opening wrong wrffirechemi* file

[Xin Zhang]

Hi everyone,

When I used fire_emissions in WRF-Chem, I encountered a problem.

I set these options in namelist.input (2015_08_06_18:00:00-2015_08_07_18:00:00):

 &time_control
  io_form_auxinput7                   = 2,
  frames_per_auxinput7                = 1, 1, 1,
  auxinput7_inname                    = 'wrffirechemi_d<domain>_<date>',
  auxinput7_interval_m                = 60, 60, 60,
 &domains
  time_step                           = 135,
  max_dom                             = 3,
  dx                                  = 27000, 9000, 3000,
  dy                                  = 27000, 9000, 3000,
 &chem
  biomass_burn_opt                    = 2, 2,2,
  plumerisefire_frq                   = 30, 30,30,

So, WRF-Chem should read fire_emission data one houy by one hour. But, I got this error:

d01 2015-08-06_19:00:45 open_aux_u : opening wrffirechemi_d01_2015-08-06_19:00:45 for reading. DATASET DATASET=AUXINPUT7
d01 2015-08-06_19:00:45 calling wrf_open_for_read_begin in open_u_dataset
d01 2015-08-06_19:00:45 module_io.F: in wrf_open_for_read_begin
d01 2015-08-06_19:00:45  NetCDF error: No such file or directory
d01 2015-08-06_19:00:45  NetCDF error in wrf_io.F90, line        1077
...........
d01 2015-08-06_19:00:45  ---- ERROR: Could not find matching time in input file wrffirechemi_d03_2015-08-06_19:00:00
NOTE:       1 namelist vs input data inconsistencies found.

Could you help me with this problem?

Thank you,
Xin

[Stacy Walters]

Xin,

The error you encountered is due to the fact that your time step for domain 1 is

135 seconds.  This results in WRF trying to read wrffirechemi_d01 at a time of

19:00:45 which does not exist on the dataset.

I recommend you try setting the namelist variable time_step as follows:

time_step = 120

Of course this will increase your execution time by ~ 12%. 

Stacy

[Xin Zhang]

Hi Stacy,

After setting time_step = 120, it works fine now. So, I should ensure that 3600/time_step = integer.

But, I find this when checking the rsl.error.0000: d03 2015-08-07_02:59:33+01/03 .......

Fraction appears in the calculation. Is this caused by 120 which isn't an integral multiple of 27 (dx)? Does that matter?

Best regards,

Xin

[Stacy Walters]

Xin,

The "_02:59:33+01/03" results from:

(a)  time_step = 120

(b)  parent_time_step_ratio = 1,3,3

The combination of (a) and (b) results in the following dt values

grid                    dt
-----                    --
01                      120
02                        40
03                        13+01/03

If you want to have a "whole" value for dt on domain 03 you could set:

parent_time_step_ratio = 1,3,4

which result in dt = 10 for domain 03.

Stacy

[Bo Wang]

Hi everyone,

I got similar error (error opening wrrfirechemi*) with time_step = 180. Could you help me to fix the problem, please? Any help would be greatly appreciated!

I'm running WRF-Chem v3.9.1. I have prepared meteorological datafiles by WPS and hourly fire emission files named: wrffirechemi_d0N_2016-11-12_HH:00:00 from FINN inventory (by fire_emis processor) for a test case. 

With the following namelist.input:

&time_control 

frames_per_auxinput7                = 1,       1,

auxinput7_inname                    = 'wrffirechemi_d<domain>_<date>',

auxinput7_interval_m                = 60,     60,

io_form_auxinput7                   = 2

&domains

time_step                           = 180,

max_dom                             = 2,

dx                                  = 30000, 10000,

dy                                  = 30000, 10000,

&chem

kemit                               = 2,

chem_opt                            = 2,        2,

biomass_burn_opt                    = 3,        3,

plumerisefire_frq                   = 30,       30,

When I run real.exe, I got the following information in rsl.error.0000 :

STARTING WITH BACKGROUND CHEMISTRY

  14 Subroutine input_chem_profile:

  13 d01 2016-11-12_12:00:00  input_chem_profile: exit subroutine

  12 READING EMISSIONS DATA OPT 3

  11 d01 2016-11-12_12:00:00  DEBUG wrf_timetoa():  returning with str = [2016-11-12_12:00:00]

  10  mediation_integrate: med_read_wrf_chem_fireemissions: Open file wrffirechemi_d01

   9 d01 2016-11-12_12:00:00  NetCDF error: No such file or directory

   8 d01 2016-11-12_12:00:00  NetCDF error in wrf_io.F90, line         1077

   7 -------------- FATAL CALLED ---------------

   6 FATAL CALLED FROM FILE:  <stdin>  LINE:    2907

   5  med_read_wrf_chem_fireemissions: error opening wrffirechemi_d01

   4 -------------------------------------------

   3 application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0

   2 [unset]: write_line error; fd=-1 buf=:cmd=abort exitcode=1

   1 :

2166 system msg for write_line failure : Bad file descriptor

Could you help me with that problem, please?

Another question is:

I'm also confused that, in one of my test run, without setting up any auxinput7 information in namelist.input, I could still get simulation result seems with fire emission input. Does model has default configuration for reading wrffirechemi* fires, please?

Thanks!!

Bo

[Gabriele Pfister]

Hello Bo,

it seems you are mixing up a few things in your namelist. 

First, FINN emissions processed through fire_emiss should really only be used for MOZART chemistry options. This is, because when using the online plumerise model, WRF-Chem expects as input the smoldering component only whereas FINN provides the total emission strength. For the MOZART chemistry options this is accounted for within the code, but has not been done for other chemistry options. Please see the README to fire_emiss for information. 

Second, I see you are running with chem_opt=2. Unfortunately we cannot provide support for chemistry options other than the MOZART schemes and I will have to refer you to wrfchemhelp. I am not familiar with the other chemistry options, but it seems from your namelist that you are running with a non-KPP chemistry option (2, RADM2), yet set emiss_opt=5 (GOCART RACM_KPP emissions). I suggest looking through the WRF-Chem Tutorials and Tutorial presentations (https://ruc.noaa.gov/wrf/wrf-chem/Tutorial.html) to see examples of how the namelist needs to be configured for different chemistry options. 

Gabriele 

--

================================
Gabriele Pfister

Atmospheric Chemistry Observations & Modeling

National Center for Atmospheric Research

Email: pfi...@ucar.edu

Phone: +1 303 497 2915

Web: https://staff.ucar.edu/users/pfister

[Bo Wang]

Hi Gabriele,

Appreciate your prompt reply so much! That's clear enough and very helpful.

Indeed, I felt somewhat conflicting when I setting up the namelist.input to use FINN.

Now I realized that I need to compile WRF-Chem with KPP library to use MOZART chemistry options. I'll do that and let you know if any other questions.

Thanks again!

Bo

[Bo Wang]

Hi Gabriele,

Thanks for your help last time!!

It read in wrffirechemi* datafiles (from FINN) successfully based on your comments--I recompiled WRF-Chem with KPP, compiled MOZART,  and modified namelist.input,

Now a quick question for using the antro_emiss utility:

Besides global inventories (such as EDGAR-HTAP) on lat/lon projection, can I use it to create WRF-Chem ready emission files from NEI 2011 inventory on Lambert projection, please?

I could put the question in the anthro_emiss discussion group if need. Thanks for your time!

Bo

  

[Gabriele Pfister]

Hi Bo, 

I am sorry, but anthro_emiss is not setup to work on the NEI 2011 emissions that are provided with WRF-Chem. 

Gabriele

[Gabriele Pfister]

Bo,

PM2.5 and PM10 are diagnostic species and are calculated by summing up individual aerosol components. So you will have to increase all aerosol components.
You can look at the module module_gocart_aerosols.F for more information.

Gabriele

On Thu, Dec 6, 2018 at 12:03 PM Bo Wang <bw0...@uah.edu> wrote:

Dear Gabriele,

May I ask a question about how PM2.5 and PM10 works in MOZCART scheme, please? Are they passive tracers in GOCART aerosol scheme (I think so by reading materials and the module code in WRF-Chem), please? If not, how would they participate into chemical reactions and influence O3 formation?

I asked this question because: Recently I started fire emission correction to improve the model performance, after I increased PM2.5 and PM10 in wrffirechemi* files (I haven't changed oc, bc and sulf yet), the previously underestimated O3 was increased and got closer to observation (compared to surface station and ozonesonde). Other changes include enhanced PM2.5 and PAN, decreased NO and DMS in WRF-Chem output.

I'm trying to make my question simplified to answer. But please let me know if any other need to make it clear.

Thanks for your time!

Best,

Bo

On Sun, Sep 30, 2018 at 12:31 PM Bo Wang <bw0...@uah.edu> wrote:

Gabriele,

Thank you very much for the prompt reply!

Your suggestions are important to me. I'll try to follow it to get a clearer picture.

Bo

On Sun, Sep 30, 2018 at 11:38 AM Gabriele Pfister <pfi...@ucar.edu> wrote:

Bo, 

all of the options you mention could be a reason plus uncertainties in the satellite retrievals. You have to do some deeper analysis to try and find the cause and also compare to other data products (e.g. AERONET, MISR plume height, other satellite AOD retrievals, surface PM, ...) to get a clearer picture. 

GOCART does not simulate SOA or nitrate aerosols. You can try and use MOZART with MOSAIC aerosols and see if this makes a difference. I would think though that specifically near and over the fire locations the contribution of SOA is small and nitrate aerosols from fires are also generally smaller. 

Gabriele

On Fri, Sep 28, 2018 at 3:32 PM Bo Wang <bw0...@uah.edu> wrote:

Hi Gabriele,

I hope everything is going well with you! Thanks for the help on your IDL code (to go from NEI 2011 to wrfchemi files) early this year very much. I'm using the code with WRF-Chem to help my PhD research about wildfire impacts on ozone production.

Now I'm confused with the underestimation of simulated AOD on a fire case using FINN inventory.It would be greatly appreciated if you could give some suggestions for me to improve the model performance.

Attached slides are the AOD comparsion between WRF-Chem and MODIS, and the model configuration I'm using. The namelist.input file has been attached as well. The comparison results show that AOD near the fire area (around smoky mountain) are underestimated by 0.5 or more. 

Would any of the following parts cause that large underestimation, please? 

(1) using GOCART aerosol scheme (instead of MOSAIC on the fire study)

(2) possible mistakes in namelist.input, especially on the biomass burning and plume rise options.

(3) possible fire emission underestimation by Finn inventory

Please let me know if any other information needed. Thanks for your time!

Best regards,

Bo Wang

Graduate Research Assistant

University of Alabama in Huntsville  

On Mon, Jan 29, 2018 at 11:05 AM Bo Wang <bw0...@uah.edu> wrote:

Got it! Appreciate!

Bo

On Mon, Jan 29, 2018 at 11:00 AM, Gabriele Pfister <pfi...@ucar.edu> wrote:

I added them to the ftp side. 

On Mon, Jan 29, 2018 at 9:55 AM, Bo Wang <bw0...@uah.edu> wrote:

Good morning, Gabriele!

I'm writing this email to ask for your further help on the IDL code if possible.

I'm looking at the IDL code your shared (to go from NEI 2011 to wrfchemi files), when I find 3 subroutines missing in ftp://ftp.acom.ucar.edu/user/pfister/FRAPPE/Emily/:

ijll_lc.pro 

llij_lc.pro

create_netcdffile_mozmam.pro

Do you have those subroutines available to share, please? 

I could only search out ijll_lc.pro and llij.pro in the following link:

https://svn-dares-dart.cgd.ucar.edu/DART/tags/wrf-chem.r11690/models/wrf_chem/run_frappe_diagnostics/

Are they the exact subroutines used in prepemis_gocart_2domains_NEI2011_addedVOC.pro ?

Thanks for your time!

Best,

Bo

On Wed, Jan 24, 2018 at 1:03 PM, Gabriele Pfister <pfi...@ucar.edu> wrote:

Hi Bo, 

I can share IDL code with you that I am using to go from NEI 2011 to wrfchemi files. It is not code I want to have public but happy to share with individuals. Hence I did not want to reply to the Discussion Forum. 

 

You can get the code from:

ftp://ftp.acom.ucar.edu/user/pfister/FRAPPE/Emily/prepemis_gocart_2domains_NEI2011_addedVOC.pro

and the needed subroutines are under:

ftp://ftp.acom.ucar.edu/user/pfister/FRAPPE/Emily/

You will have to adjust the code for your needs and make sure that you are using the correct species assignment based on the chemical scheme you are using. 

Gabriele

--

================================
Gabriele Pfister, Deputy Director

Atmospheric Chemistry Observations & Modeling

National Center for Atmospheric Research

Email: pfi...@ucar.edu

Phone: +1 303 497 2915

Web: https://staff.ucar.edu/users/pfister

--

================================
Gabriele Pfister, Deputy Director

Atmospheric Chemistry Observations & Modeling

National Center for Atmospheric Research

Email: pfi...@ucar.edu

Phone: +1 303 497 2915

Web: https://staff.ucar.edu/users/pfister

[elham baranizadeh]

HI Gabriele,

from your email above you mention that "First, FINN emissions processed through fire_emiss should really only be used for MOZART chemistry options..." . is this still the case? I didn't find any indication of this from the user guide or documents online. Please let me know. thanks so much!

-Elham

[Gabriele Pfister]

Dear Elham

to my knowledge the fire emission scaling has not been extended to any other chemistry options aside from the MOZART schemes. You can check on ./chem/module_plumerise1.F to see how the scaling is done and if needed duplicate what is done for MOZART Schemes for another scheme you want to use. 

I also suggest including ebu_* fields in the model output and comparing these to the input fire emissions to ensure the total emissions agree. 

Gabi

--

================================
Gabriele Pfister, Deputy Director

Atmospheric Chemistry Observations & Modeling

National Center for Atmospheric Research (NCAR)

Email: pfi...@ucar.edu

Temporary Home Office Phone: +1 303 241 1329

Work Phone: +1 303 497 2915

Web: https://staff.ucar.edu/users/pfister