Concept question

Things that do not fit in any of the other topics.

Important Note:
If a moderator feels that this topic belongs in one of the topics above, it will be moved but a note will be put in this forum pointing to its new location.

Concept question

Postby Mike189 » Mon Apr 07, 2014 7:47 am

Hi

I am really new to WRF and I have a general question regarding the way I would achieve a medium range forecast.
Assuming that I am using GFS forecasts in 384 hours ahead as inputs, in order to downscale them at the regional level of interest using wrf. What I will get from simulating a specific period will be the cold start of the model, right? Does this represent an actual forecast of this period? So can I achieve in this way a forecast from the already forecasted global data without going through any nudging techniques?

Thank you
Mike189
 
Posts: 1
Joined: Sun Mar 16, 2014 5:21 pm

Re: Concept question

Postby Phillip » Thu Apr 10, 2014 9:18 am

I am not sure if I understand what you mean. You want to do a regional forecast using GFS as input but your first timestep will not be the first timestep of GFS. For example:

GFS times plus:
00 h
03 h
06 h
...
192 h
195 h
198 h
201 h

Normally you would initialize WRF with the first step +00 h, but you want to use a different, later timestep.

Short answers:
Yes, you can do that. They should contain all the information you need.
No, you don't have to use nudging. In fact, you can't, because future observations do not yet exist.

Downscaling a global forecast as far as 10+ days in the future can surely be done, however I would not be sure what to expect from it, as smaller scale phenomena tend to be badly forecasted much sooner than the synoptic situation. And after more than 10 days you are almost guaranteed to start with a significant error in your synoptic situation, resulting in totally different weather on a smaller scale.

And using the last available timestep as input will only work if you use it to initialize your own global simulation. Otherwise you will not have any boundary conditions for you regional model.
Phillip
 
Posts: 46
Joined: Thu Jul 19, 2012 3:29 am


Return to Miscellaneous

Who is online

Users browsing this forum: No registered users and 2 guests

cron