Skies and Universes Johan Comparat (UAM CSIC IFT) SelGIFS workshop, Universidad autonoma de madrid november 2016
Skies and universes, rationale To overcome the cultural barrier and communicate or co-work efficiently, one needs a solid structure that in the case of astrophysics or cosmology can be reduced to a cluster of computers where data, code and resources are in open access and documented. It aims to be a very special place to connect ideas to data and theory in the most versatile, fluid manner. Currently exist many high-quality databases related to telescopes and virtual observatories: www.sdss.org, http://vizier.u-strasbg.fr/, etc. But these are not directly connected with scientific code repositories, theoretical predictions nor computing facilities. This multiplication of interfaces between the scientist and these three entities, leads to important time losses. Such a co-working space also enables better reproducibility and easier teamwork. 11/21/2016 J. Comparat 2
http://projects.ift.uam-csic.es/skies-universes/ Collaborators: J. Comparat F. Prada G. Favole S. Rodriguez-Torres P. Zarrouk E. Burtin data simulations C. Maraston V. Gonzalez-Perez code 11/21/2016 J. Comparat 3
Software & Database DATA RELATED Class GalaxySurvey{SurveyName} SCIENCE RELATED • Overall calibration methods Class LineLuminosityFunction • Converts flux => luminosities Classes • derives LF from a catalog: completeness, weighting, uncertainty Class GalaxySpectrum{SurveyName} Class ModelLuminosityFunction • Opens spectra • fitModel: Schechter or Saunders • Spectra calibration • plotting Libraries • filterList • Decam, megacam (cfht), sdss broad band filter curves Libraries • LineFittingLibrary • Integrates spectrum, fits emission line (single or doublet) models (Gaussian, Lorentzian, pseudoVoigt) • lib_plot • Produce standard plots from the catalogs • lineListAir / Vac • Line list to consider. Depends on pyneb 11/21/2016 J. Comparat 4
Emission-line-luminosity-selected galaxy samples Based on Comparat et al. 2016. MNRAS. Arxiv 1605.02875 11/21/2016 J. Comparat 5
DATA • Use major spectroscopic magnitude limited surveys that cover more than a square degree • Fold data into a common data base (skies and universes) to enhance control over flux limits determination, weighting and completeness 11/21/2016 J. Comparat 6
DATA 11/21/2016 J. Comparat 7
Workflow • flux calibration for DEEP2 spectra • fit emission line to spectra (VVDS, DEEP2, VIPERS) • create emission line catalogs DATA • convert observed flux into luminosity • Compute TSR and SSR • Determine magnitude limit and flux limit • Estimate line luminosity function SCIENCE • Convert references to Planck cosmology • Fit Saunders and Schechter models 11/21/2016 J. Comparat 8
Fitting Example Signal to noise is low therefore convergence is tricky. It depends on * resolution * redshift * line flux * estimation of the continuum * estimation of the uncertainty on the flux * … 11/21/2016 J. Comparat 9
Results, [OII] LF 11/21/2016 J. Comparat 10
[OIII] 5007 LF 11/21/2016 J. Comparat 11
H beta LF: new ! 11/21/2016 J. Comparat 12
Fit results 11/21/2016 J. Comparat 13
Results • Skies and universes started a year • Emission line LFs ago • The models plugged-in to an analysis allow comparison to any observed • It hosts a few projects: field: “where do I stand with respect • 2 published (1507.04356, to the mean density of line emitters.” 1605.02875), the latter is 100% end- • Great tool to plan observations, to-end reproducible. understand and forward model • 2 in the referee process (1611.05457) completeness • 4 in preparation • In preparation • And many more to come :-) • relation between line LF and stellar • New collaborators are welcome ! populations models • Conditional emission line LFs 11/21/2016 J. Comparat 14
Exercise Arxiv: 1507.07979 11/22/2016 J. Comparat 15
Fitting emission line models • In the folder “ELG - • 1. With the composite spectrum: composite ” taken from • 1.a. Fit emission line models to the the [OIII] 5007 line existing Zhu, Comparat et al. in the composite spectrum provided here : arxiv:1507.07979. It • 1.b. plot the best emission line models found along with the contains : data, discuss the chi2 / degrees of freedom found. • 1.c. Artificially increase / decrease the uncertainty on the flux • High SNR ELG stack: (errors) and look at how the line model parameters vary. average properties of Discuss the importance of the error. observed ELGs • 1.d. Fit models to other lines: [OII] doublet, H beta. Discuss • Two scripts the challenges corresponding to each lines. • Three data files • 2. With the stellar population model (SPM): • Open and read “pysu_tutorial.py” • 2.a. compare the SPM spectrum provided and the composite spectra. Is it a reasonable model ? • 2.b subtract the SPM from the composite and fit emission line models to the residual spectrum. Be careful with the estimation of the errors on the residual spectrum 11/21/2016 J. Comparat 16
11/22/2016 J. Comparat 17
Recommend
More recommend
