WRF forcings: precipitation

[Sara B]

Dear WRF-Hydro users,

I run WRF-Hydro forced by WRF (i.e. FORC_TYP=3). My concern is about the precipitation forcing.
I read in the documentation the required forcings for WRF-Hydro (Table 1.1. and 5.4), and I looked in ../trunk/NDHMS/Routing/module_lsm_forcing.F in order to see exactly which variables of WRF are used.
Here, I could see that the variables RAINC and RAINNC (from the files wrfout* of WRF) are used, where 

RAINC = ACCUMULATED TOTAL CUMULUS PRECIPITATION

RAINNC = ACCUMULATED TOTAL GRID SCALE PRECIPITATION. 

Nevertheless, the precipitation rate should be the forcing according to the Tables.

Thus, I would like to ask if the model WRF-Hydro computes the precipitation rate from the accumulated quantities.

Then, I would like to know if it does matter if WRF run and WRF-Hydro run have different start date.

I am asking this because I am in the following situation. WRF-Hydro is forced for 4 years with a WRF simulation that is not a continuous one, instead, I have available four 1-year WRF simulations (of four consecutive years, let's say from 2015 to 2018). Therefore, accumulated quantities are set to zero every first of January. 

Since WRF-Hydro is forced from 01/01/2015 to 31/12/2018, should I sum the precipitation variables so that RAINNC and RAINNC are always accumulated quantities from 01/01/2015?

Thank you very much!!

Sara

[Arezoo RafieeiNasab]

Hi Sara, 

I personally have not used that option, so I am not fully sure if this answers your question or not but here is an explanation I got from a colleague. I have CC'd him in case you have follow up questions.  

I don't think that WRF-Hydro can directly ingest the WRF outputs when run continuously, especially for precipitation since WRF writes out the accumulated precipitation while WRF-Hydro requires hourly values and there is some preprocessing needed. You will need to process the WRFoutputs and create the LDASIN files that WRF-Hydro uses. The precipitation amount should be the sum of RAINC and RAINNC but for hourly values (if forcing timestep is going to be hourly). 

Hope this helps!

Arezoo

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/2ecfaca6-19bc-4884-a3b5-00492d4daae5n%40ucar.edu.

--

Arezoo Rafieei Nasab, Ph.D.

NCAR/RAL Project Scientist II

office: 303-497-2888

[Sara B]

Dear Arezoo,

thank you very much for your reply.

I am slightly worried about your sentence "I don't think that WRF-Hydro can directly ingest the WRF outputs when run continuously". 

I thought that this was exactly the case for FORC_TYP=3 (namelist.hrldas), although some preprocessing could be required for preparing some variables (like precipitation). What would be run "not continuously"?

As far as I could understand, one provides WRF output files (of type wrfout*) to WRF-Hydro when  FORC_TYP=3. In fact, in this case, .../trunk/NDHMS/Routing/module_lsm_forcing.F looks for RAINC and RAINNC of WRF (RAINRATE is instead used when *LDASIN* are given, like FORC_TYP=1 and 2). 

Thus, given the hourly WRF forcing for WRF-Hydro, do you think that RAINC and RAINNC should be the accumulated precipitation of 1 hour in mm (basically hourly precipitation in mm)?

Thank you very much again!

Sara

[Arezoo RafieeiNasab]

Hi Sara,

My apologies! I looked into the code, and you are right. The option 3, reads the wrfout_d0 files (meaning the spatial resolution is the same as WRF-Hydro), and pulls the RAINC and RAINNC from two consecutive time steps, sums the two (RAINC and RAINNC) to calculate the precipitation at a given time step. And then calculates the delta of the two consecutive run as follows: 

PRCP1=(PRCPnew-prcp0)/wrf_dt   !Adjustment to convert accum to rate...(mm/s)

PRCPnew is time step t and prcp0 is the t-1. So this answers your first question (Thus, I would like to ask if the model WRF-Hydro computes the precipitation rate from the accumulated quantities.). The answer is yes, it calculates it from the accumulated quantities. 

As for the second question, I have to look further to answer your question. 

Thanks!

Arezoo