Error running wrf.exe with chem_opt=109: excessive amount of PM2.5 (sulfate) simulated

[mike_c]

Dear colleagues,

I ran wrf.exe with chem_opt = 109, which can handle the aqueous phase aerosol variables.
But the model stopped after a few forecast hours, with error message in rsl.error* files like below.


Below is the snapshot of PM2.5 and so4ai distribution at the lowest model level from 1-hr forecast wrfout file.
As you can see, an excessive amount of PM2.5 is simulated and is located randomly over the domain. (Max. = 5.67e9 ug m-3)
Especially, sulfate-related variables (so4ai, ao4aj, and sulf) caused this weird value of PM2.5 (Max. of so4aj = 166,267.4 ug kg-1),
but aqueous phase variables (so4cwi and so4cwj) seemed normal (Max. of so4cwj = 7.252 ug kg-1).



This issue was found in different versions of WRF-Chem I tested, v4.5.2 and v4.6.1 (the latest one).

Can you help me solve this problem?


Thank you,

Mike.

 

[Ioga Lazuardi]

Dear colleagues,

I ran wrf.exe with chem_opt = 109, which can handle the aqueous phase aerosol variables.
But the model stopped after a few forecast hours, with error message in rsl.error* files like below.
View attachment 17527

Below is the snapshot of PM2.5 and so4ai distribution at the lowest model level from 1-hr forecast wrfout file.
As you can see, an excessive amount of PM2.5 is simulated and is located randomly over the domain. (Max. = 5.67e9 ug m-3)
Especially, sulfate-related variables (so4ai, ao4aj, and sulf) caused this weird value of PM2.5 (Max. of so4aj = 166,267.4 ug kg-1),
but aqueous phase variables (so4cwi and so4cwj) seemed normal (Max. of so4cwj = 7.252 ug kg-1).

View attachment 17525 View attachment 17526

This issue was found in different versions of WRF-Chem I tested, v4.5.2 and v4.6.1 (the latest one).

Can you help me solve this problem?


Thank you,

Mike.

can you tell us which data of chem that you use, also meteorological dataset that you use as the input

 

[Ioga Lazuardi]

The forrtl: error (69): process interrupted (SIGINT) in WRF-Chem simulation likely caused by one of the following: insufficient memory, numerical instability, parallelization issues, or incorrect input data.

i try to find out and found it that can trigger that error

 

[mike_c]

The forrtl: error (69): process interrupted (SIGINT) in WRF-Chem simulation likely caused by one of the following: insufficient memory, numerical instability, parallelization issues, or incorrect input data.

i try to find out and found it that can trigger that error

can you tell us which data of chem that you use, also meteorological dataset that you use as the input

Thank you for your quick response!

- I used FNL data as meteorological input.
- No chemicla initial / boundary conditions from external global model were provided.
- For emission data, KORUS anthropogenic inventory and MEGAN biogenic emission were applied.

Oh and I forgot to mention that when running wrf.exe under the same options but with chem_opt = 108 (no aqueous chemistry), there was no issue.