Opened 6 months ago

Closed 4 months ago

#2731 closed help (fixed)

Another data request

Reported by: shakka Owned by: willie
Component: Data Keywords: ECMWF, grib, SST, sea ice, ERA-Interim
Cc: Platform: Monsoon2
UM Version: 11.1

Description

Hi Willie,

I'm afraid I need to ask you for more data. I've been advised that I should use ERA-Interim at 6-hourly resolution for my hindcast runs. Am I right in thinking that these do not include sea ice and SST fields? Are you able to supply me with sea ice and SSTs separately?

As before, I would need ten years' data from 2008-2017, including a month's spin-up, so 20071201_00 to 20171231_18.

If it helps reduce the amount of data you need to download, I know that James Pope has daily ECMWF data for some of the period I require:

i.e. the period 20071201_00 to 20161231_00, at 24 hourly intervals

and that Andrew Orr has 6-hourly data, including sea ice and SSTs, from 20140701_00 to 20141031_18.

I could easily copy this from their workspaces to save time.

I know this is going to be a huge amount of data so I'm happy to do it in batches if this helps.

Best wishes,
Ella

Change History (22)

comment:1 Changed 6 months ago by willie

Hi Ella,

I've started to get the first year's worth. These will be ERA-I start dumps including SST and sea-ice fields. They work out at 55MB each, so the full ten years will be about 785 GB and will comprise 14,610 files.

Willie

comment:2 Changed 6 months ago by shakka

Thank you so much Willie, you're a hero.
Ella

comment:3 Changed 6 months ago by willie

Hi Ella,

The first years worth are now in /projects/nexcs-n02/frmy/data_xfer. Please copy them to you own work space as soon as possible.

Let me know how you get on and I will generate and transfer the rest.

Willie

comment:4 Changed 6 months ago by shakka

Hi Willie,

Thank you for this - I have now copied all of the first year's data over.

Just to confirm: because you have kindly provided me with ECMWF sea ice and SSTs, I shouldn't need to reconfigure the glm to update this from ancillaries, correct?
i.e. within glm_um → namelist → Reconfiguration and Ancillary Control → Configure ancils and intialise dump fields I should only have to add in 0,32 Sea ice depth and not reconfigure SST or sea ice fraction at all ?

Best,
Ella

comment:5 Changed 6 months ago by willie

Hi Ella,

In principle, yes but in this case it might be no. The problem is that the original ancillary file n768/seaice/hadisst_6190/v1/qrclim.seaice has three fields,

  • Fraction of sea ice in sea
  • ice edge in the ancillary file only (sic)
  • sea ice depth

so originally it would have been updating all three from the ancillary file. But you are only replacing one. Without looking at the code, it seems to me that you will need the other two fields to be in the start dump too, before you can switch the ancillary off. I'll ask an expert tomorrow.

Willie

comment:6 Changed 6 months ago by shakka

Hi Willie,

Thanks for this, I look forward to hearing back tomorrow.

Best,
Ella

comment:7 Changed 6 months ago by willie

  • Owner changed from um_support to willie
  • Status changed from new to assigned

Hi Ella,

Sorry for the delay.

You are currently configuring the fraction of sea ice and the sea ice depth from the ancillary file in u-be285. But you want to run with regular start dumps that contain fraction of sea ice, but not the ice depth, so the ancillary file is not really necessary provided the ice depth can be calculated.

This can be done on the reconfig → ancil options page: select the the AMIP II method of updating SST and sea ice. This will generate the sea ice depth from the ice fraction in the dump.

Willie

comment:8 Changed 5 months ago by shakka

Hi Willie,

Thanks for this (and apologies for late reply - I've been away). I have selected the AMIP 2 method of calculating sea ice depth. Do I still need to reconfigure the sea ice depth as I have previously done for other fields from the GRIB file by adding a section in glm_um> namelist> reconfiguration and ancillary control> configure ancils and initialise dump fields? When I try to run without I get an error immediately.

Cheers,
Ella

comment:9 Changed 5 months ago by shakka

Hi Willie,

Do you have any updates on the remainder of the data, or on the sea ice/SST reconfiguration issue?

Thanks,
Ella

comment:10 Changed 5 months ago by willie

Hi Ella,

What's the suite id?

The next year of data is currently being generated.

Willie

comment:11 Changed 5 months ago by shakka

Hi Willie,

Thanks for the update. The suite ID is u-be285.

Ella

comment:12 Changed 5 months ago by willie

Hi Ella,

The 2009 data is now on Monsoon in /projects/nexcs-n02/frmy/data_xfer/. Please copy it to your own workspace as soon as possible.

As to the sea ice/SST reconfiguration issue, I fear we are getting lost. I had a look and Andrew Orr's u-bf120 suite, which is quite similar to yours. His GRIB start dumps do not include SST or ice fraction. He does not use AMIP II processing. But he does have a separate ancillary file, in UM fields file format, with STASH 31,32,38 (ice fraction, ice depth and ice edge) at 12 hourly intervals.

While looking at your suite, I noticed that the GRIB reconfiguration ignores the SST field in the start dump, claiming that it cannot find it.

If you really need SST, then I can look at this further, but I suggest Andrew's method is the way forward.

Willie

comment:13 Changed 5 months ago by shakka

Hi Willie,

I was worried you'd say that! I was hoping to save myself some work, but that's alright: I will follow Andrew's method and hopefully will be able to make some headway. Thanks very much for the 2009 data - I will copy it over to my own workspace now.

Best,
Ella

comment:14 Changed 5 months ago by willie

Hi Ella,

See Andrews comment at ticket:2739#comment:34 regarding ancillaries and general running.

Willie

Version 0, edited 5 months ago by willie (next)

comment:15 Changed 5 months ago by shakka

Hi Willie,

Thanks for this - I've spoken to Andrew about this and will hopefully be able to borrow some code to produce some similar ancils.

Cheers,
Ella

comment:16 Changed 5 months ago by willie

Hi Ella,

UMDP 303 says that the GRIB reconfiguration doesn't read the SST in the GRIB file: it calculates it from the surface temperature field, but only if the sea ice field is present in the GRIB file.

Willie

comment:17 Changed 5 months ago by shakka

I see, so you mean that if I have sea ice and SSTs in the GRIB file, it ignores SSTs and then calculates it? i.e. that to get the UM to use ECMWF SSTs, I need to feed in separate ancils as Andrew has done?

Ella

comment:18 Changed 5 months ago by willie

Hi Ella,

The 2010 data is now on Monsoon in /projects/nexcs-n02/frmy/data_xfer/. Please copy it to your own workspace as soon as possible.

The sea-ice field in the GRIB start dump causes the generation of SST - the SST in the start dump is always ignored. So you need to copy Andrew's method.

Willie

comment:19 Changed 5 months ago by shakka

Hi Willie,

Thanks for the clarification on the SSTs/SIC. I have now copied over the 2010 ERA-Interim data.

Best,
Ella

comment:20 Changed 4 months ago by willie

Hi Ella,

Is it still useful to transfer the remaining 2011-2017 ERA_interim data?

Willie

comment:21 Changed 4 months ago by shakka

Hi Willie,

Andrew Orr has just informed me that he has already got a copy of all the ERA-Interim data from 1979 in /projects/polar, so I can point the model to that instead of copying it all to my workspace. I think to save disk space I will continue to do this. Thank you for checking!

Best wishes,
Ella

comment:22 Changed 4 months ago by shakka

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