Skip to content

jhan_temp_UKMOtrip

Jump to bottom
JhanSrbinovsky edited this page Jun 3, 2019 · 2 revisions

NOTES for UKMO Trip

My email to Rachel, Iann, Danny - plans for MO: who to see-

  • Martin Best of course.

  • Richard Gillam

  • Douglas Boyd (we have met a while back but I think he had a different role)

  • Charlene (I don't believe I know Charlene. I know the name but I don't think we've ever met)

There have been other people involved with the Tickets processed so far. I'm hoping Richard might set me up with some of these people. The more discussions/opinions we can have the better. There is a session on Lfric and I am hoping to catch up with some people around this

I'm not exactly sure what is happening with Kerry, or a replacement. There are people I should speak with about navigating/enforcing their ticketing system. It will immediately affect us assoon s we upload CABLE (I suspect). Which is a good reason for finalizing as much as we can before it goes in - else every change may/will have to go through JULES etc.

Ian's response

A couple of topics that you may want to discuss

  • directory structure
  • documentation and timing [I wouldn’t be happy if we end up writing material specifically forJaC - other than around technical details of the implementation]
  • managing CABLE edits to the output routines (e.g. diagnostics_bl) some of the trickier technical stuff*

I’ve let Andy P know about the licencing quirk if that comes up in discussions.

*For example I’m thinking of the implications of CABLE running its soil moisture/temperature update under the implicit part whereas JULES runs the equivalents under the extra part. Depending on the standards/LFric there could be a fair bit of rearrangement to do there. (… and I suspect that Martin/Richard don’t appreciate the need for CABLE to run its soil schemes under implicit)

another topic that I remembered for you to discuss.

the confusion around parallelisation in standalone JULES – and consequently where CABLE declares and initialises its copies of the state variables

Clarify at committee meeting.

The licensing 'quirk'? My understanding was more or less resolved depending on official advice from here? I will endeavour to get a definite answer from them - however I htought their position was that they were happy for CABLE files to be covered by a CABLE license. A Crown license was not appropriate.

The updated soil moisture - i dont understand where the problem is? Thats how we do it now i.e. we do it in implicit, they do it in hydrol. Although - would there be any difference if we did it in hydrol. I mean it may suit JULES because thats where all there stuff is, and it suits us to do it in implicit because it makes sense for CABLE to update soil/snow and include this in returned fluxes etc. But assuming we could get everything we need to the hydrol pathway and we did just call CABLE again (particularly soil/snow routines) would anything be different? i.e. Is there any update to "forcing" between implicit and hydrol pathways? I dont know but I dont think that there is. So we could just unpack %wb etc? In addition to the unpacking it does now, which is all that it does.

Clone this wiki locally