#2701 closed help (fixed)

Forcing long runs with ERA-Interim data

Reported by: shakka Owned by: um_support
Component: Data Keywords: ERA-Interim, ERA-5, ECMWF, reanalysis
Cc: Platform: Monsoon2
UM Version: 11.1


Hi Willie,

I am about to embark on the mammoth task of running a 10-year hindcast over the Antarctic Peninsula. I'm downscaling from the global model at N512 to a 4 km nest, and would like to force the driving model with ERA-Interim/ERA-5 fields.

My intention is to run 48 hour forecasts, archiving only the first 24 hours, and initialising daily at 00Z, such that I get a constant time series of t+24 to t+48 forecasts.

I want to run the model for the period 2008-2017, so will need 00Z ERA data from 20071231T0000Z until 20171230T0000Z with the above set-up.

I understand that this is a rather large amount of data! Will it be possible to organise getting this? Also, is it possible to use ERA-5 reanalysis rather than ERA-Interim? I am keen to use the best available product (as long as it isn't too much of a headache to set up).

I have also spoken to James Pope (and read ticket #2295: http://cms.ncas.ac.uk/ticket/2295) who has done this before, but would you be able to offer some guidance on how to set up the UM at vn11.1 to work with ERA data?

Thanks (and apologies) in advance.

Change History (7)

comment:1 Changed 19 months ago by willie

Hi Ella,

Daily ERA5 start dumps for ten years works out at 3653 dumps each of 122.7MB or 438GB in total. I have started to get these. It will take about 13 days just to extract from the ECMWF archive and then will need to be transferred to Monsoon. So mid January to get the data.

Using the dumps with the model is not too difficult. You need to select GRIB as the type of dump rather than UM file.

These files have only eight fields, so the reconfiguration will fail with messages about missing fields. You then need to supply the field via an ancillary file or set the field to a fixed value. In some cases if you are not interested in the field you can switch the corresponding science off. You will probably have to do several failed reconfigurations to set up the model.


comment:2 Changed 19 months ago by shakka

Hi Willie,

That's less data than I thought it might be - great news. Is ERA-5 data a bit smaller than the corresponding ERA-Interim files then? I will have a go with setting up the recon - James has done this and briefly showed me yesterday so I can ask him to give me a hand with that too.

Thanks for sorting this so swiftly.

comment:3 Changed 19 months ago by willie

Hi Ella,

The first few files are now on Monsoon in /projects/nexcs-n02/frmy/data_xfer if you want to try setting up a model.

I am not sure about the relative sizes of ERA5 and Era-Interim files.


comment:4 Changed 19 months ago by shakka

How exciting! Thanks Willie, I will download those now.


comment:5 Changed 18 months ago by willie

Hi Ella,

You data is now ready in /projects/nexcs-n02/frmy/data_xfer. Please copy it to your own work space as soon as possible.


comment:6 Changed 18 months ago by shakka

Hi Willie,

Just to confirm - I have now copied all of these over.

Many thanks!

comment:7 Changed 18 months ago by willie

  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.