1

System wide optimization for dark energy science: DESC-LSST collaborations

Tony Tyson

LSST Dark Energy Science Collaboration meetingJune 12-13, 2012

2

Multiple LSST probes of dark energy

• Use the same LSST survey data products

• Analyzed for different signals

• Multiple cross checks

• Combination is far more powerful than root mean square

• Maximally sensitive to new physics

Primary LSST probesWeak Lens shear cross correlation tomography Weak Lens magnification cross correlation tomography 2-D Baryon Acoustic Oscillations Supernovae Shear peak statistics Galaxy cluster counts

Secondary LSST probesTime domain tomography of QSOs and AGNs

Anisotropy of WL+BAO and SN signals

New Energy or New Gravity?

3

DE Probes and DESC Tasks

• Measure geometry with Baryon Acoustic Oscillations– Photo-z effects, Non-linear corrections

• Measure mass and geometry with Weak Lensing Tomography– Measure and control systematics

• Combine multiple LSST probes – Break degeneracies– Minimize sensitivity to systematics

• Even better precision: DESC R&D– Photo-z and shear estimation algorithms, and their validation

• System-wide systematics relevant to DE– Optics, Detectors, Wavefront sensor, Guider, Calibration, …

4

DETF – Science Book – Astro 2010 – CD-1

4

5

DESC

WHY NOW?

Getting started:https://www.lsstcorp.org/sciencewiki/index.php?title=Getting_Started

Register for the All Hands Meeting (Aug 13-17):https://www.lsstcorp.org/ahm2012/  

6

Optimize LSST Survey for Dark Energy

LSST PROJECT

LSST DARK ENERGY SCIENCE

COLLABORATION

Get close to the experiment: System-wide involvementValidate algorithms, develop new algorithms

Ensure Simulator fidelity Test the Simulator and System ComponentsExplore cadence scenarios and systematics

7

Modes of Interaction

LSST PROJECTLSST DARK ENERGY

SCIENCECOLLABORATION

STUDENTS

DESC students work with existing LSST R&D groups Help validate algorithms, develop new algorithms

Help ensure Simulator fidelity Help test the Simulator and System Components

Explore cadence scenarios and systematics

LSST PROJECT

LSST PROJECT

LSST PROJECT

STUDENTS

8

Tasks

WEAK LENS SHEAR• PSF Systematics• PSF Control• Wavefront sensing• Stack-Fit and Multi-Fit• Galaxy shear• End-to-end simulations

PHOTOMETRIC REDSHIFT• Algorithms,• Photometry, Calibration• Required precision• Systematics, Simulations

BARYON ACOUSTIC OSCILLATIONS• Simulations

WIDE AREA ISSUES• Dither, Patch effects

SUPERNOVAE• Third parameter

EXTRACTING DE SCEINCE• Joint WL+BAO • Cross-calibration• All probes

MAPPING ONTO THEORY• Experiment constraints• Statistical inference

DATA MANAGEMENT• Automated DQA: document LSE-63• Computation

LSST Hardware systematics• CCDs, Optics

CALIBRATION• Spec/Phot

PRECURSOR AND TEST DATA• Subaru• f/1.2 LSST Camera Beam tests

9

Example near-term focus areas

WAVEFRONT SENSING

DESC students work with existing LSST R&D groups Help validate algorithms, develop new algorithms

Help ensure Simulator fidelity Help test the Simulator and System Components

Explore cadence scenarios and systematics

PHOTOMETRIC CALIBRATION

IMSIM: LSS WL

STACKFIT DEVELOPMENT AND TESTS

10 10

F. Roddier, Applied Optics, 27, 1223, 1998

More intense Less intense

Wavefront curvature sensing

ScienceFocal Plane

I I

I I

2 r

R

11

Ghost rays (about 3% of direct intensity)Direct rays

Photometric calibration: flat field

12

DM Pipelines

Solar System Cosmology

Defects

MilkywayExtended Sources

Transients

Base Catalog

All Sky Database

Instance CatalogGeneration

Generate the seed catalog as required for simulation. Includes:

Metadata Size Position

Operation Simulation

Type Variability

Source Image Generation

Color Brightness Proper motion

Introduce shear parameter from cosmology metadata

DM Data base load simulation

Generate per FOV

Photon Propagation

Operation Simulation

Atmosphere

Telescope

Camera

Defects Formatting Generate per Sensor

Calibration Simulation

LSST Sample Images and Catalogs

IMAGE SIMULATIONS

DESC

13

Shape measurements on galaxy and star images

14

Measuring faint galaxy shear: Stack-Fit

• Measure the shape of galaxies whose apparent shape is distorted by the point-spread function (PSF)

• PSF varies within CCDs and between CCDs and between exposures due to optics and atmosphere variations

• The Stack-Fit Algorithm: 1. Measure PSF within each CCD for each exposure 2. Separately make weighted co-add of all dithered images of the

field 3. Co-add with same weights the CCD PSF eigenfunctions 4. Use this PSF co-add map to interpolate the PSF at each galaxy’s

position 5. Convolve this PSF with a galaxy model, and fit.

• Test performance on end-to-end LSST image simulations• Test via comparing HST and Subaru WL mass reconstruction

15

Subaru-HST shear component comparison @ 40 source galaxies/arcmin2

e1 Subaru-e1 HST e2 Subaru-e2 HST

Binned by magnitude

Systematic offset test: post Stack-Fit galaxy-by-galaxy distribution of Subaru-HST shear components

Will Dawson, UCD

16

Test Stack-Fit using full LSST image simulations

LSST cosmic shear residual systematic errors

Jee & Tyson 2011

Initial tests on LSST image simulations, including atmosphere and focal plane errors, suggest that residual systematic shear correlations may be reduced below the shot noise in ~100 images of a field.

R&D on this and similar algorithms is planned, using the end-to-end LSST survey image simulations -- including realistic lensing power spectra and observational systematics.

17

Testing CCD systematics: f/1.2 beam

18

DETF FoM(t) during ten year survey

19

Testing general models of dark energy

Science Book Ch. 15.1