Using ERA-40 to force the LAM

[jonathan.day@…]

Hi,
I'm a PhD student in Bristol. I'm aware that members of your department are using ECMWF ERA-40 data to create LBC files to force the limited area model. I would like to force the vn4.5 LAM in this way and was wondering if files of this type are obtainable from NCAS for a given region.

Many Thanks

Jonny

[willie]

Hi Jonny,

I can provide ECMWF ERA 40 start files if necessary. However, if you are using data from the past five years, Met Office start files are preferred. Normally I get the old ECMWF files and then reconfigure them for use with UM6.1, so I don't yet see why it shouldn't be feasible to do the same for UM4.5. I can get the Met Office files too.

If you could let me know the precise dates you need I will try to get the data. Also it would be good to know what project this is for.

Regards,
Willie

[umdoc]

Hi Willie,
My project is called 'High Resolution climate modelling of Svalbard's Glaciers'. But I am individually NERC funded rather than through a larger project.

I wish to force the LAM with ERA-40 over the period Sept 1957 to Aug 2002. I'm a little confused however as to how this has been achieved in Reading. Is it sufficient to simply have the ERA-40 start files for each time step to produce the lateral boundary conditions?

Cheers

Jonny

[lois]

Hello Jonny,

At Reading there are a couple of users who have used the rather ad-hoc system for doing this, they are Tim Woollings and Charlie Williams. Charlie wrote a local guide for doing this which you can see on the NCAS CMS web page here
​http://ncas-cms.nerc.ac.uk/content/view/372/169/
The paper is "Idiot's Guide to setting up and Running HADRM3P on HPCx"
You will need to log in to access these pages I will email you separately with this information.

You will see from Charlie's notes that it uses a collection of perl scripts, with various add-ons to UM vn4.5 (handedits, mods, environment variables) but it generates ancillaries and dumps using PRECIS (the PC version of HADRM3 for climate change scenarios).

You may want to consider if this is the best way modelling mechanism for your project. You could email Tim (t.j.woollings@…) or Charlie (c.j.r.williams@…) before you consider going down this route.

Lois

[umdoc]

Thanks Lois,
I'll get in contact with them directly.

Cheers
Jonny