Opened 7 years ago
Closed 7 years ago
#1379 closed help (completed)
Creating a new SST ancillary file
Reported by: | s1251469 | Owned by: | jeff |
---|---|---|---|
Component: | UM Model | Keywords: | ukca;ancillary;SST |
Cc: | Platform: | ARCHER | |
UM Version: | 8.4 |
Description
Hello,
I am wanting to replace the SST ancillary file in my job. I have created a netcdf of the daily SSTs I want to use.
on archer:: /home/n02/n02/s1251469/work/ANCILS/sst_cmip5_N96_90s.nc
I am having trouble creating an ancillary from this that will run.
I have not set the time in the netcdf which could be a problem but since you can set it in xancil I dont see why it should be. If I set the start date/time as 01/12/1981 00:00:00 and take the first 360 timesteps I get the errors output in /home/n02/n02/s1251469/output/xklee000.xklee.d14276.t112447.leave
If I set the start date as 16/1/0 as seems to be suggested by the errors in the previous output then I get the following errors
/home/n02/n02/s1251469/output/xklee000.xklee.d14275.t152003.leave
I am sort of guessing how to create these SST ancillaries. Is there a webpage which clearly explains? Otherwise perhaps someone could offer some guidance as to what I'm missing.
Many thanks,
Declan
Change History (3)
comment:1 Changed 7 years ago by jeff
- Owner changed from um_support to jeff
- Status changed from new to accepted
comment:2 Changed 7 years ago by s1251469
Hi Jeff,
Thanks for the ideas.
I have the ancils for ice and sst working now. I am still learning about how the times and dates in the umui job are used so any web info on ancillaries would be useful to know about.
Since things are working at the moment you are welcome to close this ticket.
Thanks again,
Declan
comment:3 Changed 7 years ago by jeff
- Resolution set to completed
- Status changed from accepted to closed
Hi Declan
I think you have created the ancillary file correctly in terms of the meta-data. I think there could be a couple of possible problems with the data which maybe causing your model to crash.
There could be a problem with coastal values, i.e. land only points being used as sea or partial sea points. It would be best to extrapolate the sea values over the land points like how its been done with the reynolds.sst file (/work/n02/n02/ukca/ANCILS/surf/reynolds.sst.avg2000).
There could also be a problem with sea-ice points, as you didn't create the sea-ice ancil file from the sst data there could be sea-ice points which have sst values which are too high to be ice this can upset the model. This can also happen the other way around. It's possible to create the sea-ice field from the sst field in xancil, if you do this you will have to look at the sea-ice field to make sure it is suitable.
Jeff.