_alt_co2

 

[mlevy]

Hi, did you solve this? I now have the same error.

Can you include the error message you are seeing? ecosys_driver:ecosys_driver_set_interior: NaN in dtracer_module, (i,j,k)=( 63 , 89 , 42 ) indicates that MARBL returned a NaN when computing tracer tendencies at the given (i,j,k) location on the grid; in the line

DI13C    2337.5002441406250                              NaN

DI13C is the name of the tracer, 2337.5002441406250 is DI13C tracer concentration at that grid location, and NaN is the tendency returned by MARBL. Typically this comes about due to a mismatch in the land mask between the model itself and either the initial condition file or a forcing file... so the model thinks that it is in an active ocean cell but some of the input data is masked out.

 

[mlevy]

Hi,

I do not get this error anymore, but I have not gotten my model to run either.

I added ./xmlchange DEBUG=TRUE to my start code to see more error details.

I made edits to my marbl file code to add 4 new tracers and I get other code errors in my cesm.log now.

The output you included doesn't contain any error messages, just warnings that the iterative solver in the carbonate module didn't find a solution on a given iteration.

 

[jcampbell2548]

Can you include the error message you are seeing? ecosys_driver:ecosys_driver_set_interior: NaN in dtracer_module, (i,j,k)=( 63 , 89 , 42 ) indicates that MARBL returned a NaN when computing tracer tendencies at the given (i,j,k) location on the grid; in the line

DI13C    2337.5002441406250                              NaN

DI13C is the name of the tracer, 2337.5002441406250 is DI13C tracer concentration at that grid location, and NaN is the tendency returned by MARBL. Typically this comes about due to a mismatch in the land mask between the model itself and either the initial condition file or a forcing file... so the model thinks that it is in an active ocean cell but some of the input data is masked out.

Thanks for responding, but I actually don't have this issue anymore! I am still stuck on the "bounding bracket for pH solution not found" error though. Even when I use an initial conditions file, river flux file, and nitrogen deposition file with everything set to 0 I somehow am only able to run it for 2 days. Changing the time step doesn't help and everything else should be the correct (paleo) geography. The paleo simulation runs without problem when the ocean ecosys is turned off. Are there any other files that absolutely need to be changed for ocean ecosys other than initial conditions, river nutrient flux, and nitrogen deposition? Whether I have these files set to paleo geography or completely set to 0, I run into the same error.

 

[mlevy]

You should be okay with using 0 river nutrient flux, but I don't think you want to initialize the BGC tracers to 0 (and I'm not sure about the nitrogen deposition). There are also iron sediment forcing and iron vent forcing files, but those should be okay being set to 0 for initial testing as well.

If you could run with the paleo geography and then post the bounding bracket error message you get that would be really helpful. What we've found is that the iterative solver for pH in the carbonate module is typically the "canary in the coal mine", and if something starts to go off the rails in either the ocean physics or the BGC it results in the bounding bracket issue. For example, I've seen runs where temperature is -100 C in some grid cells, and we only realize it because the pH solver fails (and prints out temperature, salinity, and a few BGC tracer values). We've also seen runs where some of the BGC tracers are very very large, and it turned out the longitude coordinates in the grid for the initial condition file were offset from the model grid by some number of degrees so it looked like the paleo continents were correct but they were actually in the wrong place.

 

[shiannder]

Hi,

I do not get the Nan value from DI13C, DI14C, DO13C, and DO14C now after several code additions.

By the way, I am trying to implement 4 new tracers (DO13C, DO13Cr, DO14C, DO14Cr) in marbl code to split up DO13Ctot and DO14Ctot because of the spurious fractionation effect I am seeing in DO13Ctot with the deep ocean (should be a vertical constant -22 permil, but it moved to -35 permil in the deep ocean. Additional remin and production tracers are added too.

My most model (jsl.144) recent error message (cesm.log in /mnt/lustre/letscher/jsl1063/jsl.144/run) I am getting is related to a floating point error, which i assume is a mass balance problem in my equation coded up.

Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.

I checked my initial condition file Nan values for the land, and it should match the land points. My initial condition has the 4 new tracers added based on just splitting the DO13Ctot and DO14Ctot. DO13C and DO13Cr= DO13Ctot*0.5, while DO14C and DO14Cr= DO14Ctot*0.5.

The new IC file is in /mnt/lustre/letscher/jsl1063/initcond/cisodocsl as "jsl137_ciso_ecosys_jan_IC_gx3v7_20180308.nc"

I think this floating point error is related to my code changes.

I have attached the cesm.log file but I can email it to you because the file may be too large.

I also have a word doc that tracks changes for the files I made edits to. Let me know if you want me to email it for better access.