2017-06-28

Present:

2017-06-15

Present: Anna ShlyaevaXin Zhanguser-11260, Bryan Flynt, Jing Guouser-cf056John MichalakesThomas AuligneYannick Tremoletuser-0955e

Anna Shlyaevauser-11260 and user-cf056 gave an overview of the existing steps for computing the obs-background for T, radiances and AOD. Their analysis is documented on the GSI Observation Operators page.

Anna Shlyaeva presented her thoughts about the interfaces that will be needed for EnKF:

 

2017-05-31

Present:  Anna ShlyaevaXin Zhanguser-11260Bryan Flynt Jing Guo Julie Schrammuser-cf056Yannick Tremolet

We discussed the preliminary interfaces required by JEDI and described here.

It was noted that it is not yet clear where data thinning, data selection or superobbing will go, should it be in the QC or earlier in the constructor of the obs_space or both?

The interface of the QC needs more details. It will be necessary to investigate what the inputs to the QC are in the GSI.

To define steps between the current polymorphic structure in the GSI and the future JEDI interface, it was decided to look at the precise data flow for a few observation types and how it would fit in the proposed structure for presentation at the next meeting.

Actions for next meeting:

2017-04-26

Present:  Anna ShlyaevaXin Zhanguser-11260 Bryan Flyntuser-23cc3 Jing GuoYannick Tremolet

The first goal of the first meeting was to introduce everybody.

Then user-23cc3 and  Jing Guo presented the work they have done to introduce polymorphism in the GSI observation operators:

  • No labels