Welcome to the workshop
``Strong Gravitational Lensing in the Next Decade’’
Cogne, Aosta, Italy
Strong Gravitational Lensing:
A Past, A Present, A Future?
Léon Koopmans
(Kapteyn Astronomical Institute)
Note:
This talk will be my own, sometimes biased,
view of the current state of strong lensing. I hope by the end of the workshop that we have a more common/
shared view where where strong lensing is now and where it should be heading in the next decade.
Some questions to be addressed during this workshop
•
What has strong lensing accomplished in the last three decades in terms of scientific results?•
Where has strong lensing had the most impact?•
What is the current status of the field compared to other techniques?•
Where will strong lensing have the most impact in the coming decade?•
How many strong lenses do we know and do we cover enough of “parameter space’’?•
From where will we obtain new strong lenses?•
Should the field get more organized (science cases)?Some Accomplishments of Strong Lensing over the last 3 Decades
•
Cosmography: H0, Ωm, ΩΛ•
Galaxy Structure & Evolution•
Mass Substructure•
Central SMBHs•
Galaxy ISM•
Source Structure•
Alternative Gravity Theories?•
More?Some ramblings on different topics...
Lens Surveys
(Jackson, Wucknitz, McKean, Marshall, Courbin, Dobler)
For a field to flourish, it requires not only excellent theory (GR) and detailed modeling, but also a lot of data.
Do we have enough lenses?
There are several (traditional) ways to go about finding new strong gravitational lenses
•
Imaging: Radio/Optical (efficient)•
Spectroscopy: Looking for different redshifts•
Variability? Looking for correlated variations•
Search for lenses or search for sources?(if Nlens > Nsrc, search for the src, and visa versa!)
Lens Surveys
Lens Surveys
To find strong lenses, several requirements have to be met:
(1) The telescope has to have sufficient spatial resolution (<1” typically) to resolve multiple images/arcs/rings
(2) The telescope has to have sufficient survey speed/depth (targeting many objects at once or fast one-by-one;
i.e. blind vs targeted, depending on Nsrc/lens) (3) Preferably the telescope has a wide FOV for blind surveys.
(4) Contrast of lens vs source has to be such that it can be recognized as a lens in the first place.
(5) Luck!
Lens Surveys
Recent/Past Surveys
★ CFHT-LS
★ Muscles
★ Haggles
★ Cosmos
★ SLACS
★ JVAS/CLASS
★ SLS-AEGIS
★ OLS
★ PANELS
★ SDSS
★ 2dF Lens Survey
★ HST Snapshot Lens Survey
★ FKS Lens Survey
★ NOT Lens Survey
★ APM Lens Survey
★ MG Survey
★ Discover through luck: Q0957+561
★ SKA
★ LOFAR
★ E-MERLIN
★ E-VLA
★ Meerkat/ASKAP
★ LSST
★ JDEM/EUCLID
★ PANSTARRS
★ more?
Future Survey Instrument
CLASS SLACS
Lens Surveys
CASTLES
Lens Surveys
Current Status:
(1) Around ~200 lenses are known (e.g. CASTLES+SLACS);
most surveys discover several to several tens of lenses.
(2) Roughly half of the lenses are source selected (QSOs) and half are lens selected.
(3) Most source-selected surveys have heterogeneous lens samples (lens-selected surveys are more uniform).
(4) It has remained hard to split samples in bins and make detailed statements about trends (small # statistics)
(5) Small number statistics has plagued us for a long time
Lens Surveys
Questions:
(1) Do we have enough lenses to keep us going for the next decade and is follow-up sufficient, or not?
(2) Do we need larger samples that cover a wider range in
redshift, galaxy/src properties, environment, etc? How large?
(3) Which telescopes/surveys will yield sample of lenses >>100 in the coming decade, or even >>104?
(4) Are we prepared or are we preparing for these and should we push harder to get strong lensing “on the agenda” and into science cases for new telescopes/surveys?
The Hubble Constant
(Coe, Suyu)
The Hubble Constant
This was traditionally one of the main science goals of strong lensing. Will/should it remain so and why?
The Hubble Constant is inverse proportional to
the scale of the Universe; light travel time is proportional to the same scale. If we can measure light-travel-time
to an object (or differences), we would know H0.
With lenses we measure time difference between lensed
images, which is a function of geometry (where is the src wrt the images and what Shapiro delays do they have?;
typically these terms are roughly equal).
The Hubble Constant
Hence two ingredients are required:
(1) A good mass model (where is the source and what is the Shapiro delay) of the lens, the field and the l.o.s
(2) A good time-delay (few % accuracy)
The Hubble Constant
Monitoring campaigns to get time-delays:
(1) Radio: VLA, MERLIN, [WSRT]
(2) Optical/IR: NOT, COSMOGRAIL, Maidanak (3) X-ray: ???
Follow-up of the lens for modeling
(1) Deep imaging: lens & image structure, color info, etc (2) Spectroscopy: redshifts, kinematics, etc.
Oguri et al. 2007
Current Status:
(1) Around 20 systems with time-delays are now known (2) The Hubble Constant still lies between ~60-80 km/s/
Mpc sometimes even lower, when assuming SIE models (3) Too little effort is often going into mass modeling once a time-delay is measured.
(4) All galaxies with time-delays are QSO/AGN and often the lens-galaxy is hard to study, limiting detailed mass modeling. Visa versa, lenses that are easy to study
often have non-varying sources.
The Hubble Constant
The Hubble Constant
Questions:
(1) Why are lens H0 values still “all over the place”?
Is this due to wrong time-delays, wrong mass models, or wrong physics?
(2) Should we concentrate on more lens system or invest more time in systems where both a good time-delay and a good mass model can be determined?
(3) Is H0 from lensing still interesting and competitive, or is that the wrong question and agreement tests our fundamental understanding of cosmology?
Cosmography
Cosmography
Also this in the 1990s was a hot topic in strong lensing, before SNae/CMB/LSS measurements took over. It seems that very little has happened in this field the last 5 years.
How does it work: If one assumes no (or known)
evolution in the comoving number density of lenses and their properties, then the physical number of lenses
(hence the accumulative lensing cross-section) is a function of the volume/length between us and a source at redshift z.
This volume is a function of cosmology and hence the number of lensed objects at a given z depends
on cosmography (e.g. increases with increasing Λ).
Cosmography
To do lensing statistics properly, once requires a very
well-understood lens survey that is “statistically complete”.
Not many such surveys exists and JVAS/CLASS is probably still the largest survey for these purposes. It has 22 lenses.
Most other surveys are looking to maximize the number of lenses w/o regards to being complete in some sense or
being able to quantify selection effects (e.g. dust, variability, lens/src contrast, survey-brightness limits, flux-ratios, etc).
Chae et al. 2002
Cosmography
Why is it difficult?
The difference between a (Ωm=1, ΩΛ=0) cosmology and a (Ωm=0.3,
ΩΛ=0.7) cosmology is a factor of two smaller volume to z=1. With
22 lenses a Poisson error of ~4.5 is expected, making the difference
between 22 or 11 lenses only significant at the 2.5-σ level for
an assumed flat Universe even if no systematics are
included (evolution/biases, etc).
Questions:
(1) Whereas in the 1990s lensing statistics was still competitive, is this still the case?
(2) Could it become competitive again the in the next decade with 100s of new lenses being discovered?
(3) Should we turn around the question and fix the cosmography and use lensing statistics to study
galaxy evolution (i.e. marginalize of cosmography or over galaxy evolution)?
(4) Will lensing statistics ever become competitive again?
(5) Can cosmography be done differently with lenses?
Cosmography
Galaxy Structure & Evolution
(Auger, Chantry, Halkola, Czoske, Barnabe, vdVen, Grillo, Faure, Nipoti, Meneghetti, Lombardi [clusters])
Galaxy Structure & Evolution
Strong lensing allows for a precise measurement of galaxies masses and set constraints on their density profiles over a wide range in redshift and moderate
range in galaxy masses.
This provides a powerful method to study galaxy structure & evolution over 8 Gyr in look-back time.
Lensing studies can also be combined with other techniques (e.g. dynamics/X-rays, etc) to improve
precision and accuracy.
Galaxy Structure & Evolution
Applications vary wildly:
(a) Total density profiles (b) DM density profiles (c) Shape of haloes
(d) Fundamental Plane Studies (e) Calibrating stellar IMF
(f) etc.
Galaxy Structure & Evolution
Koopmans et al. 2006, 2009
It is not well understood why baryonic & DM add to a combined ~1/r2 density
profile, despite different
physics and starting conditions.
!γLD! " = 2.085+0.025−0.018 (68% CL)
Isothermal, but with intrinsic spread of <10%
Galaxy Structure & Evolution
Kochanek
Joint constraints can be set on stellar mass-
fraction and DM density slope.
Galaxy Structure & Evolution
Bolton et al; Koopmans et al.
A Mass Fundamental Plane has been found, which sheds
light on the tilt in the FP and the structure of ETGs
ISM
(Ellisadottir)
ISM/Scattering/Microlensing
Compact (possibly extended) sources allow on the study the content of the lensing galaxy (e.g. stars, gas, dust, ionized gas, etc). However, this requires disentangling the lensed source structure from what is
happening to it in the lens.
Also scattering/dust/gas, etc can affect the surface brightness of the lensed image and change it
(microlensing does not).
ISM/Scattering/Microlensing
Some of the main questions addressed in these studies:
(1) A considerable amount of work has been done on
(differential) spectroscopy to look at the gas/dust content in lens galaxies -> provides a detailed look at the ISM
of high-z galaxies hard to get in other ways (e.g. emission).
(2) Some effects of scattering have been found in radio lenses (eg CLASS 0128) that indicates a dense ionized medium (e.g. HII regions, SF, etc).
Biggs et al.
Surface brightness is not conserved
Motta et al.
ISM/Scattering/Microlensing
Extinction survey at high z
ISM/Scattering/Microlensing
Questions:
(1) One man’s signal is another man’s noise: Should we worry in lensing studies about microlensing, dust,
scattering etc (e.g. for substructure/time-delay studies).
(2) What have we learned from these studies about the ISM/*IMF of distant galaxies?
(3) Are these studies competitive/complementary/better to/than other studies? Could they provide evolutionary studies of the galaxy ISM/IMF?
Substructure
(Keeton, Chen, Vegetti, Fassnacht)
Substructure
Smooth mass models predict a distinct relation between the magification/fluxes of merging cusp/fold images: i.e.
the cusp relation (sum of magnification is zero).
This relation is violated in many instances.
!N
i=1
µi = 0
Mass substructure is generic prediction of the LCDM cosmography and lensing is a unique method to find it.
Substructure
B2045+265 is a classical example of the cusp violation
Fassnacht et al. McKean et al.
B2016-112: violation is due to a visible dwarf satellite;
could this hold for many/most anomalous lenses?
Substructure
Substructure
Dalal & Kochanek
Current predictions are fDM<1% inside Reinst
Limits have been set but some of the data has been questioned, so the jury is still out.
Questions:
(1) What is current level of substructure and is this interesting for lensing studies?
(2) Is lensing competitive or even unique for substructure studies in the future?
(3) What techniques are there (e.g. flux-ratio/direction
imaging) and what are their complementarities/strength?
(4) How many lenses do we need to pin down the fraction of DM and is mass function?
(5) Could there be too much high-mass substructure?
Substructure
Central SMBHs
Central SMBHs
The absence of central (3rd/5th) images in lens galaxies implies that the central density of the lens must be very high.
µ = 1
(1 − κ)2 − γ2 ≈ 1 κ2core
µ = 1
(1 − κ)2 − γ2 ≈ 1
γSMBH2
Due to density core
Due to central SMBH
If a SMBH exists, the image formed near the center of the galaxy will have properties determined by the
SMBH mass (plus external shear).
Central SMBHs
Current status/Requirements:
(1) Only one detected central image (Winn et al.) (2) Deep high-resolution and high contrast images (3) Most likely only reachable with VLBI radio
experiments.
(4) Currently still factor ~10 away from required sensitivity.
(5) Asymmetric lens geometries are requires (high flux-ratio), which are even rarer.
Central SMBHs
Extra images can form (two on one side!) or central image can be
destroyed.
These images or absence can be signs of SMBHs and be used
to quantify their mass in a geometric manner.
Mao et al.; Rusin et al.
Central SMBHs
Questions:
(1) Is measuring the mass of SMBHs interesting and competitive if ever doable?
(2) Can it provide a model-independent mass measurement?
(3) Can it probe SMBH-mass evolution?
(4) How many lenses are needed to find a single case?
(5) Could one probe gravity in the strong regime?
Lensed Sources & Their Structure
(Volino, Bradac)
Source Structure
The high macro/microlensing magnification allows one to study the source structure in great detail.
Caution: Magnification bias works for unresolved sources, but not for resolved source, where the real limit is
not the flux-limit but the surface brightness limit.
Hence lensing provides really only a way to super-resolve a source, but not to gain much in integration time (only because of multiple imaging)!
Source Structure
Magnification of 200-300 !
Study the counter jet of an AGN at z>3.
More et al.
Source Structure
Optical image is not anomalous X-ray image is anomalous
Explained by microlensing; requires QSO optical source size >> x-ray size
Source Structure
Questions:
(1) Can the increasing number of light-curves with
microlensing provide interesting limits on the stellar content (M/L) in lenses?
(2) Are microlensing results being recognized in the AGN community are being interesting?
(3) Will lensing being leading or complementary in studying source structure because of the small numbers of lenses and the serendipity of the source structure?
Some new surveys & instruments coming online that could provide more lenses?
(1) optical: LSST/Panstarss/...
(2) radio: e-MERLIN/e-VLA/LOFAR/Meerkat/Askap/SKA (3) space: JDEM/Euclid/...
Possible fields of large impact?
(1) Galaxy Structure & Evolution (2) Mass substructure
(3) Central SMBHs?
(4) ... Pick your favorite ...
A Future? Yes!
But we have to keep working
