At the DESC meeting, I chaired a meeting to help try to define standard algorithms for measuring various galaxy properties. There are many benefits to this.
- We all can speak the same language.
- Validation is simplified (although published measurements may have to be translated)
- The same tools can be used everywhere for calculating observables
- We have the luxury of choosing the best definitions for our situation. I.e. measurements need to be compared between observed (noisy) and model (optimal) realizations.
One point made by @jbosch is that when we write down a measurement (e.g. size) it’s very important to also write down the profile used to compute that measurement for models.
Following are some questions that came up and were not fully answered due to time.
- What do we do about the fact that the naive galaxy model is a linear combination of profiles and point sources?
- What definitions should we use and which measurements should be defined?
Following is a list that we came up with at the session.- Size – sqrt(a*b) where a and b are the semi-major and semi-minor axes of the ellipse enclosing half the light from the profile.
- Shape – b/a and position angle with b and a as defined above and position angle the angle of the semi-major axis relative to North increasing through East.
- variability – no standard was suggested
- weak lensing – complex shear (gamma) and convergence (kappa) for each object
Next steps are to define others and write up a document with recommendations. Concerned parties include @danielsf @jbosch and @connolly