1. Carlos Hernández-Monteagudo CE F CA 1 CENTRO DE ESTUDIOS DE FÍSICA DEL COSMOS DE ARAGÓN (CE F CA) J-PAS 10th Collaboration Meeting March 11th 2015 Cosmology with (not so) accurate photo-z surveys

2. Outline The impact of an error in redshift: radial smearing of scales of the order/smaller than σ z A clear example in a CMB context: inverting the Poisson equation and producing an ISW map from upcoming A-PLUS & J- PAS surveys Reconstructing the velocity field with J-PLUS/J-PAS photo-z errors: kinetic Sunyaev-Zel’dovich studies? BAO reconstruction? Cosmology with A-PLUS ?? 2 J-PAS 10th Collaboration Meeting March 11th 2015

3. J-PAS 10th Collaboration Meeting March 11th 20153 σ z  σ r = c/H(z) σ z ~ 30 Mpc/h (σ z / 0.01) at z=0 r ~ 2π / k  k~2π/r ~ 0.2 h/Mpc Eisenstein et al. 2007

4. Carlos Hernández-Monteagudo4 The ISW arises in linear (large) scales if gravitational potentials evolve in time. In particular, in the LCDM scenario, when the accelerated expansion of the Universe makes linear gravitational potentials shallower: Due to the Laplacian present in the Poisson equation, the potential fluctuations arise at scales typically larger than the density field. This effect arises mostly in the redshift range z \in [0.2, 1.2] A couple of facts about the Integrated Sachs Wolfe effect (ISW)... This effect probes Dark Energy!! Teruel, June 16th 2010

5. Carlos Hernández-Monteagudo5 Evolving potential wells are populated with galaxies, clusters of galaxies and other probes of matter overdensities  there must exist a correlation between the ISW and those tracers Teruel, June 16th 2010

6. Carlos Hernández-Monteagudo6 ISW CMB Projected matter density (simulated maps)‏ (ISW arises at z \in [0.2,1.2])‏ Teruel, June 16th 2010

7. J-PAS 10th Collaboration Meeting March 11th 20157 Depending on the error in the photo-z’s, we may attempt to invert the Poisson equation and provide a map of the gravitational potential field, and then, an ISW map: OUR NEW APPROACH:

8. J-PAS 10th Collaboration Meeting March 11th 20158 Since the gravitational potential wells are coming, due to the k -2 factor in the Poisson equation, from large scales (> 100 Mpc/h), photo-z errors are NOT critical! (σ z = 0.015  σ r = 45 Mpc/h) This approach may/must be applied to A-PLUS! How important are photo-z errors ? For Planck, we did try with the 2MASS Photometric redshift catalogue (2MPZ), having σ z = 0.016

9. J-PAS 10th Collaboration Meeting March 11th 20159 The 2MASS Photometric Survey (2MPZ) σ z = 0.016

10. J-PAS 10th Collaboration Meeting March 11th 201510

11. J-PAS 10th Collaboration Meeting March 11th 201511 The 2MPZ ISW map σ z = 0.016Very shallow survey!

12. J-PAS 10th Collaboration Meeting March 11th 201512 The 2MASS Photometric Survey (2MPZ) σ z = 0.016

13. J-PAS 10th Collaboration Meeting March 11th 201513 Since we are looking at linear scales, one can easily assume Gaussian statistics for the fluctuations and conduct easy simulations in order to test for the errors of this approach. In our case, we conducted two sets of simulations: One can generate an initial density field from a fiducial Power Spectrum One then inverts the Poisson equation and computes the corresponding potential field One integrates along the line of sight and obtains an ideal ISW map For the SAME initial Gaussian density realisation, one also generates a galaxy distribution (using normal or log-normal distribution, it does not really matter), following the same selection function and shot noise level as the real catalogue, augment the empty regions (e.g., Poissonian augmentation), invert the resulting galaxy map into a density and potential map, and finally produces a realistic ISW map. One can repeat this for many mock Gaussian/log-normal realizations … Ideal sims Realistic sims

14. J-PAS 10th Collaboration Meeting March 11th 201514 Recovered vs predicted power spectra: Compared RMS of power spectra: Photo-z errors have little impact on this plot! Ideal sims Realistic sims

15. J-PAS 10th Collaboration Meeting March 11th 201515 For each simulation, how do the ideal and the realistic ISW maps compare? Correlation coefficient: /SQRT( ) Could A-PLUS reconstruct the ISW in redshift shells up to z~0.3—0.4 ?

16. J-PAS 10th Collaboration Meeting March 11th 201516 ANOTHER EXAMPLE: RECOVERING VELOCITIES WITH PHOTOMETRIC SURVEYS

17. J-PAS 10th Collaboration Meeting March 11th 201517 Statistical studies of the kinetic Sunyaev-Zeldovich effect (kSZ) Studies of the RSDs in 2-point functions Reconstruction of the BAOs ANOTHER EXAMPLE: RECOVERING VELOCITIES WITH PHOTOMETRIC SURVEYS

18. J-PAS 10th Collaboration Meeting March 11th 201518 How accurate does this inversion work for photometric surveys?

19. J-PAS 10th Collaboration Meeting March 11th 201519 How accurate does this inversion work for photometric surveys? Perfect, J-PAS like, A-PLUS like Correlation coefficient: /SQRT( ) Millennium sim.

20. J-PAS 10th Collaboration Meeting March 11th 201520 J-PAS should be able to recover the velocities and conduct BAO reconstruction With A-PLUS we may detect some RSDs on larger scales if we sample enough volume: Next step: testing in realistic mocks Eisenstein et al. 2007

21. J-PAS 10th Collaboration Meeting March 11th 201521 Photo-z non-critical problems: Angular clustering, f NL if systematics under control – mostly due to QSOs and high-z ELGs (can we get them right?? – if so, f NL constraints would be among the most competitive from LSS!) Cluster catalogue, magnification bias, number counts tSZ on clusters and groups ISW x-correlation and reconstruction Cross-correlation with Planck CMB lensing map  *independent* bias estimation for galaxy populations like QSOs or ELGs ? Photo-z accuracy critically dependent problems: BAO measurement (seems hopeless for A-PLUS) RSD measurement (only on the largest scales ? ) – we would need large volume sampling kSZ & diffuse baryon studies on those larges scales ?  look at the sims!! Open questions: Should we right away combine A-PLUS with other all sky surveys like WISE? Would we gain significant photo-z accuracy? What about using the shift among filters as different filters (A.Orsi, Txitxo)? What about using the difference of two shifted filters? Is that not a two very narrow filter for very bright objects? What cosmology could we do with very narrow redshift shells? More brainstorming at the Cosmology Group meeting … COSMOLOGY WITH A-PLUS ?