Consolidate and understand the findings from the current Y5 spatial variation results

[tms-epcc]

26/MAY/22

• have all the results for Y5 and it is all homogeneous
• the Y1 and Y2 data is not readily available and Ellen may need to run photoZs on the Y1 data. This will be computationally expensive so investigating ways to reduce the compute costs. This may be a substantially more amount of work than originally planned for.
• Using DC2 and DR 6, have examined properties such as galaxy number density, magnitude error and colour as functions of different observation systematics
• the results have been presented to DESC WGs for feedback
• Next steps
  • wrap up the results thus far in a document for others to have access to
  • examine with HSC data, since this is real rather than simulated, and check if results are reasonable
  • look at cross-correlation and see how the systematics affect that

[tms-epcc]

16/JUN/22

• Ellen has created a note explaining the analysis. This will form part of D3.6.1 (see #12)
• The note has been shared with the DESC photo-Z WG
• Regarding the previously indicated intention to examine with HSC data, there are now concerns about the possible extent of such an investigation.
• next steps
  • get comments on the note and complete the final analysis
  • An effect is observed with the idealised simulations. The real effect may be worse. This needs further discussions with WG.
  • still to do cross-correlation. This is actually part of the WP3.6.2 (ie #21).

[tms-epcc]

14/JUL/22

• unfortunately there has been no feedback on the analysis note from the members of the DESC photo-Z WG.
• Will discuss the way forward with the WG leaders
• To wrap up the task considering either
  • looking at real data, or
  • discussing with the receipient group the contribution to the RAIL package regarding magnitude erorr dependence on observational systematics.

[tms-epcc]

28/JUL/22

• discussed with WG convener whether to look at real data and the current status. Agreed to look at Y1 data since Y5 effects are such a concern. Want to check if same for Y1.
• have started looking at Y1 data, and ran on a small dataset.
• Plots from this have been sent to WG and @joachimi for feedback.
• Data suggesting that there is not an effect to worry about but need confirmation from WG.
• github repo (see https://github.com/lsst-uk/DC2-observation-systematics) has been updated with the documentation for Y5.
• if WG agree to wrap up then will add Y1 to same repo and documentation. This will be then a DESC note for others to refer to.
• Expect to wrap up and close ticket by September

[tms-epcc]

11/AUG/22

• conclusion for Y5 is the systematic effect is not so significant (ie. well within Rubin requirements for DESC) and so will focus more on Y1 ie #17.
• Now believe this ticket can be closed but will first confirm with @joachimi

[tms-epcc]

15/SEP/22

• @joachimi confirmed can close this ticket