Christoph Becker (Durham/ICC)
Nan Li (Nottingham)
2019.06.11
(e.g. Sekiguchi+10, Linder11, Freedman+12, Weinberg+12)
E.V.Linder 11
(e.g. Sekiguchi+10, Linder11, Freedman+12, Weinberg+12)
shsuyu.github.io/H0LiCOW/site/, 2019
Independent methods are needed to overcome systematics, especially the unknown unknowns.
Variable sources: e.g. SNe, Quasars
Advanteges:
Time Delay
Time Delay Distnace
Lens potential
\( D_l \)
\( D_s \)
\( D_{ls} \)
Observable
Goal
Inferable
More l.o.s. structures         larger \( {\color{blue}\kappa_{\text{ext}}} \)         larger \(  {\color{blue}D_{\Delta t}} \)
Observable
Goal
Inferable
Angular diameters are perturbed by large scale structure relative to the homogeneous prediction
LSST will find 3000 strong lensed quasars with time-delays
(Oguri&Marshall 10)
LSST will find 3000 strong lensed quasars with time-delays
Impossible to have follow up wide, deep, spectroscopic imaging for all of them!
(Oguri&Marshall 10)
1. Construct light-cone with & without LOS-structure
3. Lens mass modelling
2. Ray-tracing through the light-cone
4. \( H_0 \) estimation
Semi-Analytic model cosmoDC2*:
*github.com/LSSTDESC/cosmodc2
   \( z_{\text{obs}} \)= 0.0         \( z_{\text{lens}} \)= 0.5         \( z_{\text{src}} \)= 2.0
Oguri & Marshall 2010
   \( z_{\text{obs}} \)= 0.0         \( z_{\text{lens}} \)= 0.5         \( z_{\text{src}} \)= 2.0
* Nan Li + 16
Gravlens*
SIE
(10 param.)
Glafic**
Softened Power-Law (SPL)
(11 param.)
lens-model:
cross-validate
*Gravlens by Keeton, **Galfic by Oguri
input:
perturbation: Â Â Â Â constant shear
image-positions, -time delays, -magnifications, redshifts
errors on input: Â Â assume best observation conditions (sub-%)
prior on \( H_0 \) = [0.2, 1.2]
Argonne National Laboratory (USA/IL)
Physics: Gravity
Volume: \( 4.225 \text{Gpc}^3 \)
Particle Nr.: \( 10240^3 \)
Dark matter resolution: \( \sim 2.6 10^9 M_{\odot} \)
Semi-Analytic model named cosmoDC2*:
*github.com/LSSTDESC/cosmodc2
Observed:
Modelled:
S. Suyu + 14
PDF from Millenium Simulation
Gregory Dobler et al. 2014
More variables than knows !!!
Options 1:
Collect better data
Options 2:
Collect more data
Fermat Potential difference
Observable
Goal
Inferable
Observable
Goal
Inferable