11:00 Eastern; 08:00 Project/Pacific Time. bluejeans.com/426716450

Attendees

Notes

Merlin Fisher-Levine

  • Finished work on the ComCam calibration products.
    • Including good test coverage.
    • Now being looked at by Andrés.
    • This includes a good defect task, but needs more work on the PTC task.
    • Plan:
      • Split DM-19709 into multiple tickets, one which covers defects, and another which covers PTC.
      • Then have the PTC & linearity work on a new ticket (or two new tickets: the simple version, and the fancy version) which Andrés can own.
      • Then close DM-19709, without code review, keeping the defects task on a branch for now.
      • Make another ticket to get the defects task review.
  • Now working through the backlog of atmospec tickets.
  • Will be going to DESC PCWG meeting in Paris next month; would like to have atmospec in a better shape before then
  • Following the DESC meeting, we should review plans/timelines.

Andrés Alejandro Plazas Malagón

  • Agrees with the Merlin-plan above; will be porting code from Pierres.
  • RHL has been pushing him to look at AuxTel data.
    • Following change of system configuration.
    • It would be good to chat with Robert to set up a set of goals for this.
    • John Swinbank will set up a call with Robert Lupton to discuss this.
  • DM-19336 can be marked as done, following comments on the ticket.
  • Resolving most outstanding AuxTel tickets required better data; maybe related to the RHL comments above.

Unknown User (aguyonnet)

  • Nick Mondrick visited Paris for 10 days.
    • Worked on Gemini observation with help from August on cloud coverage.
    • Also calibrating Harvard CBP.
  • Using Long short-term memory (LSTM) to model MERRA-2 time series.
    • This is a “recurrent neural network for deep learning”.
    • Using https://keras.io/
    • Seems to work; needs more input about the use-cases for this.
  • Is comparing MERRA-2 data with ERA-5, which is effectively the European counterpart of MERRA-2.
    • Still working to find/access all of the data.
  • Using MERRA-2 retrieval method for other observatories.
    • In other words, providing atmospheric transmission to other sites on request.
  • Plans to submit a proposal for a fellowship based on this work.
    • H2020. 
    • Will deliver a website where you input the observatory location & date, and get the atmospheric transmission.
    • Publication is vital for this; we must move ahead with the paper.
    • Has been informally asking for feedback from potentially referees.
    • Will present the paper to SNFactory and PCWG this week.
    • Will need to add a comparison with real data. A comparison with DES photometry looks excellent. Will work with the PCWG on this.