Alexey-Ben meeting 2018-10-01

[ashiklom]

Alexey's goals from last week:

• Frontiers paper revisions (10%) – Done. Waiting on co-authors
• PEcAn (~40%)
  • Finish documentation PR – done.
  • Solidify and document Docker workflow for developers – lots of Docker documentation written, and general workflow outlined. But not yet working. Hit some strange technical hiccups, but will try to resolve before Docker demo for Ben tomorrow (10/2).
  • Test model runs at UMBS and SERC – didn't get to
• FoRTE (~30%)
  • Background reading – yes. Read several papers by Gough, Hardiman.
  • Organize/share paper ideas – not organized or shared, but reading helped solidify some more ideas.
• Hector (~20%)
  • Background reading – didn't get to
  • Running different scenarios – didn't get to
• Other notes:
  • Attended Mike's NSF talk on Thursday. Ate up a lot of time getting to/from Alexandria, but was worth it for the inspiration!

Alexey's goals for this week:

• Submit multivariate traits paper to Ecology Letters (~6 hours)
  • Co-authors requested some minor revisions
• PEcAn
  • Solidify Docker development workflow (~6 hours)
  • Test model runs at UMBS and SERC (~16 hours)
    1. SIPNET (pre-installed, few input requirements)
    2. ED2 (needs vegetation initial conditions. Should be able to get them from FIA)
    3. PreLES, DALEC, MAESPA, Linkages (need to be installed, but should be easy)
• FoRTE
  • Background reading (~5 hours)
  • Outline first paper (~3 hours)
• Hector
  • Simple parameter sensitivity analysis for vegetation parameters (~4 hours)

[ashiklom]

We should also discuss plans for PIC nodes. Now that we're in FY19, can we move ahead? I think I just have to give the green light to Wendy. Briefly, we are buying 1 cloud "unit" with 4 CPUs, 64 GB RAM quota, 1 TB cloud storage. That runs us ~$200/month (~$2400/year). Once we're comfortable with the setup, we can look at adding HPC storage, and setting things up for "remote" execution (via the PIC queue).

Full discussion here: https://jira.pnnl.gov/jira/servicedesk/customer/portal/4/ICS-1919

[ashiklom]

Additional notes from meeting:

• Possible paper idea: Review paper on lineage of terrestrial ecosystem models
• Field work -- In the future, PNNL can rent me a car and hotel; space shouldn't be an issue.
  • Next trip will be tree girdling sometime around April 2019. Then, actual measurements that summer.
• PIC
  • We have the money, so can move ahead. Use the FoRTE charge code.
  • Keep Mina in the loop
• In general, OK to continue devoting much of October to infrastructure setup, etc. (though hopefully will not take all that time)
• First FoRTE paper
  • Alexey should follow up on Biogeosciences paper from Ben, Chris, others on modeling disturbance
  • Chris is the real data resource for UMBS -- chronosequences, FASET data, flux tower, etc.
  • Would also be interesting to look at 20th century CMIP5 outputs
  • Alexey's idea -- this paper would be a piece of a bigger-picture Nature/Science paper on modeling the biodiversity-productivity relationship, which is controversial in ecology
• Ben wants to start working on adding Biome-BGC to PEcAn this winter
• Tomorrow's PEcAn/Docker demo: Focus on Docker and how it works with PEcAn. Also, a deeper look at how PEcAn works, how to go about adding models, etc.
• Hector
  • Alexey's parameter sensitivity analysis sounds good
  • In general, OK to put project slightly on back burner for now
  • Corinne should start being back at least part-time in ~3 weeks. Good date to keep in mind for initial Hector tests.