Pushing the envelope of lensing with resolved kinematics Eric Huff - - PowerPoint PPT Presentation

Pushing the envelope of lensing with resolved kinematics

Eric Huff Tim Eifler, Elisabeth Krause, Chris Hirata, Matt George, David Schlegel

Kinematics break degeneracy between shape and shear

face-on image rotation curve face-on, but sheared inclined, but not sheared

γ× = 0.2

Shear changes the orientation of an ellipse But shear has no solid-body rotation component.

Lensing mis-aligns the kinematic and photometric axes

−2 −1 1 2 x (arcsec) −2 −1 1 2 y (arcsec)

qint = 0.75, γ+ = 0, γ× = 0

−60 −45 −30 −15 15 30 45 60 v (km/s) −2 −1 1 2 x (arcsec) −2 −1 1 2 y (arcsec)

γ× = 0.2

−60 −45 −30 −15 15 30 45 60 vLOS (km/s)

Consider the Tully-Fisher relation.

Reyes et al 2011 Reyes et al 2011 Schlegel (private comm.)

With spectroscopy, the Tully-Fisher relation tells us the inclination angle.

Blue points: not corrected for inclination Red trendline: TF relation, which we treat as given

For a disk, sin(i) tells us what ellipticity we should measure in the absence of lensing.

log [sin(i)]

Shear messes up the inclination correction.

Tully-Fisher: For a disk: sin(i) =

✓ e 1 + e ◆ 1

2

The effect of a shear:

e 7! e + γ vobs = vTF sin(i) + σTF sin(i)|γ = sin(i)|γ=0 + γ 2 p e(1 + e)3

2

The reduction in shape noise can be very large… …For face-on disks, factors of 10.

normal shape noise

σeff = 4.4 p e(1 + e)3σTF

A spectroscopic weak lensing measurement with slit spectroscopy

−2 −1 1 2 x (arcsec) −2 −1 1 2 y (arcsec)

γ× = 0.2

−60 −45 −30 −15 15 30 45 60 vLOS (km/s)

A spectroscopic weak lensing measurement with slit spectroscopy

−2 −1 1 2 x (arcsec) −2 −1 1 2 y (arcsec)

γ× = 0.2

−60 −45 −30 −15 15 30 45 60 vLOS (km/s)

Less rotation along the major axis than TFR would predict

A spectroscopic weak lensing measurement with slit spectroscopy

−2 −1 1 2 x (arcsec) −2 −1 1 2 y (arcsec)

γ× = 0.2

−60 −45 −30 −15 15 30 45 60 vLOS (km/s)

More rotation along the minor axis than TFR would predict

Simulating the measurement: Slit Spectroscopy

Keck-DEIMOS June 30, 2014 simple galsim-based simulation ———————- consistent with DES/BigBOSS estimates

Simulating the measurement: Slit Spectroscopy

shear (+) shear (x) extremely crude Fisher estimate:

• 8m telescope
• 1000 s
• Paranal sky, atm

———————- gain factor of ~30 measurement precision

• ver shapes alone

Simulating the measurement: fiber Spectroscopy

shear (+) shear (x)

(work ongoing)

For this level of per-galaxy shape noise:

Shape noise: ∝ σe √ngal For LSST: ngal ≈ 25 gal arcmin−2 σe ≈ 0.2 For kinematic lensing, equivalent shape noise with: σe ≈ 0.025 ngal ≈ .25 gal arcmin−2

ngal ≈ .25 gal arcmin−2 This is achievable with SuMIRe/PFS or DESI ∼ 103 deg2 = ⇒ 107 spectra

• Photo-z’s
• Intrinsic alignments
• Shear measurement
• PSF correction

None of the usual lensing systematics matter

We’ll have spectra for every source galaxy

• Photo-z’s
• Intrinsic alignments
• Shear measurement
• PSF correction

None of the usual lensing systematics matter

Intrinsic alignments don’t contribute (at ~leading order) to the kinematic signal

• Photo-z’s
• Intrinsic alignments
• Shear measurement
• PSF correction

None of the usual lensing systematics matter

Low means we can target bright, resolved galaxies

¯ ngal