moo select options

[LesleyGray]

Dear CMS Helpdesk

I'm trying to extract some zonal wind data using:

moo select —fill-gaps queryFile_U_1850 moose:/crum/u-bd839/ap6.pp /gws/nopw/j04/solar/lesley/temp_bd839

and my queryfile looks like this:
begin

stash=30201
year=[1858..1859]

end

It works fine, but the suite has output U in various different combinations e.g. with different time averaging, and on different numbers of model levels. So I get pp files that contain 5 different U fields.

If I xconv the pp file the fields all have the same title (U component of wind on p level) but have different numbers of nz levels. I only want the field that is on 39 levels. I've been making .nc files of this field using xconv but this is really laborious - I'm trying to extract 40+ years of daily data, in 1-yr chunks.

Is there a way to modify my query file to select only the field that's available on 39 levels? If I can just extract the one field that I want, then I could use some of your CF python scripts to convert to .nc.

I found this documentation ​http://artefacts.ceda.ac.uk/badc_datadocs/um/umdp_F3-UMDPF3.pdf that looked hopeful (e.g. list on page 28 with LBLEV options). I tried a few options but nothing worked - I don't really understand how these things work - and hoping you can help me.

thanks very much
Lesley

[grenville]

Hi Lesley

Sorry for the delay in replying. We don't have sufficient moo expertise/experience to provide a quick answer.

I shall forward your query to the MO team.

Grenville

[david]

Hi Lesley,

You could use a cf-python 2-liner to do this:

>>> import cf
>>> fields = cf.read('file(s)').select_by_construct('Z')
>>> f = cf.FieldList(x for x in fields if x.coordinate('Z').size == 39)

The first line picks out all the fields with Z coordinates, and the second line throws away all those without exactly 39 levels.

Hope that helps,
David

[andy]

Hi David,

This is the same issue that was reported a while back but awaiting a fix. cf-python cannot aggregate the fields properly and so you don't end up with four fields of u but many slices of data one levels. I think Sadie was going to look at this as per our Slack discussion on Oct 6th.

Cheers
Andy