[Wrf-users] Long-run Simulation

Mickias Kebede ws.micky at gmail.com
Thu Feb 23 13:07:09 MST 2017


Hello! Garg,

Thank you very much for your email. How do I run the real without having
the met_em files from metgrid?  And when will the model use my
boundary forcings (i.e. NCAR CESM Global Bias-Corrected CMIP5 Output to
Support WRF/MPAS Research)?

Thanks again for your help,

Kind regards,

Mickias,

On Thu, Feb 23, 2017 at 12:13 PM, Nikhil Garg <nikhilgarg.gju at gmail.com>
wrote:

> i suppose you dont need to worry about metgrid, and also about real, you
> can create wrfbdy file for batches, as it only stores boundary condition.
>
> you have to run real.exe few times for each set of simulation and then
> copy the files from real.exe output to separate directories, so that you
> dont overwrite them.
>
> also, if you dont want to use initial conditions again and again, and have
> a continuous simulation despite doing it in batches, possibly "restart" is
> what you want, just store the restart file at the end of each run, and
> remove the old wrfbdy file and put new wrfbdy and start the run with
> restart file.
>
> it will be a continuous run. you can write a shell script (or perl or
> python) depending on what you prefer, to do the whole batch run, where the
> script
> can modify the namelist entry, copy files around etc
>
>
> On 24 February 2017 at 01:26, Mickias Kebede <ws.micky at gmail.com> wrote:
>
>> Hello! Tabish,
>>
>> Thank you very much for your email. Yes, I've used the model output from
>> NCAR to enforce my model (NCAR CESM Global Bias-Corrected CMIP5 Output to
>> Support WRF/MPAS Research) and I'm interested in simulating from 1970 to
>> 2005 for the past and do the same in the mid-century. No need to do
>> ungrib in this case as the data is already in intermediate format and I've
>> downloaded the data for the entire time frame. My concern is, for that
>> one-year successful simulation I did, the size of wrfbody is about 20 GB
>> and I'm concerned how big it's going to be for the entire 30 years? I'm
>> wondering if there is a way to break the 30 years simulation into fewer
>> number of years such as three 10 years while I run the metgrid.exe and
>> real.exe?
>>
>> Thank you very much again and any help would be highly appreciated.
>>
>> Kind regards,
>>
>> Mickias,
>>
>>
>>
>> On Thu, Feb 23, 2017 at 9:35 AM, Tabish Ansari <
>> tabishumaransari at gmail.com> wrote:
>>
>>> Hi Mickias,
>>>
>>> I assume that you had your initial and boundary conditions for
>>> meteorology for the one year that you ran the model for, in your wrfinput
>>> and wrfbdy files. You must have produced these files through ungrib and
>>> metgrid utilities, and must have downloaded a reanalysis or global model
>>> data like FNL/GFS or ECMWF to create these files.
>>>
>>> You cannot run a regional model (WRF) without initial and (more
>>> importantly) boundary conditions. So you'll have to first run a global
>>> model for the 30 years past and 30 years future, or alternatively contact
>>> someone who has run it and has the output available for that period. Then
>>> use that data (instead of FNL etc.) to create new wrfinput and wrfbdy
>>> files. This would require a fair amount of restructuring and reformatting
>>> of data if you want it to be processed by WPS. Once you've produced your
>>> wrfinput and wrfbdy files for the required period, you can go ahead and run
>>> WRF.
>>>
>>>
>>> Good luck!
>>>
>>> Tabish
>>>
>>> Tabish U Ansari
>>> PhD student, Lancaster Environment Center
>>> Lancaster Univeristy
>>> Bailrigg, Lancaster,
>>> LA1 4YW, United Kingdom
>>>
>>> On 23 February 2017 at 10:53, Mickias Kebede <ws.micky at gmail.com> wrote:
>>>
>>>> Hello!
>>>>
>>>> I did a successful one-year simulation of three different parallel
>>>> cases of simulations. I would like to do the same three parallel runs but
>>>> this time with 30 years in the past and 30 years in the future. Can I just
>>>> adjust the dates in my namelist.wps to my 30 years span timeframe and run
>>>> geogrid and do the rest of the running processes as one continuous
>>>> simulation? Or is there any another trick that I should do perhaps to break
>>>> the long 30 years simulation into smaller such as three 10 year simulations?
>>>>
>>>> Which one is the most effective way of doing such type of long
>>>> simulations?
>>>>
>>>> Thank you very much and any help would really be appreciated.
>>>>
>>>> --
>>>>>>>>
>>>>
>>>>
>>>>
>>>> *Best regards,Mickias*
>>>>
>>>> _______________________________________________
>>>> Wrf-users mailing list
>>>> Wrf-users at ucar.edu
>>>> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>>>>
>>>>
>>>
>>
>>
>> --
>>>>
>>
>>
>> _______________________________________________
>> Wrf-users mailing list
>> Wrf-users at ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>>
>


--
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20170223/d03e43da/attachment.html 


More information about the Wrf-users mailing list