Can MCIP generate same domain (row and column) but not reduced WRF domain?

MCIP
1

Hi, I am a new user for running MCIP and SMOKE.
NEI2017 data show the surrogate file for 12km grid is 516x444, so I set 517x445 in WRF (I may need to set 516x444), BTRIM=0 in MCIP run script, but output from MCIP is 514x442 (I know that “This affects the output MCIP domain dimensions by reducing meteorology domain by 2BTRIM + 2NTHIK + 1” in MCIP run script description).
I wander could I get the same domain for MCIP output like those in wrfout files from WRF? or I have to set 520x448 in WRF?

Thank you in advance.

2

Normally, good modeling practice indicates that you should have at least a 3-cell border within the met-modeling grid all around the AQ modeling grid…

Maybe what you should do is to window the surrogates to the MCIP-output file? M3Tools programs bcwndw and mcwndw are your friends for that…

3

As @cjcoats mentioned that good modeling practice indicates that you should have at least a 3-cell border within the met-modeling grid all around the AQ modeling grid. If you want to use NEI2017 data with the surrogate file for the 12km grid with 516x444 grid cells, you should set your WRF domain for 523*451. Then, you can set the boundary cut value as 3 in your MCIP script and get the results of MCIP for 516x444 grid cells.

4

Thank you for this suggestion! I am not familiar with M3Tools, but I can try that later if I have not good method.

5

Thank you for this suggestion! I will set 523x451.

6

Please also note that you should keep the projection parameters of WRF the same as the emission files (ref_lon, stand_lon, etc.). To get these values of NEI emission, just execute “ncdump -h filename.nc” then you will see the values printed on your screen.

7

Thank you! Ryan. I have not generated emission file using NEI2017. I checked the surrogate files in NEI2017 (LAMBERT meters 33.000000 45.000000 -97.000000 -97.000000 40.000000) and set these:
ref_lat = 40,
ref_lon = -97.0,
truelat1 = 33.0,
truelat2 = 45.0,
stand_lon = -97.0.

8

Looks good. Just go ahead and I think it will work well.

9

Hi Ryan,

It may be not appropriate to ask this question here. Just curious, do you have suggestion about any feasible combinations of physical options (such as sf_sfclay_physics, sf_surface_physics, sf_urban_physics, bl_pbl_physics) in WRF to get meteorological data in 2017 for MCIP? Recently, I ran two combinations with every 5 days reinitialization but both did not work well.

Thank you.

10

It largely depends on your model domain and reanalysis data that drives WRF. If you are modeling North American, the NARR dataset would be a good fit which has a 12km spatial resolution (the model performance is much better than NCEP FNL). Regarding the physical parameterizations schemes, I suggest you search the US EPA meteorological assessment report and adopt their selection. It won’t be hard to find the project report (you just need to google it.

11

Yeah, it is easy to find. Thanks a lot for these suggestions.