3
$\begingroup$

I have earlier used FNL, CFSR, Reanalysis-2 data set as input in WRF. However, I am finding it difficult to download ECMWF data as there are so many variables. Can anyone with experience in ECMWF let me know from where to download ECMWF data for WRF?

I need ERA-Interim Project to input into WRF. I have the link, I just don't know which variables to download and how to input in WRF.

$\endgroup$
3
  • 1
    $\begingroup$ Can you get access to any machine near a data archive where the data are readily available (for example, CEDA in the UK)? $\endgroup$
    – gerrit
    Mar 22, 2017 at 11:56
  • 1
    $\begingroup$ forum.wrfforum.com/viewtopic.php?f=22&t=1001 I believe this answer on this forum is a starting point for those wanting to go on this road. $\endgroup$
    – user1066
    Mar 23, 2017 at 3:11
  • $\begingroup$ I've written some notes on downloading ERA20C data and inputting it into WRF here: dvalts.io/2016/12/23/ECMWF-data-in-WRF.html hopefully they may be adaptable to your needs. (and I might get around to posting in the form of an answer if I get time.) $\endgroup$
    – decvalts
    Dec 14, 2017 at 11:20

3 Answers 3

4
$\begingroup$

For any of the basic forecast runs (10 day, ensembles, monthly/seasonal forecasts), I do not believe ECMWF data access is general controlled to ECMWF.INT. Generally most data, including anywhere near enough to use in a WRF forecast, is only available for purchase at extreme fees (generally we're talking 5+ figures per year). The only option you might have is if you are doing your modeling for strictly research purposes, in which case you may qualify for a research license at reduced or free cost, or be able to access it from servers at the institution you are associated with.

It does appear that reanalysis data and various other specialized products are available free with an account at their website as described here. I believe reanalysis would only get you initial conditions which I don't believe is sufficient to run a mesoscale model like WRF(?), and that they are also poorer quality/resolution (please correct me if I'm wrong on any of these beliefs!!) Archived actual ECMWF output is still pay (see the bottom option on the license page).

I worked with ECM for a few years while producing content at StormVistaWxModels and definitely indeed found most things with them to be quite a challenge, limitation, and pity! At the time it was fairly abstruse to find what you needed or communicate with them, though it appears to have improved some.

$\endgroup$
4
  • 1
    $\begingroup$ OP has not specified what kind of ECMWF data he is wanting to supply as input to WRF. There are several categories. There is ERA interim and ERA 40 and then the forecast data and so on. $\endgroup$
    – user1066
    Mar 22, 2017 at 7:52
  • 2
    $\begingroup$ If one knows OPeNDAP you can get ERA data from UCAR as well provided you are affiliated with a university or research institution $\endgroup$
    – user1066
    Mar 22, 2017 at 7:57
  • 1
    $\begingroup$ @gansub Apologies, you're right. I completed overlooked that topic, sorry. I haven't worked with reanalysis... but tried to update to include mention of those details. Please do confirm if I have treated that topic correctly, thanks :-) $\endgroup$ Mar 22, 2017 at 15:53
  • 1
    $\begingroup$ RDA, from UCAR (rda.ucar.edu) may be a good site, if OP is in academia, or involved in research. $\endgroup$ Mar 22, 2017 at 16:51
1
$\begingroup$

Please go through the Variable table vtable.ECMWF . Download all the variables in that list. These are the minimum variables required to initiate WRF.

$\endgroup$
1
  • 2
    $\begingroup$ This probably could be a decent answer with a bit more detail and a less terse nature. $\endgroup$ Sep 14, 2017 at 10:11
1
$\begingroup$

Here is a Python Script showing how to download the required variables automatically. http://conorsweeneyucd.blogspot.com/2015/01/download-era-interim-data.html

$\endgroup$
2
  • 1
    $\begingroup$ Please expand your answer. As stated it is basically a link-only answer and it might be deleted as such. $\endgroup$
    – arkaia
    Sep 20, 2018 at 1:04
  • $\begingroup$ Urgh. Technically yes, this is a link-only answer, and as such ought to be deleted according to SE guidelines. But there isn't any point in "summarizing" or reproducing part of a Python script or the instructions for using it. This is a useful signpost, and I think we should retain it. $\endgroup$ Sep 23, 2018 at 15:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.