Strong Gravitational Lensing of Explosive Transients
Strong Gravitational Lensing of Explosive Transients
Strong Gravitational Lensing of Explosive Transients
Masamune Oguri1,2,3
1 Research Center for the Early Universe, University of Tokyo, Tokyo 113-0033,
arXiv:1907.06830v1 [astro-ph.CO] 16 Jul 2019
Japan
2 Department of Physics, University of Tokyo, Tokyo 113-0033, Japan
3 Kavli Institute for the Physics and Mathematics of the Universe (Kavli IPMU,
time delays constrain H0 at better than 3% precision more fundamental than the geometric optics (e.g.,
[49–52]. [58]). The wave optics effect produces interesting
Recently measurements of H0 attract a lot of observable features such as the interference pattern,
attention given a possible tension among them. One which may provide additional useful information on the
of the most traditional methods to measure H0 is lensing object. In order for this effect to be observed,
the so-called distance ladder (e.g., [53–55]), with the the source must be sufficiently compact, as the finite
most recent measurement yielding the best-fit value of source size smears the interference pattern. Since these
H0 = 74.03 ± 1.42 km/s/Mpc including systematics explosive transients have compact sizes as compared
[56]. On the other hand, H0 can also be inferred with quasars and galaxies, strong lensing of explosive
from observations of cosmic microwave background transients may open up the possibility of using the
anisotropies, yielding H0 = 67.4 ± 0.5 km/s/Mpc wave optics effect as additional applications.
assuming the standard Λ-dominated cold dark matter In this review article, we focus on strong lensing of
model [57]. The discrepancy between these two explosive transients, which will be discovered in large
measurements might suggest new physics such as numbers in the future. We discuss how these events
additional relativistic particle species, or might be can be used to address several outstanding questions in
attributed to unknown systematic errors in either or modern cosmology, such as the nature of dark matter
both of these two measurements. While the latest and dark energy. In addition, strong lensing can be
measurement of H0 from 6 quasar lens time delays is used to understand these explosive transients better,
H0 = 73.3+1.7
−1.8 km/s/Mpc [52] and is consistent with with help of the gravitational lensing magnification.
the distance ladder result, more accurate and precise We also discuss the prospect for detecting these events
measurements of H0 from time delays as well as the in the future. We note that this review article focuses
exploration of a possible dependence of the constraints on a limited aspect of strong lensing, and in fact there
on redshifts are important to understand the origin of are many reviews and textbooks [58–69] that are useful
the H0 tension. to cover the broader aspects of strong lensing.
In fact, the Refsdal’s original proposal assumed to The rest of this review article is organized as
use strong lensing of supernovae, rather than quasars, follows. In Section 2, we briefly review basic theory of
to measure H0 from time delays. Because of the strong lensing. In Section 3, we summarize explosive
relatively small number of distant supernovae observed transients that we discuss in this review article. In
so far, strong lensing of supernovae has not been Section 4, we discuss possible applications of strong
discovered until recently, which is the reason why lensing of these explosive transients. In Section 5, we
strongly lensed supernovae have not been used to summarize observations so far and also present future
constrain H0 . However, strong lensing of supernovae prospects. We give a brief summary in Section 6.
has several advantages over strong lensing of quasars, Unless otherwise stated, we assume a flat cosmological
as will be discussed below. These advantages make model with matter density ΩM = 0.3156, cosmological
strongly lensed supernovae an alternative powerful constant ΩΛ = 0.6844, and the dimensionless Hubble
probe of the Universe. constant h = 0.6727 [70].
In addition to supernovae, there are other types
of explosive transients known, including gamma-ray 2. The basics of strong gravitational lensing
bursts, fast radio bursts, and gravitational waves from
compact binary mergers. These transients are observed 2.1. Multiple images and the Einstein radius
at cosmological distances, and therefore are subject to
strong lensing applications. A notable difference of We begin with a brief overview of the formulation of
these transients from supernovae is that their typical strong lensing. While the path of light rays in arbitrary
time scales of light curves, seconds or milliseconds, are matter distributions in the Universe is calculated by
much shorter than the time scale of supernova light the geodesic equation in General Relativity, in most
curves, a month to several months. The shorter time astronomical situations where deflection angles are
scales indicate that the measurement precision of time small we can linearize the geodesic equation to obtain
delays is much better and that they can in principle the so-called lens equation. The lens equation can be
probe much smaller mass scale of the lensing object. regarded as mapping between positions of the source
In most applications of strong lensing, we (that would be observed in absence of the gravitational
can assume geometric optics, which is a good lensing effect) and the image (that is actually observed)
approximation when the wavelength is sufficiently on the sky. In the analysis of strong lensing, it is also
small compared with the scale of the structure of common to assume that the deflection is dominated by
interest. However, there are cases where we have a single object along the line-of-sight whose size is thin
to take account of the wave optics effect, which is as compared with cosmological distances. Under these
approximations, the gravitational lensing effect is fully
Strong lensing of explosive transients 4
image
from which the magnification µ of each image is
computed as
1
µ(θ) = . (6)
θ detA(θ)
source
β
This means that the image at θ is magnified by a factor
lens observer of |µ(θ)|. The sign of µ corresponds to the parity of
Dls Dol the image such that the parity of the image is flipped
Dos when µ is negative.
Equation (6) indicates that magnification factors
Figure 1. Schematic illustration of a gravitational lens system. formally diverge at points satisfying detA(θ) = 0.
These points in the image plane form closed curves,
which are called critical curves. Corresponding curves
described by the following lens equation in the source plane obtained via the lens equation (1)
β = θ − α(θ), (1) are called caustics. These curves are important in
strong lensing studies because they are closely related
where two-dimensional vectors β and θ denote to the image multiplicity. When a source is located
positions of the source and the image on the sky, far from caustics, there is only one image. Once a
respectively, and α is the deflection angle source crosses a caustic, the number of images increases
1
Z
θ − θ0 or decreases by 2. Therefore, we can easily infer the
α(θ) = dθ 0 0
κ(θ ), (2) number of images and image configuration by checking
π θ − θ 0 2
the position of a source with respect to caustics. We
where κ, which is sometimes referred to as convergence, show an example in Figure 2, in which we can see 5
is essentially the projected surface mass density images as the source crosses caustics twice (see also
distribution of the lensing object Σ(θ) normalized by [71]).
the critical surface density Σcr When the mass distribution of the lensing object is
Z ∞ spherically symmetric (i.e., κ(θ) = κ(θ)), the deflection
Σ(θ) 1
κ(θ) = = dz ρ(θ, z), (3) angle (2) reduces to
Σcr Σcr −∞
Z θ
θ 2θ
c2 Dos α(θ) = α(θ) = 2 dθ0 θ0 κ(θ0 ), (7)
Σcr = , (4) θ θ 0
4πG Dol Dls
where ρ is the three-dimensional density profile of the and therefore the lens equation (1) reduces to the one-
lensing object, z denotes the line-of-sight direction, dimensional equation
c is the speed of light, G is the gravitational
constant, and Dos , Dol , and Dls are angular diameter β = θ − α(θ). (8)
distances from the observer to the source, from the
observer to the lens, and from the lens to the source, This indicates that in the limit of β → 0 the lensed
respectively. Figure 1 shows a schematic illustration of image forms a ring with the radius θEin that satisfies
a gravitational lens system, including definitions of the θEin − α(θEin ) = 0. (9)
angles involved in the lens equation.
The lens equation (1) predicts the image position This radius θEin is called the Einstein radius. We can
θ for the source position β. Importantly, the lens rewrite equation (9) to obtain
equation is in general nonlinear in θ, which suggests
θEin
that multiple θ can satisfy the lens equation for a
Z
given β. These multiple solutions of the lens equation M (< θEin ) = 2
Dol dθ0 2πθ0 Σ(θ0 ) = πDol
2 2
θEin Σcr ,
0
correspond to multiple images. Such multiple images (10)
can be produced where the deflection angle α is which indicates that the Einstein radius probes the
sufficiently large, i.e., in high density regions such as total projected mass of the lensing object within the
centers of galaxies and clusters of galaxies. Einstein radius, as long as lens and source redshifts
Gravitational lensing not only changes the ob- are known.
served position on the sky, but also changes the ob- In fact, equation (10) has important implications.
served brightness of the source. The change of the It has been known that image separations between
brightness is determined by the Jacobi matrix from multiple images are typically twice the Einstein radius,
the lens equation which is approximately true even when the lens
∂β mass distribution is not spherically symmetric. This
A(θ) = , (5) means that we can measure the Einstein radius fairly
∂θ
Strong lensing of explosive transients 5
2 found to
caustics
r
1 M
source θEin =
Dol πΣcr
1 1/2 −1/2
00 −6 M Dol Dos /Dls
∼ 1.63 × 10 (11)
,
M 3.06 Gpc
β2 [arcsec]
0
where distances are normalized to values at the lens
redshift zl = 0.5 and the source redshift zs = 1.0.
Another lens model that is commonly used is a
singular isothermal sphere (SIS) model whose three-
−1 dimensional radial density profile is given by ρ(r) =
σ 2 /2πGr2 , where σ is the velocity dispersion. The
Einstein radius of the SIS model is computed as
−2 4πσ 2 Dls
−2 −1 0 1 2
β1 [arcsec] θEin =
c2 Dos
σ 2 D /D
2
00 ls os
∼ 0.492 , (12)
critical curves 200 km s−1 0.426
images where the distances are again normalized to values
1 at the lens redshift zl = 0.5 and the source redshift
zs = 1.0.
θ2 [arcsec]
100
4GM (< θEin )
∆tfid = (1 + zl ) , (19)
c3
−1
10
which suggests that ∆tfid is on the order of the light
crossing time of the “gravitational radius” for the mass
10−2 defined by M (< θEin ). Again, by normalizing distances
to those at the lens redshift zl = 0.5 and the source
10−3 redshift zs = 1.0, ∆tfid is estimated as
1012 1013 1014 1015
Mvir [M]
θEin
2
(1 + zl )Dol Dos /Dls
∆tfid ∼ 128 day × .
100 4.59 Gpc
Figure 3. The Einstein radius θEin as a function of the halo (20)
mass Mvir for an NFW profile. The lens redshift is fixed to
Therefore, for typical galaxy-scale strong lens systems
zl = 0.5, whereas source redshifts of zs = 1 (solid red) and
6 (dashed blue) are considered. Here the mass-concentration with θEin ∼ 0.500 − 100 , we expect time delays on the
relation presented in [76] is adopted. The shaded regions order of a month to a few months. For reference, by
represent the uncertainty of θEin originating from 1σ scatter inserting the typical Einstein radius of a solar mass
(σln c = 0.3) of the concentration parameter.
compact object (see equation 11), θEin /100 = 10−6 , to
equation (20), we obtain ∆tfid ∼ 1.1 × 10−5 sec for the
fiducial distances. More generally, from equation (19)
gravitational potential of the lens. Note that we can we can estimate ∆tfid for a point mass lens with mass
observe only the time delay between image i and j, i.e., M as
∆tij = ∆ti − ∆tj .
Again, it is useful to present time delays for some −5 M
∆tfid ∼ 1.97 × 10 sec × (1 + zl ) , (21)
lens models. Using equation (13), we can rewrite ∆tij M
as
which can also be applicable to other lens models if we
∆tij = ∆tfid Φ(θ i , θ j ), (15)
replace M to M (< θEin ).
On the other hand, the function Φ(θ i , θ j ) depends
1 + zl Dol Dos 2 on the assumed mass model. For a point mass lens,
∆tfid = θ , (16)
c Dls Ein
θj2 − θi2
θj
Φ(θ i , θ j ) = ln + , (22)
(θ i − β)2 (θ j − β)2 φ(θ i ) φ(θ j ) θi 2θi θj
Φ(θ i , θ j ) = 2 − 2 − 2 + 2 .
2θEin 2θEin θEin θEin and for an SIS lens
(17)
The factor ∆tfid represents a typical size of the time 2(θj2 − θi2 )
Φ(θ i , θ j ) = , (23)
delay for the lens, and Φ(θ i , θ j ) is a O(1) function (θj + θi )2
that represents the dependence of the time delay on the
image configuration. For instance, the multiple image where θi = |θ i | and θj = |θ j |. Again, no simple
configuration is symmetric, we have |θ i | ∼ |θ j | and analytic expression of Φ for the NFW profile is known.
|β| ∼ 0, leading to Φ(θ i , θ j ) ∼ 0. Put another way, As mentioned in Section 1, time delays provide
time delays are larger when the image configuration a powerful means of measuring the Hubble constant
is more asymmetric, and are smaller when the image H0 , which is sometimes referred to as time delay
configuration is more symmetric. cosmography. Given that θEin is well constrained from
We can also reinterpret the gravitational lensing the data, observations of time delays between multiple
effect from this expression of the time delay [77, 78]. images put direct constraints on the distance ratio
Fermat’s principle in geometric optics states that a Dol Dos /Dls , which is inversely proportional to H0 ,
light ray takes a path with a stationary path length. but only if Φ is accurately known. The examples
This immediately suggests that observed images should above already indicate that values of Φ depend on
satisfy the following condition the underlying lens mass model, which implies that
accurate determinations of lens mass distributions are
∇θ ∆t = 0. (18) a key for the successful time delay cosmography.
Strong lensing of explosive transients 7
When multiple images are observed, we can observed image positions unchanged. In order to mea-
constrain the lens mass distribution from positions and sure H0 robustly from time delays, it is essential to
flux ratios of multiple images. However, in most cases explore these degeneracies carefully, and to make use
these constraints are insufficient to robustly constrain of additional observational constraints that can break
the lens mass distribution, and we need additional these degeneracies.
constraints. For instance, host galaxies of quasars
or any explosive transients are also expected to be 2.3. Lensing rates
lensed into extended arcs, which may provide useful
additional constraints (e.g., [79]). Furthermore, the Strong lensing is a rare event that occurs only when
velocity dispersion of the lensing galaxy, which can be the light ray from a distant object passes through
observed by deep spectroscopy of the lensing galaxy, is high density regions such as centers of galaxies and
sometimes used as additional constraints on the lens clusters. The chance probability of strong lensing
mass distribution (e.g., [80]). can be calculated as long as the density profile and
However, there is a fundamental difficulty in the abundance of putative lensing objects are known. From
strong lensing analysis, which originates from various the density profile one can derive the lensing cross
degeneracies inherent to the lens equation. One such section, i.e., the area on the sky within which strongly
example is the mass-sheet degeneracy [81], in which the lensed multiple images are produced, which is on the
2
following transform is considered order of θEin .
Historically, strong lensing probabilities are calcu-
θ2 lated assuming lensing by galaxies. For instance, the
φ(θ) → (1 − κext )φ(θ) + κext , (24)
2 detailed calculation in Turner et al. [85] indicates that
strong lensing events are dominated by those due to
β → (1 − κext )β, (25) field elliptical galaxies. They also show that strong
lensing probabilities are a steeply increasing function
where κext is constant. It is straightforward to see that
of the source redshift. Calculations of strong lensing
this transform keeps the lens equation (1) unchanged.
probabilities have been improved partly due to im-
This transform corresponds to an operation that
proved measurements of velocity dispersion functions
rescales the mass of the lensing object and instead
of galaxies in observations [13, 14, 86–94].
inserts a constant mass sheet κext . Importantly, this
It has been known that clusters of galaxies also
transform also changes time delays (13) between any
produce strong lensing. While individual clusters have
multiple image pairs as
larger lensing cross sections than galaxies, clusters are
∆tij → (1 − κext )∆tij , (26) much less abundant than galaxies. Narayan and White
[95] discussed the image separation distribution in the
which indicates that H0 estimated from observed time standard cold dark matter (CDM) cosmology to argue
delay should scales as H0 → (1 − κext )H0 . Therefore, that the contribution of clusters to the total strong
H0 measured from time delays is subject to the lensing probability is small but non-negligible. This
uncertainty of κext that cannot be constrained from calculation has been updated following the improved
strong lensing observations. As we will discuss later, knowledge of the density profile and the abundance of
one way to break the degeneracy is to observe the clusters [96–110].
magnification factor µ, because the transform changes As briefly mentioned in Section 2.1, N -body
µ as simulations of the structure formation in the CDM
µ → (1 − κext )−2 µ. (27) model have revealed that the density profile of dark
Note that this transform does not change the ratio of matter halos is universal and is well approximated by
magnification factors between multiple images. the NFW profile [73, 74]. As shown in Figure 3, the
This mass-sheet degeneracy implies other approx- Einstein radius of the NFW profile is a steep function
imate degeneracies. For instance, for a power-law mass of the halo mass such that it becomes too small for
model with φ ∝ rβ (β = 1 corresponds to an SIS pro- galaxy-scale dark matter halos, Mvir . 1013 M , which
file), the change of β around β = 1 can be approxi- appears to contradict observations in which there are
mated by the mass-sheet transform with 1−κext = 2−β many strong lens systems with θEin ∼ 100 due to
(e.g., [82]), which implies that the Hubble constant isolated galaxies.
from time delays is sensitive to the radial slope of the This issue is resolved by taking proper account
density profile of the lensing object, which is difficult to of the baryonic component. Dissipative cooling of
be constrained from strong lensing observations. Fur- gas makes the spatial distribution of stars much more
thermore, the mass-sheet transform is generalized to compact that that of dark matter. At the galaxy scale
the source-position transform [83, 84], which is essen- this effect is more efficient such that the total density
tially a global mapping of the source plane that keeps profile of dark matter and the baryonic component
Strong lensing of explosive transients 8
resembles an SIS profile that was also mentioned in Here we present some examples of calculations of
Section 2.1. Indeed calculations based on this idea strong lensing probabilities, following recent calcula-
successfully reproduce the observed image separation tions presented in Oguri [123]. In short, we compute
distribution of strong lenses for a wide mass range from strong lensing probabilities due to single galaxies, be-
the galaxy to cluster scales [23,111–121], which suggest cause galaxies dominate the total strong lensing prob-
that the contribution of clusters to the total strong ability as discussed above. The strong lensing proba-
lensing probability is ∼ 1−10% and that strong lensing bility Psl (zs ) for a source at redshift zs is computed as
events are dominated by those due to single galaxies.
In practice, we need to take account of selection zs ∞
d2 V
Z Z
dn
effects when we compare expected strong lensing Psl (zs ) = dzl dσ Bσsl (σ), (30)
probabilities with observations. The best-known 0 dzl dΩ 0 dσ
example is the magnification bias [122], which where d2 V /dzl dΩ is the comoving volume element per
originates from the fact that in any survey objects are redshift and steradian, dn/dσ is the velocity dispersion
detected only above some flux threshold. Because of function of galaxies, σsl (σ) is the strong lensing cross
gravitational lensing magnifications, faint objects that section in units of steradian for galaxies at redshift
fall below the threshold in absence of gravitational zl with the velocity dispersion σ, and B encapsulates
lensing can in fact be observed thanks to the various selection effects such as the magnification bias.
magnification. This effect increases observed strong To compute the strong lensing cross section, we assume
lensing probabilities and hence should be taken into that the mass distribution of lensing galaxies follow a
account. For sources with the flux f and differential Singular Isothermal Ellipsoid, which is an extension of
number counts N (f ) = dN/df , the magnification bias an SIS to include the ellipticity in the projected mass
factor B is computed as distribution. We also add external shear perturbation.
1
Z ∞
dµ dP The velocity dispersion function dn/dσ is taken from
B= N (f /µ), (28) the one measured in the Sloan Digital Sky Survey [124]
N (f ) µmin µ dµ
with the redshift evolution predicted by the Illustris
where dP/dµ denotes the magnification probability cosmological hydrodynamical simulation [125]. The
distribution. As a simple example, assuming power- strong lensing probability is derived in the Monte-
law number counts N (f ) ∝ f −α and an SIS lens for Carlo approach in which many lenses and sources are
which dP/dµ = 8/µ3 (µ > µmin = 2), we can compute randomly generated and the lens equation is solved
B as numerically using glafic [72]. Interested readers are
2α referred to [123] for more details.
B= . (29)
3−α Figure 4 shows the strong lensing probability
From this expression it is found that steeper number Psl (zs ) computed with the setup described above,
counts (larger α) lead to the larger magnification bias without any selection bias i.e., B = 1. As noted in
factor. e.g., [85], the strong lensing probability is a steeply
In addition to the magnification bias, there are increasing function of the source redshift at low
other possible selection effects. Multiple images with redshifts, z . 1. The dependence on the redshift
large differences in their fluxes are difficult to be becomes somewhat weaker at higher redshifts. The
identified in observations, and any cut on the flux redshift dependence
Rz mainly comes from the total
ratio of multiple images reduces the strong lensing volume, 0 s dzl d2 V /dzl dΩ in equation (30), which
probability. For lens systems with small Einstein radii, suggests that the redshift dependence is ∝ zs3 at low
image separations of multiple images can be too small redshifts. We find that the strong lensing probability
to be resolved in observations, depending on spatial shown in Figure 4 is crudely approximated by the
resolutions of observations. Also when the size of following functional form
the source is comparable or larger than the Einstein
(5 × 10−4 )zs3
radius, the gravitational lensing effect on the source Psl (zs ; B = 1) ≈ , (31)
(1 + 0.41zs1.1 )2.7
is quite inefficient. These effects remove strong lenses
with small image separations from the sample, leading which may be useful for quick estimates of the occur-
to the smaller strong lensing probability. Depending rence of strong lensing events in various situations.
on the threshold on the image separation, they can As discussed above, in most cases we have to
significantly change the relative contribution of cluster take account of the magnification bias, which can
lenses to the whole strong lens sample. In order significantly enhance the strong lensing probability.
to make fair comparisons with observations, any For instance, the radio source sample used in Cosmic
theoretical calculations of strong lensing probabilities Lens All-Sky Survey [8,9] approximately has power low
should take proper account of these selection effects. number counts with α = 2.1, which leads to, from
Strong lensing of explosive transients 9
equation (29), the magnification bias factor of B ∼ 4.8 In this review article, thus far we implicitly assumed
for an SIS lens. geometric optics in all the calculations of gravitational
The magnification bias factor is larger when lensing. Indeed, geometric optics serves as an excellent
number counts are steeper. An extreme example is approximation in most astronomical situations of
found in the bright ends of number counts or luminosity interest. However, there are some exceptional cases
functions. For many sources, there are exponential where the wave effect plays a crucial role (e.g.,
cutoffs in their number counts or luminosity functions, [127–133]), especially for strong lensing of explosive
and beyond those exponential cutoffs the magnification transients that is a topic of this review article . Here we
bias is infinitely large as without gravitational lensing briefly review the wave aspect of gravitational lensing
magnifications we would not expect any sources theory. For more details, see e.g., [58, 134].
observed far beyond the cutoffs. Therefore at these We consider the propagation of monochromatic
luminosity or flux ranges, almost all the observed waves ψ(x, t) = ψ̃(x)e−2πif t with frequency f . In
sources are strong lensing events, suggesting that the the presence of a weak gravitational field that is
strong lens search among such brightest sources is characterized by the gravitational potential U (x)
highly efficient. A good example of this is brightest (|U | 1), the propagation equation is
galaxies in the submm wavelength, which indeed have (∇2 + ω 2 )ψ̃ = 4ω 2 U ψ̃, (32)
been found to be dominated by strong lensing (e.g.,
[24]). with ω = 2πf . Defining the amplification factor as
In Figure 4, we also show probabilities of
strong lensing events with magnifications µ > 10. ψ̃ L
F = , (33)
Since magnification probabilities are approximately ψ̃
dP/dµ ∝ µ−3 for most situations, we naively expect
Psl (zs ) ∝ µ−2 where ψ̃ L and ψ̃ denote wave amplitudes with and
min . In our fiducial case without the
selection effect we have µmin ≈ 2, suggesting that without gravitational lensing, respectively, we obtain
the larger magnification threshold of µmin = 10 leads the diffraction integral formula for the amplification
to ≈ 1/25 smaller strong lensing probabilities (i.e., factor of gravitationally lensed waves in the expanding
B ≈ 1/25), which appears to hold approximately in Universe
Z
Figure 4. 1 + zl Dol Dos f
F (f, β) = d2 θ exp [2πif ∆t(θ, β)] ,
However, there are some subtleties in computing c Dls i
the magnification bias. The magnification factor (34)
used in the calculation of the magnification bias can where ∆t is the arrival time defined by equation (13).
Note that the wave intensity is amplified by |F |2 . We
Strong lensing of explosive transients 10
can simplify this equation by defining the dimensionless where Φ(θ j , θ k ) is defined in equation (17) and ∆njk =
parameter w using ∆tfid defined in equation (16) nj − nk . The first term in the right hand side
of equation (42) agrees with magnifications in the
1 + zl Dol Dos 2
w = 2πf ∆tfid = 2πf θ . (35) geometric optics, whereas the second term represents
c Dls Ein the wave effect and is the interference between multiple
From equation (19), it is found that w is also expressed images. In the limit f → ∞, however, this term rapidly
as oscillates such that e.g., averaging over a small finite
4GM (< θEin ) source size easily eliminates this term.
w = 2πf (1 + zl ) . (36)
c3 We caution that equation (42) is valid only
approximately, and in order to take full account of the
By defining θ̂ = θ/θEin and β̂ = β/θEin , we can rewrite
wave optics effect we should evaluate equation (37)
equation (34) as
directly. For instance, in the case of a point mass
Z
w h i lens, the amplification can be computed analytically
F (f, β) = d2 θ̂ exp iwT (θ̂, β̂) , (37)
2πi [135, 136]
2
where T is similar to Φ defined in equation (17) and is
2 πw 1 F1 i w, 1; i wβ̂ 2 , (43)
described as |F (f, β)| = −πw
1−e 2 2
(θ̂ − β̂)2 φ(θ̂) where 1 F1 is the confluent hypergeometric function. In
T (θ̂, β̂) = − 2 , (38)
2 θEin this case, the dimensionless parameter w reduces to
which is also dimensionless. 4GM (1 + zl )
w = 2πf
The geometric optics limit corresponds to f → ∞. c3
In this limit, we can evaluate equation (37) using the M f
≈ 1.24 × 10−4 (1 + zl ) . (44)
stationary phase approximation, where only critical M Hz
points satisfying Equation (43) indicates that the maximum amplifica-
∇θ̂ T (θ̂, β̂) = 0, (39) tion at β̂ = 0 is
2 πw
contribute to the integral in equation (37). This is |F (f, β = 0)| = , (45)
1 − e−πw
same as equation (18) and hence the lens equation,
2
indicating that the contributions comes from only which becomes |F (f, β = 0)| → 1 for w → 0. This
multiple image positions θ̂ j . In this limit we can is essentially the diffraction of waves i.e., any obstacle
approximate equation (37) as whose size is much smaller than the wavelength does
X1 not affect the propagation of waves. From this
−1/2 iπσ
F (f, β) ≈ |detA(θ j )| exp iwT (θ̂ j , β̂) + , expression it is found that the gravitational lensing
j
i 4 magnification becomes quite inefficient for w . 1 due
(40) to the wave effect. We note that the similar analytic
where j runs over multiple images, A(θ) is the Jacobi expression of the amplification factor for an SIS lens is
matrix defined in equation (5), and σ is the signature also available [137]
of A i.e., the number of positive eigenvalues minus the 2
∞
number of negative eigenvalues. Given the definition 2
X Γ(1 + n/2)
|F (f, β)| = g(w, β̂) , (46)
of the magnification factor (6), we can simplify this n!
n=0
further as
X h i
1/2
F (f, β) ≈ |µ(θ j )| exp iwT (θ̂ j , β̂) − iπnj , n/2
n i
j g(w, β̂) = 2we(3π/2)i 1 F1 − , 1; w β̂
2
, (47)
2 2
(41)
where µ(θ) is the (signed) magnification factor and where the dimensionless parameter w for an SIS
nj = 0, 1/2, and 1 correspond to the cases where θ j reduces to
is a minimum, saddle, and maximum point of T (θ), 2
1 4πσ 2
respectively. From this expression, we can derive the Dol Dls
w = 2πf (1 + zl ) . (48)
amplification of the wave intensity as c c2 Dos
2
X
|F (f, β)| ≈ |µ(θ j )| Equations (42) and (43) suggest that the gravi-
j tational lensing amplification shows an oscillating be-
X 1/2
+2 |µ(θ j )µ(θ k )| cos [wΦ(θ j , θ k ) − π∆njk ] ,(42) havior as a function of the source position or the fre-
j<k quency of waves. Figure 5 shows some examples. If
Strong lensing of explosive transients 11
β2/θEin = 0.15 which can be observed only when the source size Rs
satisfies the condition given by equation (49). We will
discuss specific examples in Section 4.4.
4
3. Explosive transients
2 3.1. Supernovae
A supernova is an explosion associated with the death
of a star. Observations of supernovae have a long
0
−3 −2 −1 0 1 2 3 history, for example some supernovae took place in the
β1/θEin Milky Way were observed even in the naked eye and
were recorded in the literature. Here we provide a brief
Figure 5. The magnification factor for a point mass lens in wave overview of supernovae. Interested readers are referred
optics, which is computed from equation (43), as a function of to reviews (e.g., [138]) and textbooks (e.g., [139]) for
the source position. Here we fix β2 /θEin = 0.15 and change β1 more details.
to see how the magnification factor change as a a function of
the source position, for w = 0.1 (dash-dotted magenta), 1 (thick
Observationally there is a great deal of diversity
solid red), and 10 (dashed blue), where w is the dimensionless in properties of supernovae, including their light curves
parameter defined by equation (44). The magnification factor and spectral features. First, supernovae are classified
for the geometric optics case is shown by the thin solid gray line. based on the presence or absence of hydrogen lines.
Supernovae without hydrogen lines are classified as
Type I, whereas those with hydrogen lines are classified
observed, it provides a direct evidence of the wave as Type II. Type I supernovae are further divided
effect in action. As mentioned above, however, the into subclasses based on the presence or absence of a
wave effect may be suppressed due to the finite source singly ionized silicon line (SiII) such that those with the
size (e.g., [137]). Here we discuss the finite source strong silicon line are Type Ia and those with the weak
size effect using equation (42), from which it is found or no silicon line are Type Ib/c. Type II supernovae
that the oscillating behavior comes from wΦ. For rea- are also classified into e.g., Type IIP, IIL, and IIn,
sonably small β, equations (22) and (23) imply that depending on their shapes of the light curves and/or
Φ ∼ (θ̂j − θ̂i ) ∼ β̂ = β/θEin . Therefore the width of the presence of absence of narrow line features in their
interference oscillations in the source plane is on the spectra.
order of θEin /w. In order for the interference pattern We can classify supernovae on more physical basis,
to be observed, the source size in the angular unit, depending on their explosion mechanisms. Type Ia
βs = Rs /Dos , should satisfy βs . θEin /w.‡ This condi- supernovae are thought to be thermonuclear explosions
tion yields of white dwarfs near the Chandrasekhar mass, ≈
Dos θEin 1.4M . The explosion of a white dwarf is triggered by
Rs . . (49)
w the matter accretion from a companion star. There is a
In the case of the point mass lens, this condition is long controversy whether the companion star is a non-
expressed as degenerate star such as a red giant or a main sequence
−1 star (single degenerate scenario) or the companion star
1 + zl
Rs . 2.24 × 1015 km is also a white dwarf i.e., a Type Ia supernova is
1.5 trigger by the merger of two white dwarfs (double
−1/2 −1 1/2
M f Dos Dls /Dol degenerate scenario). See e.g., a review by Maoz et
× , (50) al. [140] for more details on this topic. On the other
M Hz 0.949 Gpc
hand, both Type Ib/c and Type II supernovae are
where distances are normalized to values at the lens
thought to be produced by the core collapse of massive
redshift zl = 0.5 and the source redshift zs = 1.0.
stars. There are several possible mechanisms to trigger
To summarize, the wave effect suppresses the
the explosion, including the development of an iron
gravitational lensing magnification when w . 1 due
core that exceeds the Chandrasekhar and leads to the
‡ Near the fold caustic, the time delay between merging image collapse and bounce of the core. After the bounce the
pairs scales as ∆t ∝ β 3/2 , where β here is the distance from outgoing shock is heated by neutrino emitted from the
the caustic. Therefore, in this situation this condition should be
modified as (βs /θEin )3/2 . 1/w. core, which is thought to be a key ingredient for the
Strong lensing of explosive transients 12
−18 IIn
with redshift information, one can constrain the
Hubble constant H0 as well as the cosmic expansion
history out to sufficiently high redshifts. For instance,
−16
luminosity distance measurements out to z ∼ 1 with
Type Ia supernovae led to the direct confirmation of
the accelerated expansion of the Universe and hence
−14 the significant amount of dark energy in the Universe
[144, 145]. Type Ia supernovae also play a crucial role
in the measurement of H0 with the so-called distance
−30 0 30 60 90 120 150 ladder method (e.g., [53]).
days after peak brightness Because of their importance, a number of super-
nova surveys have been conduced, including Super-
Figure 6. Light curves of various types of supernovae. We show nova Legacy Survey [146], Sloan Digital Sky Survey
template light curves of Ia (thick solid red), Ib/c (dashed blue),
IIP (dot-dashed magenta), IIL (dotted cyan), and IIn (thin solid II Supernova Survey [147], Palomar Transient Fac-
green). The light curve templates in V -band are taken from the tory [148], Hubble Space Telescope Cluster Super-
webpage https://c3.lbl.gov/nugent/nugent templates.html. nova Survey [149], Pan-STARRS Medium Deep Sur-
Absolute magnitudes at the peaks correspond to typical vey [150], All-Sky Automated Survey for Supernovae
magnitudes for these supernova types [142].
[151], Dark Energy Survey Supernova Program [152],
and Hyper Suprime-Cam Transient Survey [153]. The
total number of supernovae discovered by now amounts
successful explosion. Extensive numerical simulations to O(104 ) both for Type Ia supernovae and for core-
to understand the explosion mechanism of core-collapse collapse supernovae (e.g., [154]).
supernovae are ongoing (see e.g., [141]). These supernova surveys also revealed new classes
Figure 6 shows template light curves of various of supernovae. Among others, an interesting class
supernovae. It is found that supernovae are of supernovae that may be relevant for this review
luminous. Peak luminosities of luminous supernovae article is a superluminous supernova (e.g., [155, 156]).
are comparable to galaxy luminosities, which indicate One of the first examples of this class, SN 2005ap,
that we can observe supernovae out to high redshifts, was discovered by the Texas Supernova Survey in
z & 1. The Figure also indicates that the typical time 2005 [157]. Superluminous supernovae have peak
scale of the light curves is a month, if we define the absolute magnitudes less than −21, and hence are
time scale by the full-width-half-maximum (FWHM) much more luminous than normal Type Ia and core-
of the light curve. Their shapes are simple with a rise collapse supernovae (see Figure 6). Their light curves
and a fall, although details are different for different are also wider, with the typical time scale of up to
types of supernovae. We note that these are template ∼ 100 days in the rest frame rather than a month.
light curves in the supernova rest frame. Observed Thanks to their bright luminosities, they can be
light curves of supernovae at cosmological distances are observed out to very high redshifts of z & 2 (e.g.,
stretched due to cosmological dilation by a factor of [158, 159]).
1 + z, which indicates that we expect the time scale of Finally, we summarize event rates and sizes of
a few months in the observer frame for supernovae at supernovae, which are important for strong lensing
z ∼ 1 − 2. studies. Li et al. [160] derived supernova rates in the
Studies of supernovae are important in several local Universe as RSNIa = (3.0 ± 0.6) × 104 Gpc−3 yr−1
ways. For instance, supernovae are associated with the for Type Ia and RSNcc = (7.1 ± 1.6) × 104 Gpc−3 yr−1
death of stars, and therefore their rates as a function for core-collapse (i.e., Type Ib/c and Type II). We
of galaxy type or redshift reflect the cosmic history of note that these rates increase toward higher redshifts,
star formation. Supernovae are produced only when mainly due to the increase of the cosmic star formation
masses of progenitor stars fall in a particular range, rate density from z ∼ 0 to ∼ 2. Quimby et al. [161]
from which we can obtain information on the stellar estimated the rate of superluminous supernovae at
initial mass function. One of the most important z ∼ 0.2 to RSLSN = (1.99+1.37 2 −3 −1
−0.86 ) × 10 Gpc yr .
applications of supernovae is the measurement of the The size of a supernova changes with time because of
cosmic expansion. It is known that peak luminosities the dynamical evolution of the photosphere. In the
of Type Ia supernovae are quite similar, which is case of supernovae, the ejecta is expected to enter the
particularly true if the empirical relation between peak
Strong lensing of explosive transients 13
S/N
gamma-ray bursts with supernovae is lacking (e.g., 5
[182]). A promising scenario that explains these
properties is that short gamma-ray bursts are caused
by binary mergers of compact objects such as neutron 0
stars and black holes. This scenario is confirmed by the
discovery of gravitational waves from a binary neutron
−20 −10 0 10 20
star merger, GW170817, for which the associated short time [msec]
gamma-ray burst GRB 170817A was detected (see
also Section 3.4). From BATSE, Swift, and Fermi
Figure 8. An example of light curves of fast radio bursts. Here
short gamma-ray samples, Wanderman and Piran [183] we show a dedispersed, averaged pulse profile of FRB 121102
+2.3
derived the local rate of RSGRB = 4.1−1.9 Gpc−3 yr−1 detected with the Arecibo Observatory [200].
49 −1
for L > 5 × 10 erg s . The rate rapidly increases
with increasing redshift, at least out to z ∼ 1.
There have been many proposals to use gamma- emissions, and sizes of X-ray and optical afterglows
ray bursts as standardizable candles to probe the should be three or more orders of magnitude larger
cosmic expansion history (see e.g., [184] for a review), than the values mentioned above.
just like Type Ia supernovae. Many luminosity
correlations that can be used to standardize gamma-
3.3. Fast radio bursts
ray bursts are proposed, including the correlation
between the time variability and the luminosity [185], Fast radio bursts, which are transient radio pulses
the isotropic energy and the rest-frame peak energy with the time scale of a millisecond, are a new
[186], the luminosity and the rest-frame peak energy class of transients that was identified relatively
[187], and the peak energy and the collimated energy recently (see [192, 193] for recent reviews). The first
[188]. Cosmology with gamma-ray bursts is potentially example of fast radio bursts was discovered with the
very powerful as the Hubble diagram can be extended Parkes Observatory in 2007 by Lorimer et al. [194].
to very high redshifts out to z > 8. Discoveries of additional four events by Thornton al.
The size of the emission region has also been [195] support the astrophysical origin of fast radio
studied in the literature (e.g., [189–191]). The size is bursts. Thanks to Canadian Hydrogen Intensity
estimated as Mapping Experiment (CHIME) [196] and Australian
Rem ∼ Γ2 c∆tvar , (51) Square Kilometre Array Pathfinder (ASKAP) [197],
−1/2 the number of known fast radio bursts is now rapidly
where Γ = 1 − (v/c)2 is the Lorentz factor of the increasing, and is reaching O(100) [198, 199]. An
ejecta with velocity v and ∆tvar is the variability time example of the light curve is shown in Figure 8.
scale. The Lorentz factor is thought to be typically A key quantity that characterizes each fast
O(100). From observed variabilities of light curves, radio burst is the dispersion measure. Because of
we have Rem ∼ 1013 cm for short gamma-ray bursts dispersive effects, electromagnetic waves propagate
and Rem ∼ 1014 cm for long gamma-ray bursts, albeit through a plasma with different speeds at different
with large uncertainties. Since the relativistic effect, frequencies. More specifically, electromagnetic waves
the transverse extent of the emission region Rtv differs with frequency ν have the following delay of the arrival
from Rem by a factor of Γ i.e., time
Rem e2 DM ν −2 DM
Rtv ∼ , (52) ∆t = ≈ 4150 sec ,
Γ 2πme c ν 2 MHz cm−3 pc
which suggests that, assuming Γ ∼ 300, Rtv ∼ 3 × (53)
1010 cm = 3 × 105 km for short gamma-ray bursts and where DM is the dispersion measure, which is
Rtv ∼ 3 × 1011 cm = 3 × 106 km for long gamma- essentially the column density of free election along
ray bursts, again with large uncertainties. Since the the line-of-sight
size that is relevant for strong lensing is Rtv , in what Z d
follows we refer to Rtv as the size of gamma-ray bursts. DM = ne dl, (54)
We note that these sizes are for gamma-ray prompt 0
Strong lensing of explosive transients 15
where d is the distance to the fast radio burst. Since star formation, although it is also unknown whether
the Universe is ionized at z . 6, the large contribution repeating and non-repeating fast radio bursts have the
from the intergalactic medium (IGM) to the observed same origin.
DM after subtracting the Galactic contribution is The volumetric rate of fast radio bursts is also still
expected. A useful approximation that relates the DM very uncertain, although it gives another important
from the IGM and redshift z is [192] clue to the origin. Assuming that observed fast radio
bursts are distributed out to z ∼ 1, we crudely obtain
DMIGM ≈ 1000 × z cm−3 pc, (55) RFRB ∼ 2 × 103 Gpc−3 yr−1 [192]. The recent study
by Lu and Piro [206] suggests an order of magnitude
which is reasonably accurate at least out to z ∼ 2. higher rate, ∼ 3 × 104 Gpc−3 yr−1 (see also [207]). In
The dispersion measures of fast radio bursts either case, the high event rate of fast radio bursts is a
discovered so far are typically 100 − 1000 cm−3 pc great promise for the future.
after subtracting the Galactic contribution, which The size of the emission region of fast radio
suggest their redshifts of ≈ 0.1 − 1 according to bursts is poorly constrained. The direct upper limit
equation (55). A complication is that there may of Rtv < 0.7 pc ∼ 2 × 1013 km is obtained from
also be contributions from host galaxies and local observations of the repeating fast radio burst FRB
environments. For example, if the source is surrounded 121102 with European VLBI Network [208], although
by a dense plasma, the contribution of the local this constraint is not quite tight. Tighter constraints
environment to DM can be as large as ∼ 1000 cm−3 pc of the size will greatly help discriminate different
and hence can be comparable or larger than DM progenitor models.
from the IGM. Therefore the redshift estimated by
equation (55) should be taken as the upper limit of
3.4. Gravitational waves
the source redshift.
Accurate distances to fast radio bursts are The existence of gravitational waves was predicted
obtained if their host galaxies are successfully by Albert Einstein in 1916 on the basis of General
identified. However, identifications of host galaxies Relativity. Gravitational waves are essentially the
have been challenging due to limited localization propagation of fluctuations of curvature in spacetime,
capabilities. The host galaxy was identified for the first but the strain amplitude is so small that its
time for a repeating fast radio burst, which represents detection has been quite challenging. The first direct
a rare class of fast radio bursts with repeating pulses. detection [209] was made in 2015 by the Advance
So far only two repeating fast radio bursts, FRB Laser Interferometer Gravitational-Wave Observatory
121102 [201] and FRB 180814.J0422+73 [202], have (Advanced LIGO; [210]). The event named GW150914
been identified, and in the former case the host was produced by a merger of a binary black hole
galaxy is identified to be a low-metallicity, low-mass with masses ∼ 36 M and ∼ 29 M located at
dwarf galaxy at z = 0.193 [203], which confirms the redshift z ∼ 0.09. Figure 9 shows the waveform of
extragalactic origin of fast radio bursts. GW150914, which was detected both in the Hanford
Very recently, host galaxies have been identified and Livingston detectors. The slight offset of the
for non-repeating fast radio bursts as well. A luminous arrival times and the relative amplitudes between the
galaxy at z = 0.3214 has been identified as a host two detectors contain information on the position of
galaxy of the non-repeating fast radio burst FRB the gravitational wave source on the sky. Since then,
180924 detected by ASKAP [204]. A massive galaxy the study of gravitational waves is progressing rapidly.
with a relatively low specific star-formation rate at See e.g., [211,212] for details of theory and experiments
z = 0.66 has been identified as a host galaxy of of gravitational waves.
FRB 190523 detected by the Deep Synoptic Array ten- Gravitational waves that are detectable with
antenna prototype [205]. These results highlight the Advanced LIGO are produced from mergers of binary
possibility of the association of fast radio bursts with black holes, binary neutron stars, and black hole-
relatively old stellar populations. neutron star binaries. Since Advanced LIGO can
The mechanism to produce these fast radio bursts detect gravitational waves in the frequency range
is still unknown. Many progenitor models that explain f ∼ 101−4 Hz, only gravitational waves at the final
fast radio bursts have been proposed (see [192] for a inspiraling and merging stages are observed. After
summary). The statistical analysis of a large sample the discovery of GW150914, there are more than
of fast radio bursts is a key for discriminating these 10 gravitational wave observations from binary black
different scenarios. Another important clue will be hole mergers out to z ∼ 0.5 with the total mass
obtained by identifying many host galaxies. For ranging from ∼ 20 M to ∼ 80 M (e.g., [213] for
example the host galaxy of FRB 121102 implies the the summary of the second observing run, and there
connection between fast radio bursts and massive are more observations from the third observing run).
Strong lensing of explosive transients 16
Table 1. Summary of explosive transients discussed in this review article. See the text in each Section for details and references.
Type Subclass Number zmax Wavelength Time scale Local rate Size
(f [Hz]) [Gpc−3 yr−1 ] [km]
Supernova Ia O(104 ) ∼2 optical ∼ 30 days ∼ 3 × 104 ∼ 1010
(Section 3.1) (∼ 1014−15 )
core-collapse O(104 ) ∼2 optical ∼ 30 days ∼ 7 × 104 ∼ 1010
(∼ 1014−15 )
superluminous O(100) ∼4 optical ∼ 100 days ∼ 200 ∼ 1010
(∼ 1014−15 )
Gamma-ray burst long > 5000 ∼9 γ a few sec ∼1 ∼ 106−7
(Section 3.2) (∼ 1018−23 )
short > 1000 ∼3 γ <sec ∼ 1 − 10 ∼ 105−6
(∼ 1018−23 )
Fast radio burst ··· O(100) ∼ 3? radio ∼msec ∼ 103.5−4.5 < 1013
(Section 3.3) (∼ 109 )
Gravitational wave BBH > 10 ∼ 0.5 LIGO band .sec ∼ 10 − 100 ∼ 100
(Section 3.4) (∼ 101−4 )
BNS ≥1 ∼ 0.05? LIGO band .sec ∼ 100 − 4000 ∼ 100
(∼ 101−4 )
BHNS 0 ··· LIGO band .sec < 600 ∼ 100
(∼ 101−4 )
of lensed quasar images for many years. Microlensing fraction of the baryonic matter consists of stars.
due to stars in lensing galaxies, which we will discuss in Lensing by these individual stars can significantly
more detail later, add additional variability to the light affects magnifications of individual multiple images
curve, making the robust measurement even more chal- if the source size is comparable or smaller than the
lenging. As a result, reliable measurements of quasar Einstein radii of the stars (see e.g., [241, 242]).
time delays require ∼ 10 yr monitoring observations, To illustrate the sensitive of microlensing variabil-
and the resulting accuracy on time delay measurements ities on the source size, in Figure 10 we show exam-
is on the order of ∼ 1 day (e.g., [227–229]). ples of flux variabilities as a function of the source
In contrast, since light curves of explosive size Rsrc , which is computed using the GPU-Enabled,
transients introduced in Section 3 are simple, we do High Resolution cosmological MicroLensing parame-
not need monitoring much beyond the time delay. In ter survey (GERLUMPH) microlensing magnification
the case of gamma-ray bursts, fast radio bursts, and maps [243, 244]. It is clear that microlensing variabil-
gravitational waves, their time scales of light curves ities are suppressed at Rsrc & REin due to the finite
are less than ∼ 1 sec, which indicates that time delays source size, where REin = Dos θEin is the Einstein ra-
can be measured with an accuracy better than ∼ 1 sec, dius in the source plane. Table 1 indicates that size
much better than current measurements with lensed of the explosive transients tend to be smaller than
quasars. This point has been discussed in [230–232] for REin , which suggests the importance of microlensing
gravitational waves, in [233] for gamma-ray burst, and in strong lensing of explosive transients.
in [234] for fast radio bursts. Repeated observations of The microlensing effect can be time dependent due
very precise time delay measurements with repeating to the transverse motions as well as the change of the
fast radio burst may allow us to directly measure the size of the emitting region with time, and therefore can
cosmic expansion [235]. We also expect accurate and distort the light curve in a non-trivial manner, which is
robust measurements of time delays for strong lensing particularly significant for strong lensing of supernovae
of supernovae, even though the time scale of their light [245]. Recent studies discuss possible ways to mitigate
curves is & 30 days, because of their simple and well- the effect of microlensing on supernova lensing [246–
known light curves [126, 236, 237]. 249], which indicates that a percent level measurement
However, one complication that may affects the of time delays is possible even in the presence of
accuracy and precision of time delay measurements microlensing by taking advantage of multiband light
is microlensing, which refers to flux variabilities due curves.
to stars in lensing galaxies (see e.g., [238–240]). The However, an exception is strong lensing of
contribution to the lens potential includes both dark gravitational waves. Thanks to their long wavelength,
and baryonic matter in the lensing object, and some microlensing variabilities are suppressed by the wave
Strong lensing of explosive transients 19
the energy. Moreover, some modified gravity theories ray bursts, which is referred to as femtolensing, was
predict the propagation speed of gravitational waves proposed in [127] (see also [283, 284]). This method
that differs from the speed of light. has been applied to Fermi Gamma-ray Burst Monitor
These effects can be tested by observations of data to place useful constraints on the abundance of
explosive transients (e.g., [218, 261–263]), by checking compact dark matter in the mass range mentioned
arrival time differences between different particles or above [285]. However, the finite source size effect,
energy. However the observed arrival time difference which has been ignored before, is in fact crucial in
consists of both the intrinsic time delay and the this application [137]. The recent study by Katz et
time delay caused by the different propagation speeds. al. [191] revisited constraints from femtolensing taking
The former is usually unknown or poorly constrained, full account of the finite source size effect to find that
which makes the result somewhat uncertain. One a useful constraint on the abundance of compact dark
interesting way to overcome this intrinsic time delay matter cannot be placed from the currently available
is to make use of strong lensing. This is because data.
the contribution of the intrinsic time delay vanishes The search of echo signals due to strong lensing
if we compare the difference of time delays between is possible also with other explosive transients. For
multiple images among different particles or energy. instance, the possibility of using fast radio bursts to
This idea has been applied to strong lensing of gamma- constraint compact dark matter with M & 20 M
ray burst [264] and gravitational waves [265–268]. A has been proposed in [286] and subsequently studied
caveat is that the difference between the propagation in [287, 288]. For the mass of ∼ 20 M , we expect
speeds of gravitational waves and their electromagnetic to observe multiple bursts separated by a typical time
counterparts might also be cause by the wave effect delay of a few milliarcseconds. It is found that ongoing
(Section 2.4) in gravitational lensing, because the experiments such as CHIME can place meaningful
propagation of gravitational waves is not affected by constraints on the abundance of compact dark matter
small intervening matter due to the diffraction effect in that mass range. The wave effect in strong lensing
[269]. of fast radio bursts and its application to the compact
Another test of fundamental physics includes the dark matter search was discussed in [289] and also
time variation of fundamental constants such as the noted in [286].
gravitational constant and the speed of light. Strong Although the similar search is possible with
lensing may also help for this type of test (e.g., [270]). strong lensing of gravitational waves, their long
wavelengths and compact sizes indicate that the wave
4.4. Compact dark matter effect definitely plays an important role (Section 2.4).
Again, equation (44) indicates that we need compact
There have been long discussions on whether dark dark matter with the mass M & 10 − 100 M to
matter is composed of unknown elementary particle avoid the diffraction and to observe strong lensing
or compact objects such as primordial black holes magnifications. When the mass is near the threshold,
(PBHs). The possibility of compact dark matter the signal-to-noise ratio of lensed waveforms shows
has been tested with various observations including an oscillatory behavior as the frequency sweeps up
microlensing in and around the Milky Way (see e.g., due to the wave effect, which can be regarded as a
[271]), from which constraints on the abundance of smoking gun signature of strong lensing [128]. Strong
compact dark matter are derived as a function of the lensing of gravitational waves by compact objects is
mass of compact dark matter. recently revisited after the first direct observation of
Strong lensing of explosive transients helps gravitational waves from binary black hole mergers,
improve these constraints. For instance, strong including the rate estimate and expected constraints
lensing or microlensing of gamma-ray bursts and their on the abundance of compact dark matter [290–293].
afterglow emissions have been studied extensively as The compact dark matter scenario can be tested
a means of testing the compact dark matter scenario with strong lensing of supernovae as well. In particular,
[272–282]. The ideas include the search for echo strong lensing of Type Ia supernovae by compact dark
signals in gamma-ray bursts and the modification of matter produces a non-Gaussian tail in their apparent
the afterglow light curve due to the size dependence of magnitude distribution for a given redshift, from which
microlensing. useful constraints on compact dark matter for a wide
When the mass of compact dark matter is very mass range M & 0.01 M are obtained [294].
small, from ∼ 10−13 M to ∼ 10−16 M the wave effect In testing the compact dark matter scenario with
(Section 2.4) becomes important even in gamma-ray. strong lensing, the wave effect and the finite source
In this case, the interference between multiple images size effect sometimes become very important. To
induces an oscillating feature in the photon energy guide future studies along this line, we revisit these
spectrum. The application of this effect to gamma-
Strong lensing of explosive transients 21
108 105
106 w =1 no lensing
104 104 due to
diffraction wave optics
102
100 103
1/β̂s = DosθEin/Rs
(1 + zl)M [M]
10−2 nt
M
ere
10−4 2 diff geometric optics
10
10−6
10−8
101
10−10
10−12
100
10−14 M
nt
10−16 ere
−1 diff
10−18 GW GW 10
(space) (ground) radio IR Opt. X γ no lensing due to
FUV
10−20 −4 −2 0 2 4 6 8 10 12 14 16 18 20 22 finite source size
10 10 10 10 10 10 10 10 10 10 10 10 10 10 10−2 −2
f [Hz] 10 10−1 100 101 102 103 104 105
w
Figure 11. The relation between the frequency f and
the (redshifted) mass (1 + zl )M of a point mass lens for Figure 12. The schematic illustration of regions relevant for
the dimensionless parameter w = 1, where w is defined in geometric optics and wave optics. The solid line shows w = 1/β̂s ,
equation (44). The region below the solid line corresponds to the where w is defined in equation (44) and β̂s = Rs /(Dos θEin ) is the
case that the gravitational lensing magnification is significantly source size normalized by the Einstein radius. The region above
suppressed due to the wave effect i.e., the diffraction. the solid line corresponds to the situation where the interference
pattern due to the wave effect may be observed, whereas the
region below the solid line corresponds to the situation that the
geometric optics approximation is relevant. The shaded regions
effects introduced in Section 2.4 and discuss them more show w < 1 and 1/β̂s < 1, for which the gravitational lensing
quantitatively. magnification is significantly suppressed due to the diffraction
and the finite source size effect, respectively. The dotted lines
First, an important parameter that controls the show the direction along which parameter values change by
wave effect is the dimensionless parameter w defined in changing the lens mass M .
equation (44). When w < 1, the diffraction originating
from the wave optics becomes so effective that the
gravitational lensing magnification is highly suppressed lensing magnification is significantly suppressed due
i.e., µ ∼ 1 irrespective of the impact parameter. In to the diffraction and the finite source size effect,
Figure 11, we show the relation between the frequency respectively. Therefore in these regions we do not
f and the (redshifted) lens mass (1 + zl )M that satisfy observe any gravitational lensing effect.
w = 1 for the case of a point mass lens. We note that Figure 12 has several important implications.
the similar relation hold for other lens mass models, Since w and 1/β̂s depend on the mass
once M is replaced to the enclosed mass within the √ M of a point
mass lens as w ∝ M and 1/β̂s ∝ M , respectively,
Einstein radius (see equation 36). The region below the parameter values change along the direction indicated
line in Figure 11 corresponds to w < 1, and hence to by the dotted lines. This indicates that in the limit
the diffraction. Figure 11 clearly demonstrates that the M → ∞ the parameter values always fall in the
wave effect is particularly important for gravitational geometric optics region, which is one of the reasons why
waves. the geometric optics approximation is valid in most
As discussed in Section 2.4, in order for the astronomical situations. Figure 12 also suggests that
interference pattern due to the wave effect to be the interference pattern due to the wave effect may be
observed, the source must be sufficiently compact. observed only when 1/β̂s > 1 at w = 1, as in the case of
This condition (equation 49) is given as w . 1/β̂s , the upper dotted line in Figure 12. In contrast, in the
where β̂s = Rs /(Dos θEin ) is the source size Rs case of the lower dotted line in Figure 12, the geometric
normalized by the Einstein radius. We illustrate optics approximation is valid in all the parameter range
this condition in Figure 12. The region above the of interest. From equations (44) and (50), the necessity
line w = 1/β̂s corresponds to the situation where condition that the interference pattern is observed for
the interference pattern due to the wave effect may some lens masses is written as
be observed. We note that this is just a necessity −1/2 −1/2
condition, and in order for the interference pattern 13 1 + zl f
Rs . 3.05 × 10 km
to be observe other conditions such as the frequency 1.5 Hz
band should also be met [134]. We also note that in 1/2
Dos Dls /Dol
regions with w < 1 and 1/β̂s < 1 the gravitational × , (58)
0.949 Gpc
Strong lensing of explosive transients 22
no wave effect
For instance, strong lensing of various transients
1014 FRB can be discovered by monitoring massive clusters
1012 of galaxies, which are known to be efficient lenses.
SN Time delays obtained from measurements of multiple
1010 images of explosive transients break degeneracies
in mass models reconstructed from multiple image
Rs [km]
to detect substructures more directly. Lensing by in the high mass end of the observed chirp mass
substructures with masses ∼ 103−6 M can induce distribution [315]. Furthermore, strong lensing of
interference pattern in waveforms, as in the case of gravitational waves produce multiple images, some of
microlensing by compact dark matter discussed in which are demagnified. Such demagnified images can
Section 4.4. This possibility has been explored in [304]. be observed as apparently very high redshift events,
and hence produce a tail at the high end of the observed
4.6. The nature of explosive transients redshift distribution [123]. These examples highlight
the critical importance of gravitational lensing for the
As discussed in Section 3, the true nature of the interpretation of observe distributions of gravitational
explosive transients that are considered in this review waves.
article is yet to be fully understood. Strong lensing When multiple images of explosive transients are
may help reveal their true nature by taking advantage produced, in a sense we observe the transients multiple
of its magnifying power as well as its power to resolve times with some time differences. If we can predict the
fine structures of sources. appearance of trailing images, it opens up interesting
First, the result in Section 2.3 indicates that applications such as the detailed monitoring of early
the strong lensing probability is a steep function of light curves. This possibility was noted in [126] for
the source redshift. Therefore, we can constrain the strong lensing of supernovae, and was explored in
redshift distribution of explosive transients from their detail in [316]. A particularly interesting feature in
strong lensing probabilities. The idea was used in [305] the early light curves of supernovae is the so-called
to constrain the redshift distribution of gamma-ray shock breakout, which is a luminosity emission with
bursts. The similar idea was proposed for fast radio very short time scale.
bursts in [306]. Strong lensing can also be used to resolve fine
Since strong lensing magnifies background sources, structures of sources by e.g., taking advantage of the
it enables us to observe very distant events that size dependence of microlensing as shown in Figure 10.
cannot be observed without the gravitational lensing The ideas to resolve the jet structure of gamma-
magnification. Therefore, we can constrain the ray bursts with strong lensing have been explored in
supernova rate at very high redshifts by observations [317–320]. For strong lensing of repeating fast radio
of strongly lensed supernovae at such high redshifts bursts, one can measure the change of time delays
[126, 307, 308]. The search of lensed high-redshift between multiple images, from which the motion of
supernovae can be conducted efficiently by monitoring fast radio bursts is measured [321].
massive clusters of galaxies [309–314].
The example above immediately suggests that
5. Past observations and future prospects
the gravitational lensing magnification modifies the
observed distribution of explosive transients. This may 5.1. Past observations
be particularly important for gravitational waves from
binary black hole mergers for which redshifts are not 5.1.1. Strong lensing of supernova: Before discoveries.
directly measured in most cases. Instead, as discussed The expected event rates of strongly lensed supernovae
in Section 3.4, from gravitational wave observations in various supernova surveys have been computed
one can measure the luminosity distance to the [112, 236, 322–325], which suggest that future surveys
source. However, in presence of gravitational lensing that are aimed at finding many supernovae at z & 1
magnification µ, the observed luminosity distance is should also be able to find strongly lensed supernovae.
modified as One of the most comprehensive predictions before the
D̄L first discoveries has been made in [93], in which it was
DLobs = √ , (59)
µ argued that Pan-STARRS1 [326, 327] can find O(1)
strongly lensed supernovae, whereas Large Synoptic
where D̄L is the luminosity distance to the source in
Survey Telescope (LSST) [225] can find more than 100
absence of gravitational lensing (i.e., the luminosity
strongly lensed supernovae.
distance to the source redshift computed assuming
We can efficiently search for strongly lensed
a homogeneous and isotropic Universe) and DLobs is
supernovae by monitoring plausible sites, such as
the luminosity distance measured from observations of
galaxy-galaxy strong lens systems [328] and massive
gravitational waves. Therefore, for highly magnified
clusters [329–332]. The latter search led to discoveries
events µ 1, the redshift inferred from the luminosity
of some supernovae behind clusters, which are
distance is based low. The bias in the estimated
magnified but not multiply imaged [298–300, 333–335].
redshift directly affects the estimate of the chirp mass
via equation (56). Indeed it is pointed out that
strong lensing magnification produces an apparent tail 5.1.2. Discovery of PS1-10afx. Pan-STARRS1 Medium
Deep Survey (see [327]) is a time-domain survey with
Strong lensing of explosive transients 24
Table 2. Summary of strongly lensed supernovae discussed in this review article. See the text in each Section for details and
references. Nimg indicates the number of multiple images, mpeak is an observed peak magnitude (the total magnitude for PS1-10afx
and iPTF16geu, and the magnitude of the brightest image for SN Refsdal), µtot is the total magnification factor of all the multiple
images, which is directly measured from the observation for Type Ia, θmax is the maximum image separation between any multiple
image pairs, and ∆tmax is the maximum time delay between any multiple image pairs. Note that the values listed here can be either
observed or model predicted ones.
Name Type zs zl Nimg mpeak µtot θmax ∆tmax
PS1-10afx (Section 5.1.2) Ia 1.388 1.117 4? i ∼ 22 ∼ 31 < 0.400 < 4 days
SN Refsdal (Section 5.1.3) II 1.49 0.54 6 i ∼ 27 ∼ 74 ∼ 3200 ∼ 6000 days
iPTF16geu (Section 5.1.4) Ia 0.409 0.216 4 i ∼ 19 ∼ 52 ∼ 0.600 < 1 days
The high total magnification of ∼ 52 is partly supernovae near elliptical galaxies whose absolute
explained by the selection effect, which will be magnitudes computed from the redshifts of the
discussed in detail in Section 5.2. However, it has been elliptical galaxies are brighter than those of Type
found from detailed comparisons with theoretically Ia supernovae. This search method is based on
expected distributions that the observed magnification the fact that lensing galaxies are dominated by
is higher than expected given its redshift, even if elliptical galaxies. This method also enables rapid
we take account of the selection effect [357, 359, 362]. identifications of unresolved strong lensing events and
This issue may be related with the anomalous flux potentially increases the number of strongly lensed
ration mentioned above. It is of great importance supernovae discovered by LSST (see also [362]).
to understand the cause of the anomalous flux ratios Even if multiple images of strongly lensed
and the high total magnification for the future use of supernova are barely resolved, we may still be able to
strongly lensed Type Ia supernovae for cosmology. see its signature by carefully checking the morphology
of the supernova image to see if it is really consistent
5.1.5. Implications of the first discoveries for search with the Point Spread Function. The possibility
methods. Table 2 summarizes properties of the three of finding strong lensed supernovae by checking the
strongly lensed supernovae presented in this review ellipticity of the supernova image is discussed in [364].
article. Two out of the three events have very small Recent work by Wojtak et al. [365] explored
image separations such that they are barely resolved in how effective such new strategy to find unresolved
ground-based imaging observations. Such unresolved strongly lensed supernovae is in ongoing and future
events were not included in the calculation of [93]. time-domain surveys. It was found that finding
If we can identify these unresolved strong lensing unresolved strongly lensed supernovae increases the
events from the survey data in a timely manner, number of strongly lensed supernovae drastically for
we may be able to increase the number of strongly shallow surveys such as Palomar Transient Factory,
lensed supernovae discovered in future time-domain whereas the increase of the number is modest for deep
surveys to enhance their power for cosmological and time-domain surveys such as LSST.
astrophysical studies.
Based on the discovery of PS1-10afx, Quimby 5.1.6. Strong lensing of gamma-ray burst. One of the
et al. [338] proposed a new method to identify most comprehensive discussions on the detectability of
strongly lensed Type Ia supernovae, utilizing a color- strong lensing of gamma-ray bursts has been presented
magnitude diagram of supernovae. Specifically, it was in [366]. Although the Swift may be able to detect
found that strongly lensed Type Ia supernovae are strongly lensed gamma-ray bursts, it is argued that
well separated from unlensed supernovae in i-band detecting multiple image pairs is unlikely because of
magnitude versus r − i color diagram, which allows its inefficient duty cycle and the limited sky coverage,
us to identify unresolved strong lensing candidates ∼ 50% (see also [367]). Fermi Gamma-ray Burst
relatively securely. Rapid follow-up observations Monitor has more sky coverage and therefore may be
of these candidates may lead to measurements of suited to search for multiple image pairs in this regard.
time delays for these strong lensing events. It was Despite some explicit attempts to search for lensed
argued that this approach can significantly increase image pairs in the gamma-ray burst catalogs for a wide
the number of strongly lensed supernovae discovered range of time delays [368–371], no secure candidate of
by LSST. multiply imaged gamma-ray bursts has been identified
Goldstein and Nugent [363] proposed a slightly so far. The latest search by Hurley et al. [371] makes
different approach, in which strongly lensed Type use of the gamma-ray burst sample detected by Konus-
Ia supernova candidates are identified by identifying Wind [372], which has the high duty cycle and large sky
Strong lensing of explosive transients 27
coverage, to search for lensed image pairs. Based on effect of the Earth rotation that changes the sensitivity
the absence of any candidate of strongly lensed gamma- to a source located in a given position on the sky as a
ray burst, an upper limit of the lensing probability function of time [384]. These issues are mitigated by
of 0.0033 is placed. A caution is that microlensing increasing the number of detectors in operation.
can distort light curves of strongly lensed gamma-ray There are some attempts to explicitly search for
bursts (e.g., [373]), which may affect the efficiency strongly lensed gravitational wave events among sub-
of searching for multiple image pairs based on the threshold signals, but no promising candidate of strong
similarity of the light curves. lensing events is identified [385, 386]. Since strong
lensing of gravitational waves may be produced by
5.1.7. Strong lensing of fast radio burst. While some foreground galaxy clusters, the search of strong lensing
estimates of expected event rates of strongly lensed events can also be conduced around known massive
fast radio bursts have been presented in the literature clusters within error circles of observed gravitational
(e.g., [234, 306]), so far no systematic search for strong wave events [387, 388].
lensing of fast radio bursts has been made. Since the
number of observed fast radio bursts is very rapidly 5.2. Expected event rates
increasing, the future search in real catalogs will be
Here we present some discussions on expected event
interesting.
rates of strongly lensed explosive transients. Our
strategy here is to provide simple and concise estimates
5.1.8. Strong lensing of gravitational waves. The ex- of strong lensing rates for various explosive transients
pected rates of strongly lensed gravitational waves have in a unified manner. Since we do not fully take account
been computed both for ground based experiments of the luminosity distributions and selection functions,
(e.g., [123,374–378]) and space based experiments (e.g., these estimates are not very accurate, but a reward of
[123, 129, 379, 380]). These calculations suggest that this simple approach is that the parameter dependence
a large number of strongly lensed gravitational waves is clearer that leads to the better understanding of
from compact binary mergers will be discovered in fu- differences of expected strong lensing rates in different
ture third-generation ground-based experiments as well setups. Our approach here is also complementary to
as future space-based gravitational wave experiments. detailed calculations of event rates of strong lensing
The expected rates of strongly lensed gravitational taking full account of the luminosity distributions
waves in the previous and ongoing Advanced LIGO and the selection effects, e.g., [93, 362, 365] for strong
observing runs are predicted to be small (e.g., [123, lensing of supernovae and [123, 377] for strong lensing
377, 378]). However, Broadhurst et al. [381, 382] made of gravitational waves.
an interesting claim that roughly half of gravitational We start with the strong lensing probabilities
waves from binary black hole mergers detected by derived in Section 2.3. The strong lensing probability
Advanced LIGO are in fact strongly lensed ones. as a function of the source redshift, Psl (zs ) defined by
As discussed in Section 4.6, estimated redshifts and equation (30), is computed following the Monte-Carlo
chirp masses of highly magnified gravitational wave approach [123] assuming single galaxies as lensing
events are biased if gravitational lensing is not taken objects. The galaxy mass distribution is modeled
into account, such that highly magnified high redshift by a Singular Isothermal Ellipsoid plus an external
events are observed as low redshift events with very shear (see Section 2.3 for more details). Since groups
large chirp masses. Therefore, binary black holes with and clusters of galaxies are not included in the
relatively high masses of ∼ 30 M , if interpreted as calculation, these strong lensing probabilities are likely
highly magnified events, are in fact binary black hole to be underestimated, although their contribution to
systems with moderate masses, ∼ 10 M . In order for the total strong lensing probability is thought to be
such events to contribute to the current observation, subdominant as discussed in Section 2.3, especially for
the redshift evolution of the event rate must be very strong lensing of explosive transients whose sizes are
strong such that the event rate at z ∼ 1 − 2 is several compact.
orders of magnitude higher than the local event rate. Given the strong lensing probabilities, we can
For highly magnified strong lensing events, we compute the expected observed rate of strongly lensed
expect a pair of images with similar waveforms [383], explosive transients at z < zmax as
which are observed with a typical time difference of
Z zmax
less than a day [123]. While the absence of such pair d2 V R(zs )
events in Advanced LIGO observations may disfavor Rsl (< zmax ) = Ωsky dzs Psl (zs ),
0 dzs dΩ 1 + zs
the lensing scenario mentioned above, it is possible that (60)
such counterimages are missed due to the relatively low where R(z) is the comoving rate density of explosive
duty cycle (“glitches” in the data stream) as well as the transients as a function of redshift and Ωsky is the sky
Strong lensing of explosive transients 28
where Rloc is the local event rate and αz parametrizes Figure 16. The normalization of the observed strong lensing
the redshift evolution. See Table 1 for the current norm (< z
rate, Rsl max ; αz ) defined in equation (62), as a function
estimates of Rloc for various explosive transients. In of the maximum redshift zmax . We show Rsl norm for several
many cases, the redshift evolution of event rates of different redshift evolution parameter αz that is introduced
in equation (61). The fitting form of Rsl norm is given by
explosive transients traces the global star formation equation (63).
history of the Universe, for which the rate increases
toward higher redshifts out to z ∼ 2, with the slope
corresponding to αz ≈ 2. With these assumptions, the
We use equations (60) and (63) to compute
observed rate (60) is rewritten as
expected event rates of strong lensing of various
norm Ωsky B explosive transients as a function of zmax . For each
Rsl (< zmax ) = Rsl (< zmax ; αz )
4π 0.04 explosive transient listed in Table 1, we choose a
loc fiducial value of Rloc that is consistent with the current
R
× , (62) estimates. We also choose the redshift evolution
1 Gpc−3 yr−1
norm parameter defined in equation (61) to αz = 1 or 2 so
where Rsl (< zmax ; αz ) is calculated by inserting
that it is broadly consistent with the current estimates.
the fiducial values to equation (60) and adopting an
Table 3 summarizes our fiducial choices of Rloc and αz ,
approximation given by equation (31).
norm and resulting expected observed rates for zmax = 0.5,
Figure 16 shows Rsl (< zmax ; αz ) for several
1, 2, and 3. Since we adopt B = 0.04, these correspond
different choices of αz . It is found that the expected
to rates of strong lensing events with µtot & 10. We
observed strong lensing rate is a steep function of
note that these predictions can be easily modified to
zmax . At low redshift zmax 1, we roughly have
norm 6 those for other parameter sets by using equation (60).
Rsl (< zmax ; αz ) ∝ zmax , in contrast to the unlensed
Results in Table 3 do not take account of the
event rate which is proportional to the volume at low
3 observability. We provide a rough estimate of zmax
redshifts i.e., ∝ zmax .
for each survey as follows. For each survey, we
We find that the results shown in Figure 16 are
first estimate the redshift zlim , out to which normal
fitted by the following form
unlensed events are largely detected. Since we consider
a2
norm a1 zmax (1 + zmax )αz strong lensing events µtot & 10, we assume that each
Rsl (< zmax ; αz ) ≈ a4 , (63) multiple image is magnified by a factor of ∼ 4 or so.
1 + a3 zmax
The magnification factor of 4, for instance, indicates
√
a1 = 3 × 10−3 , (64) that the event is detected out to a factor of 4 = 2
larger luminosity distance. Therefore, for each survey
and explosive transient with zlim , we assume that the
a2 = 5.8, (65) strong lensing events are detected out to zmax that
satisfies
a3 = 4.6 + 0.35 αz (66) DL (zmax ) = 2 DL (zlim ). (68)
At sufficiently low redshifts, this relation implies
a4 = 3.1 + 0.1 αz0.6 . (67) zmax ≈ 2zlim .
Figure 17 shows expected observed rates of strong
This fitting form is derived in the range 0 ≤ αz ≤ 5
lensing of supernovae along with rough estimates of
and zmax < 5.
Strong lensing of explosive transients 29
Table 3. Expected observed rates of strong lensing of various transients computed using equation (62). See also Table 1 for the
summary of properties of these transients. The columns Rloc and αz show fiducial values of the local event rate and the redshift
evolution parameter adopted in the calculation. Expected observed rates Rsl for Ωsky = 4π and B = 0.04 (corresponding to
µtot & 10) within the maximum redshift zmax = 0.5, 1, 2, and 3 are shown.
Type Subclass Rloc αz Rsl (< 0.5) Rsl (< 1) Rsl (< 2) Rsl (< 3)
[Gpc−3 yr−1 ] [sky−1 yr−1 ] [sky−1 yr−1 ] [sky−1 yr−1 ] [sky−1 yr−1 ]
Supernova Ia 3 × 104 1 1.6 30 320 1300
core-collapse 7 × 104 2 5.4 130 2000 10000
superluminous 200 2 0.02 0.38 5.8 29
Gamma-ray burst long 1 2 < 0.01 < 0.01 0.03 0.15
short 3 1 < 0.01 < 0.01 0.03 0.13
Fast radio burst ··· 104 2 0.78 19 290 1500
Gravitational wave BBH 30 2 < 0.01 0.06 0.88 4.4
BNS 600 1 0.03 0.61 6.5 25
BHNS 10 1 < 0.01 0.01 0.11 0.4
104 101
Rsl(< zmax; B = 0.04) [sky−1yr−1]
SNIa
102 10−1
101 10−2
100 10−3
10−1 10−4
10−2 10−5
0.0 0.5 1.0 1.5 2.0 2.5 0 1 2 3 4 5
redshift zmax redshift zmax
Figure 17. Expected observed rates of strong lensing of Figure 18. Same as Figure 17, but for strong lensing of gamma-
supernovae as a function of the maximum redshift zmax ray bursts and rough estimates of zlim for Swift [172].
computed using equation (60). See Table 3 for the fiducial values
adopted in the calculation. We show all-sky (Ωsky = 4π) rates
with B = 0.04 that corresponds to strong lensing events with
µtot & 10. For each supernova type, a rough estimate of zlim estimates of zmax for Swift [172]. Although the event
and the corresponding zmax (see equation 68) for LSST [225] are
marked by left and right circles, respectively. rates of gamma-ray bursts are low, thanks to the high
mean redshift strong lensing of long gamma-ray bursts
can in principle be observed, although one limitation
is its inefficient duty cycle as discussed in [366].
zmax for LSST [225]. The LSST monitors a half sky,
Figure 19 shows expected observed rates of strong
but the survey is conduced for 10 years. Therefore
lensing of fast radio bursts along with a rough estimate
in their survey duration we expected to discover the
of zlim for CHIME [199]. Thanks to the high event rate,
significant number of strongly lensed Type Ia and core-
the expected rate of strong lensing is also high, but
collapse supernovae, which is consistent with more
CHIME observes the sky for the area of ∼ 250 deg2
detailed estimates (e.g., [93,362,365]). The calculation
and therefore a factor of 250/41200 ≈ 0.006 should
also suggests that we may be able to discover strongly
be multiplied to obtain the actual expected observed
lensed superluminous supernovae. These are results
rate in CHIME. While this suggests that O(1) strong
for the wide survey of LSST, whereas LSST is also
lensing events per a few years are expected from
planning to conduct deep drilling fields survey where
CHIME, we caution that this estimate can easily
strongly lensed supernovae at higher redshifts may be
change by an order of magnitude or more given the
discovered.
quite large uncertainties of their event rate and redshift
Figure 18 shows expected observed rates of
distribution.
strong lensing of gamma-ray bursts along with rough
Figure 20 shows expected observed rates of
Strong lensing of explosive transients 30
BBH
BNS by
0
10 BHNS dRsl z 4
lim
∝ z η−1 = z η−5 zlim
4
. (71)
dz z
10−1 By taking the ratio of equations (69) and (71), we have
10−2
η−5
dRsl /dz z
∝ >1 (η > 5). (72)
dRsl /dzlim zlim
−3
10
Since equation (72) is an increasing function of z, we
10 −4 preferentially observe strong lensing events with z
zlim i.e., µ 1, which qualitatively explains the high
magnification factors of PS1-10afx and iPTF16geu. In
10−5
0.0 0.5 1.0 1.5 sufficiently deep surveys, on the other hand, Figure 16
redshift zmax implies that the slope of the strong lensing rate
becomes shallower, η < 5, for which this argument
Figure 20. Same as Figure 17, but for strong lensing of no longer holds so that strongly lensing events with
gravitational waves and rough estimates of zlim for the Advanced
LIGO design sensitivity [210].
modest magnifications are preferentially observed,
although the detail depends also on the shape of the
luminosity function. We expect that, at least for
sufficiently shallow surveys such as CHIME for fast
strong lensing of gravitational waves along with rough radio bursts and Advanced LIGO for gravitational
estimates of zmax for the Advanced LIGO design waves, we typically observe highly magnified events
sensitivity [210]. This result indicates that, albeit the with redshifts well beyond the redshift limit of unlensed
probability is not very high, it may be possible to events. This point has also discussed in e.g., [123] in
detect strongly lensed gravitational waves of binary the context of strongly lensed gravitational waves.
black hole mergers in Advanced LIGO, which is
broadly consistent with more detailed calculations
(e.g., [123, 377]). 6. Conclusions
Finally, we discuss typical magnifications of
In this article, we have reviewed strong lensing of ex-
strongly lensed explosive transients detected in sur-
plosive transients, specifically focusing on supernovae,
veys. Table 2 indicates that magnifications of those
gamma-ray bursts, fast radio bursts, and gravitational
observed strongly lensed supernovae are high in gen-
waves from compact binary mergers. Although many
eral, µtot & 30, which may appear odd given the steep
strongly lensed quasars and galaxies have already been
magnification probability distribution of dP/dµ ∝ µ−3 .
identified, strong lensing of these explosive transients is
The probability distribution suggests that such highly
complementary to those traditional strong lensing and
magnified events are much rarer than strong lensing
enables new applications that was not possible before.
Strong lensing of explosive transients 31
In this article we have discusses possible applications discussed selection effects, and showed that highly
of these new strong lensing events, summarized the magnified strong lensing events are preferentially
current status of strong lens searches, and presented observed in shallow surveys.
expected rates of strong lensing events adopting a sim- In this review article, we have not covered all
plified approach. explosive transients at cosmological distances. For
Rapidly evolving light curves of these transients instance, the tidal disruption event is a disruption of
indicate that we expect a lot of progress in a star by the tidal force of the back hole (see e.g.,
applications of time delays between multiple images. [389]), which can also act as a source of strong lensing.
In particular for gamma-ray bursts, fast radio bursts, Furthermore, time-domain surveys may identify new
and gravitational waves, thanks to their very short types of transients that are similar to strong lensing of
time scales of . 1 sec we can drastically improve the explosive transients. One such example is a caustic
accuracy of time delay measurements as compared with crossing. When a star in a gravitationally lensed
the current accuracy for strongly lensing quasars, ∼ galaxy passes through a caustic it is magnified by a
1 day. The very accurate measurements of time delays factor of several thousands or more and hence can
open new avenues, including improved constraints on be detected even at cosmological distances. Recent
cosmological parameters such as the Hubble constant discoveries of caustic crossings of individual stars at
H0 , a probe of small-scale perturbations from dark z ∼ 1 − 1.5 [390–393] have attracted a lot of attention.
matter substructures, tests of fundamental physics The caustic crossing has the time scale of the light
from the propagation speed, and constraints on the curve near the peak as short as several days, and
abundance of compact dark matter from the search of therefore is definitely an interesting target to find in
pair events with short time delays. future time-domain surveys, in addition to strongly
The compact sizes of these explosive transients lensed explosive transients. In coming years, these
imply that wave optics effect may play an important new time-variable strong lensing events will deepen our
role. When the frequency is comparable to the inverse view of the Universe in several ways.
of the typical time delay, the wave effect induces the
interference pattern as a function of the source position Acknowledgments
and frequency, although in order for this effect to
be observed the source size must be smaller than I thank Kazumi Kashiyama for useful discussions. This
the width of the interference pattern. On the other work was supported in part by World Premier Inter-
hand, the frequency is much lower than the inverse national Research Center Initiative (WPI Initiative),
of the typical time delay, lensing magnifications are MEXT, Japan, and JSPS KAKENHI Grant Numbers
suppressed due to the diffraction. We have presented JP18H04572, JP15H05892, and JP18K03693.
detailed discussions on whether the wave effect is
relevant for strong lensing of these explosive transients, References
and argued that the wave effect can become important
for strong lensing of gravitational waves, and probably [1] P. Coles, Einstein, Eddington and the 1919 Eclipse, in
for strong lensing of fast radio bursts as well depending Historical Development of Modern Cosmology, V. J.
Martı́nez, V. Trimble and M. J. Pons-Borderı́a, eds.,
on their actual sizes. vol. 252 of Astronomical Society of the Pacific
Strong lensing may also help better understand Conference Series, p. 21, Jan, 2001,
these explosive transients. Thanks to lensing astro-ph/0102462.
magnifications, we can detect very high redshift events [2] A. Einstein, Lens-Like Action of a Star by the Deviation
of Light in the Gravitational Field, Science 84 (1936)
that are not accessible without lensing magnifications. 506.
Lensing magnifications can also be used as a [3] F. Zwicky, Nebulae as Gravitational Lenses, Physical
microscope to resolve the fine structure of sources to Review 51 (1937) 290.
[4] F. Zwicky, On the Probability of Detecting Nebulae
constrain their progenitor models.
Which Act as Gravitational Lenses, Physical Review
We have presented the current status and future 51 (1937) 679.
prospect of the strong lens search. We have described [5] D. Walsh, R. F. Carswell and R. J. Weymann, 0957 +
recent discoveries of strongly lensed supernovae, PS1- 561 A, B - Twin quasistellar objects or gravitational
lens, Nature 279 (1979) 381.
10afx [338], SN Refsdal [339], and iPTF16geu [357]. [6] R. Lynds and V. Petrosian, Giant Luminous Arcs in
We have computed expected observed rates of strong Galaxy Clusters, in Bulletin of the American
lensing of various explosive transients adopting a Astronomical Society, vol. 18 of Bull. Am. Astron.
simple and concise approach. These calculations Soc., p. 1014, Sept., 1986.
[7] G. Soucail, B. Fort, Y. Mellier and J. P. Picat, A blue
suggest that strong lensing of gamma-ray bursts, fart ring-like structure, in the center of the A 370 cluster
radio bursts, and gravitational waves, can be observed of galaxies, Astron. Astrophys. 172 (1987) L14.
in near future. Using this simple model, we have [8] S. T. Myers, N. J. Jackson, I. W. A. Browne, A. G. de
Bruyn, T. J. Pearson, A. C. S. Readhead et al., The
Strong lensing of explosive transients 32
Cosmic Lens All-Sky Survey - I. Source selection and J.-P. Kneib et al., The CFHTLS-Strong Lensing
observations, Mon. Not. R. Astron. Soc. 341 (2003) 1 Legacy Survey (SL2S): Investigating the Group-scale
[astro-ph/0211073]. Lenses with the SARCS Sample, Astrophys. J. 749
[9] I. W. A. Browne, P. N. Wilkinson, N. J. F. Jackson, S. T. (2012) 38 [1109.1821].
Myers, C. D. Fassnacht, L. V. E. Koopmans et al., [24] M. Negrello, R. Hopwood, G. De Zotti, A. Cooray,
The Cosmic Lens All-Sky Survey - II. Gravitational A. Verma, J. Bock et al., The Detection of a
lens candidate selection and follow-up, Mon. Not. R. Population of Submillimeter-Bright, Strongly Lensed
Astron. Soc. 341 (2003) 13 [astro-ph/0211069]. Galaxies, Science 330 (2010) 800 [1011.1255].
[10] M. Oguri, N. Inada, B. Pindor, M. A. Strauss, G. T. [25] J. L. Wardlow, A. Cooray, F. De Bernardis, A. Amblard,
Richards, J. F. Hennawi et al., The Sloan Digital Sky V. Arumugam, H. Aussel et al., HerMES: Candidate
Survey Quasar Lens Search. I. Candidate Selection Gravitationally Lensed Galaxies and Lensing Statistics
Algorithm, Astron. J. 132 (2006) 999 at Submillimeter Wavelengths, Astrophys. J. 762
[astro-ph/0605571]. (2013) 59 [1205.3778].
[11] N. Inada, M. Oguri, M.-S. Shin, I. Kayo, M. A. Strauss, [26] J. D. Vieira, D. P. Marrone, S. C. Chapman, C. De
T. Morokuma et al., The Sloan Digital Sky Survey Breuck, Y. D. Hezaveh, A. Weiβ et al., Dusty
Quasar Lens Search. V. Final Catalog from the starburst galaxies in the early Universe as revealed by
Seventh Data Release, Astron. J. 143 (2012) 119 gravitational lensing, Nature 495 (2013) 344
[1203.1087]. [1303.2723].
[12] A. More, M. Oguri, I. Kayo, J. Zinn, M. A. Strauss, [27] Y. D. Hezaveh, D. P. Marrone, C. D. Fassnacht, J. S.
B. X. Santiago et al., The SDSS-III BOSS quasar lens Spilker, J. D. Vieira, J. E. Aguirre et al., ALMA
survey: discovery of 13 gravitationally lensed quasars, Observations of SPT-discovered, Strongly Lensed,
Mon. Not. R. Astron. Soc. 456 (2016) 1595 Dusty, Star-forming Galaxies, Astrophys. J. 767
[1509.07917]. (2013) 132 [1303.2722].
[13] K.-H. Chae, A. D. Biggs, R. D. Blandford, I. W. Browne, [28] L. V. E. Koopmans, T. Treu, A. S. Bolton, S. Burles and
A. G. de Bruyn, C. D. Fassnacht et al., Constraints on L. A. Moustakas, The Sloan Lens ACS Survey. III.
Cosmological Parameters from the Analysis of the The Structure and Formation of Early-Type Galaxies
Cosmic Lens All Sky Survey Radio-Selected and Their Evolution since z ˜ 1, Astrophys. J. 649
Gravitational Lens Statistics, Phys. Rev. Lett. 89 (2006) 599 [astro-ph/0601628].
(2002) 151301 [astro-ph/0209602]. [29] T. Treu, M. W. Auger, L. V. E. Koopmans, R. Gavazzi,
[14] M. Oguri, N. Inada, M. A. Strauss, C. S. Kochanek, P. J. Marshall and A. S. Bolton, The Initial Mass
I. Kayo, M.-S. Shin et al., The Sloan Digital Sky Function of Early-Type Galaxies, Astrophys. J. 709
Survey Quasar Lens Search. VI. Constraints on Dark (2010) 1195 [0911.3392].
Energy and the Evolution of Massive Galaxies, [30] M. Postman, D. Coe, N. Benı́tez, L. Bradley,
Astron. J. 143 (2012) 120 [1203.1088]. T. Broadhurst, M. Donahue et al., The Cluster
[15] A. S. Bolton, S. Burles, L. V. E. Koopmans, T. Treu and Lensing and Supernova Survey with Hubble: An
L. A. Moustakas, The Sloan Lens ACS Survey. I. A Overview, Astrophys. J. Suppl. 199 (2012) 25
Large Spectroscopically Selected Sample of Massive [1106.3328].
Early-Type Lens Galaxies, Astrophys. J. 638 (2006) [31] J. M. Lotz, A. Koekemoer, D. Coe, N. Grogin, P. Capak,
703 [astro-ph/0511453]. J. Mack et al., The Frontier Fields: Survey Design
[16] A. S. Bolton, S. Burles, L. V. E. Koopmans, T. Treu, and Initial Results, Astrophys. J. 837 (2017) 97
R. Gavazzi, L. A. Moustakas et al., The Sloan Lens [1605.06567].
ACS Survey. V. The Full ACS Strong-Lens Sample, [32] D. Coe, B. Salmon, M. Bradac, L. D. Bradley, K. Sharon,
Astrophys. J. 682 (2008) 964 [0805.1931]. A. Zitrin et al., RELICS: Reionization Lensing
[17] J. F. Hennawi, M. D. Gladders, M. Oguri, N. Dalal, Cluster Survey, arXiv e-prints (2019) [1903.02002].
B. Koester, P. Natarajan et al., A New Survey for [33] C. A. Lemon, M. W. Auger, R. G. McMahon and
Giant Arcs, Astron. J. 135 (2008) 664 F. Ostrovski, Gravitationally lensed quasars in Gaia -
[astro-ph/0610061]. II. Discovery of 24 lensed quasars, Mon. Not. R.
[18] J. M. Kubo, S. S. Allam, J. Annis, E. J. Buckley-Geer, Astron. Soc. 479 (2018) 5060 [1803.07601].
H. T. Diehl, D. Kubik et al., The Sloan Bright Arcs [34] C. A. Lemon, M. W. Auger and R. G. McMahon,
Survey: Six Strongly Lensed Galaxies at z = 0.4-1.4, Gravitationally lensed quasars in Gaia - III. 22 new
Astrophys. J. Lett. 696 (2009) L61 [0812.3934]. lensed quasars from Gaia data release 2, Mon. Not. R.
[19] J. R. Brownstein, A. S. Bolton, D. J. Schlegel, D. J. Astron. Soc. 483 (2019) 4242 [1810.04480].
Eisenstein, C. S. Kochanek, N. Connolly et al., The [35] L. Delchambre, A. Krone-Martins, O. Wertz,
BOSS Emission-Line Lens Survey (BELLS). I. A C. Ducourant, L. Galluccio, J. Klüter et al., Gaia
Large Spectroscopically Selected Sample of Lens GraL: Gaia DR2 Gravitational Lens Systems. III. A
Galaxies at Redshift ˜0.5, Astrophys. J. 744 (2012) 41 systematic blind search for new lensed systems,
[1112.3683]. Astron. Astrophys. 622 (2019) A165 [1807.02845].
[20] Y. Shu, A. S. Bolton, C. S. Kochanek, M. Oguri, [36] H. T. Diehl, E. J. Buckley-Geer, K. A. Lindgren,
I. Pérez-Fournon, Z. Zheng et al., The BOSS B. Nord, H. Gaitsch, S. Gaitsch et al., The DES
Emission-line Lens Survey. III. Strong Lensing of Lyα Bright Arcs Survey: Hundreds of Candidate Strongly
Emitters by Individual Galaxies, Astrophys. J. 824 Lensed Galaxy Systems from the Dark Energy Survey
(2016) 86 [1604.01842]. Science Verification and Year 1 Observations,
[21] C. Faure, J.-P. Kneib, G. Covone, L. Tasca, Astrophys. J. Suppl. 232 (2017) 15.
A. Leauthaud, P. Capak et al., First Catalog of Strong [37] T. Treu, A. Agnello, M. A. Baumer, S. Birrer, E. J.
Lens Candidates in the COSMOS Field, Astrophys. J. Buckley-Geer, F. Courbin et al., The STRong lensing
Suppl. 176 (2008) 19 [0802.2174]. Insights into the Dark Energy Survey (STRIDES)
[22] N. Jackson, Gravitational lenses and lens candidates 2016 follow-up campaign - I. Overview and
identified from the COSMOS field, Mon. Not. R. classification of candidates selected by two techniques,
Astron. Soc. 389 (2008) 1311 [0806.3693]. Mon. Not. R. Astron. Soc. 481 (2018) 1041
[23] A. More, R. Cabanac, S. More, C. Alard, M. Limousin, [1808.04838].
Strong lensing of explosive transients 33
Structure of Cold Dark Matter Halos, Astrophys. J. quasars and supernovae in future wide-field optical
462 (1996) 563 [astro-ph/9508025]. imaging surveys, Mon. Not. R. Astron. Soc. 405
[74] J. F. Navarro, C. S. Frenk and S. D. M. White, A (2010) 2579 [1001.2037].
Universal Density Profile from Hierarchical [94] T. E. Collett, The Population of Galaxy-Galaxy Strong
Clustering, Astrophys. J. 490 (1997) 493 Lenses in Forthcoming Optical Imaging Surveys,
[astro-ph/9611107]. Astrophys. J. 811 (2015) 20 [1507.02657].
[75] M. Bartelmann, Arcs from a universal dark-matter halo [95] R. Narayan and S. D. M. White, Gravitational lensing in
profile., Astron. Astrophys. 313 (1996) 697 a cold dark matter universe, Mon. Not. R. Astron.
[astro-ph/9602053]. Soc. 231 (1988) 97p.
[76] B. Diemer and A. V. Kravtsov, A Universal Model for [96] J. Wambsganss, R. Cen, J. P. Ostriker and E. L. Turner,
Halo Concentrations, Astrophys. J. 799 (2015) 108 Testing Cosmogonic Models with Gravitational
[1407.4730]. Lensing, Science 268 (1995) 274.
[77] P. Schneider, A new formulation of gravitational lens [97] C. S. Kochanek, Gravitational Lensing Limits on Cold
theory, time-delay, and Fermat’s principle, Astron. Dark Matter and Its Variants, Astrophys. J. 453
Astrophys. 143 (1985) 413. (1995) 545 [astro-ph/9411082].
[78] R. Blandford and R. Narayan, Fermat’s principle, [98] J. S. B. Wyithe, E. L. Turner and D. N. Spergel,
caustics, and the classification of gravitational lens Gravitational Lens Statistics for Generalized NFW
images, Astrophys. J. 310 (1986) 568. Profiles: Parameter Degeneracy and Implications for
[79] C. S. Kochanek, C. R. Keeton and B. A. McLeod, The Self-Interacting Cold Dark Matter, Astrophys. J. 555
Importance of Einstein Rings, Astrophys. J. 547 (2001) 504 [astro-ph/0007354].
(2001) 50 [astro-ph/0006116]. [99] C. R. Keeton and P. Madau, Lensing Constraints on the
[80] T. Treu and L. V. E. Koopmans, The internal structure Cores of Massive Dark Matter Halos, Astrophys. J.
of the lens PG1115+080: breaking degeneracies in the Lett. 549 (2001) L25 [astro-ph/0101058].
value of the Hubble constant, Mon. Not. R. Astron. [100] R. Takahashi and T. Chiba, Gravitational Lens Statistics
Soc. 337 (2002) L6 [astro-ph/0210002]. and the Density Profile of Dark Halos, Astrophys. J.
[81] E. E. Falco, M. V. Gorenstein and I. I. Shapiro, On 563 (2001) 489 [astro-ph/0106176].
model-dependent bounds on H(0) from gravitational [101] M. Oguri, N. Inada, C. R. Keeton, B. Pindor, J. F.
images Application of Q0957 + 561A,B, Astrophys. J. Hennawi, M. D. Gregg et al., Observations and
Lett. 289 (1985) L1. Theoretical Implications of the Large-Separation
[82] O. Wucknitz, Degeneracies and scaling relations in Lensed Quasar SDSS J1004+4112, Astrophys. J. 605
general power-law models for gravitational lenses, (2004) 78 [astro-ph/0312429].
Mon. Not. R. Astron. Soc. 332 (2002) 951 [102] M. Oguri and C. R. Keeton, Effects of Triaxiality on the
[astro-ph/0202376]. Statistics of Large-Separation Gravitational Lenses,
[83] P. Schneider and D. Sluse, Source-position Astrophys. J. 610 (2004) 663 [astro-ph/0403633].
transformation: an approximate invariance in strong [103] S. Hilbert, S. D. M. White, J. Hartlap and P. Schneider,
gravitational lensing, Astron. Astrophys. 564 (2014) Strong lensing optical depths in a ΛCDM universe,
A103 [1306.4675]. Mon. Not. R. Astron. Soc. 382 (2007) 121
[84] S. Unruh, P. Schneider and D. Sluse, Ambiguities in [astro-ph/0703803].
gravitational lens models: the density field from the [104] G. L. Li, S. Mao, Y. P. Jing, W. P. Lin and M. Oguri,
source position transformation, Astron. Astrophys. Properties of wide-separation lensed quasars by
601 (2017) A77 [1606.04321]. clusters of galaxies in the Sloan Digital Sky Survey,
[85] E. L. Turner, J. P. Ostriker and J. R. Gott, III, The Mon. Not. R. Astron. Soc. 378 (2007) 469
statistics of gravitational lenses - The distributions of [astro-ph/0701801].
image angular separations and lens redshifts, [105] Q. E. Minor and M. Kaplinghat, Effects of galaxy-halo
Astrophys. J. 284 (1984) 1. alignment and adiabatic contraction on gravitational
[86] M. Fukugita, T. Futamase, M. Kasai and E. L. Turner, lens statistics, Mon. Not. R. Astron. Soc. 391 (2008)
Statistical properties of gravitational lenses with a 653 [0711.2537].
nonzero cosmological constant, Astrophys. J. 393 [106] T. J. Broadhurst and R. Barkana, Large Einstein radii: a
(1992) 3. problem for ΛCDM, Mon. Not. R. Astron. Soc. 390
[87] C. S. Kochanek, Is There a Cosmological Constant?, (2008) 1647 [0801.1875].
Astrophys. J. 466 (1996) 638 [astro-ph/9510077]. [107] M. Oguri and R. D. Blandford, What is the largest
[88] M. Chiba and Y. Yoshii, New Limits on a Cosmological Einstein radius in the universe?, Mon. Not. R.
Constant from Statistics of Gravitational Lensing, Astron. Soc. 392 (2009) 930 [0808.0192].
Astrophys. J. 510 (1999) 42 [astro-ph/9808321]. [108] A. Zitrin, T. Broadhurst, M. Bartelmann, Y. Rephaeli,
[89] K.-H. Chae, The Cosmic Lens All-Sky Survey: statistical M. Oguri, N. Benı́tez et al., The universal Einstein
strong lensing, cosmological parameters, and global radius distribution from 10 000 SDSS clusters, Mon.
properties of galaxy populations, Mon. Not. R. Astron. Not. R. Astron. Soc. 423 (2012) 2308 [1105.2295].
Soc. 346 (2003) 746 [astro-ph/0211244]. [109] M. Redlich, M. Bartelmann, J.-C. Waizmann and
[90] J. L. Mitchell, C. R. Keeton, J. A. Frieman and R. K. C. Fedeli, The strongest gravitational lenses. I. The
Sheth, Improved Cosmological Constraints from statistical impact of cluster mergers, Astron.
Gravitational Lens Statistics, Astrophys. J. 622 Astrophys. 547 (2012) A66 [1205.6906].
(2005) 81 [astro-ph/0401138]. [110] J.-C. Waizmann, M. Redlich, M. Meneghetti and
[91] D. Huterer, C. R. Keeton and C.-P. Ma, Effects of M. Bartelmann, The strongest gravitational lenses.
Ellipticity and Shear on Gravitational Lens Statistics, III. The order statistics of the largest Einstein radii,
Astrophys. J. 624 (2005) 34 [astro-ph/0405040]. Astron. Astrophys. 565 (2014) A28 [1403.4573].
[92] M. Oguri, C. R. Keeton and N. Dalal, The impact of lens [111] C. R. Keeton, II, Using Gravitational Lenses to Study
galaxy environments on the image separation Galaxies and Cosmology, Ph.D. thesis, HARVARD
distribution, Mon. Not. R. Astron. Soc. 364 (2005) UNIVERSITY, 1998.
1451 [astro-ph/0506549]. [112] C. Porciani and P. Madau, Gravitational Lensing of
[93] M. Oguri and P. J. Marshall, Gravitationally lensed Distant Supernovae in Cold Dark Matter Universes,
Strong lensing of explosive transients 35
Astrophys. J. 532 (2000) 679. [132] S. Sugiyama, T. Kurita and M. Takada, Revisiting the
[113] C. S. Kochanek and M. White, Global Probes of the wave optics effect on primordial black hole constraints
Impact of Baryons on Dark Matter Halos, Astrophys. from optical microlensing search, arXiv e-prints
J. 559 (2001) 531 [astro-ph/0102334]. (2019) [1905.06066].
[114] L.-X. Li and J. P. Ostriker, Semianalytical Models for [133] P. Montero-Camacho, X. Fang, G. Vasquez, M. Silva and
Lensing by Dark Halos. I. Splitting Angles, Astrophys. C. M. Hirata, Revisiting constraints on asteroid-mass
J. 566 (2002) 652 [astro-ph/0010432]. primordial black holes as dark matter candidates,
[115] M. Oguri, Constraints on the Baryonic Compression and arXiv e-prints (2019) arXiv:1906.05950 [1906.05950].
Implications for the Fraction of Dark Halo Lenses, [134] T. T. Nakamura and S. Deguchi, Wave Optics in
Astrophys. J. 580 (2002) 2 [astro-ph/0207520]. Gravitational Lensing, Progress of Theoretical Physics
[116] C.-P. Ma, Schechter versus Schechter: Subarcsecond Supplement 133 (1999) 137.
Gravitational Lensing and Inner Halo Profiles, [135] S. Deguchi and W. D. Watson, Diffraction in
Astrophys. J. Lett. 584 (2003) L1 [astro-ph/0211464]. gravitational lensing for compact objects of low mass,
[117] D. Huterer and C.-P. Ma, Constraints on the Inner Astrophys. J. 307 (1986) 30.
Cluster Mass Profile and the Power Spectrum [136] S. Deguchi and W. D. Watson, Wave effects in
Normalization from Strong Lensing Statistics, gravitational lensing of electromagnetic radiation,
Astrophys. J. Lett. 600 (2004) L7 [astro-ph/0307301]. Phys. Rev. D 34 (1986) 1708.
[118] M. Kuhlen, C. R. Keeton and P. Madau, Gravitational [137] N. Matsunaga and K. Yamamoto, The finite source size
Lensing Statistics in Universes Dominated by Dark effect and wave optics in gravitational lensing, J.
Energy, Astrophys. J. 601 (2004) 104 Cosmol. Astropart. Phys. 1 (2006) 023
[astro-ph/0310013]. [astro-ph/0601701].
[119] C. R. Keeton, M. Kuhlen and Z. Haiman, Gravitational [138] A. V. Filippenko, Optical Spectra of Supernovae, Ann.
Lensing Magnification without Multiple Imaging, Rev. Astron. Astrophys. 35 (1997) 309.
Astrophys. J. 621 (2005) 559 [astro-ph/0405143]. [139] D. Branch and J. C. Wheeler, Supernova Explosions.
[120] M. Oguri, The image separation distribution of strong 2017, 10.1007/978-3-662-55054-0.
lenses: halo versus subhalo populations, Mon. Not. R. [140] D. Maoz, F. Mannucci and G. Nelemans, Observational
Astron. Soc. 367 (2006) 1241 [astro-ph/0508528]. Clues to the Progenitors of Type Ia Supernovae, Ann.
[121] A. Amvrosiadis, S. A. Eales, M. Negrello, L. Marchetti, Rev. Astron. Astrophys. 52 (2014) 107 [1312.0628].
M. W. L. Smith, N. Bourne et al., ALMA observations [141] B. Müller, The Status of Multi-Dimensional
of lensed Herschel sources: testing the dark matter Core-Collapse Supernova Models, Publ. Astron. Soc.
halo paradigm, Mon. Not. R. Astron. Soc. 475 (2018) Aust. 33 (2016) e048 [1608.03274].
4939 [1801.07282]. [142] D. Richardson, D. Branch, D. Casebeer, J. Millard, R. C.
[122] E. L. Turner, The effect of undetected gravitational Thomas and E. Baron, A Comparative Study of the
lenses on statistical measures of quasar evolution, Absolute Magnitude Distributions of Supernovae,
Astrophys. J. Lett. 242 (1980) L135. Astron. J. 123 (2002) 745 [astro-ph/0112051].
[123] M. Oguri, Effect of gravitational lensing on the [143] M. M. Phillips, The absolute magnitudes of Type IA
distribution of gravitational waves from distant binary supernovae, Astrophys. J. Lett. 413 (1993) L105.
black hole mergers, Mon. Not. R. Astron. Soc. 480 [144] A. G. Riess, A. V. Filippenko, P. Challis, A. Clocchiatti,
(2018) 3842 [1807.02584]. A. Diercks, P. M. Garnavich et al., Observational
[124] M. Bernardi, F. Shankar, J. B. Hyde, S. Mei, F. Marulli Evidence from Supernovae for an Accelerating
and R. K. Sheth, Galaxy luminosities, stellar masses, Universe and a Cosmological Constant, Astron. J.
sizes, velocity dispersions as a function of 116 (1998) 1009 [astro-ph/9805201].
morphological type, Mon. Not. R. Astron. Soc. 404 [145] S. Perlmutter, G. Aldering, G. Goldhaber, R. A. Knop,
(2010) 2087 [0910.1093]. P. Nugent, P. G. Castro et al., Measurements of Ω and
[125] P. Torrey, S. Wellons, F. Machado, B. Griffen, D. Nelson, Λ from 42 High-Redshift Supernovae, Astrophys. J.
V. Rodriguez-Gomez et al., An analysis of the 517 (1999) 565 [astro-ph/9812133].
evolving comoving number density of galaxies in [146] M. Sullivan, J. Guy, A. Conley, N. Regnault, P. Astier,
hydrodynamical simulations, Mon. Not. R. Astron. C. Balland et al., SNLS3: Constraints on Dark Energy
Soc. 454 (2015) 2770 [1507.01942]. Combining the Supernova Legacy Survey Three-year
[126] M. Oguri, Y. Suto and E. L. Turner, Gravitational Data with Other Probes, Astrophys. J. 737 (2011) 102
Lensing Magnification and Time Delay Statistics for [1104.1444].
Distant Supernovae, Astrophys. J. 583 (2003) 584 [147] M. Sako, B. Bassett, A. C. Becker, P. J. Brown,
[astro-ph/0210107]. H. Campbell, R. Wolf et al., The Data Release of the
[127] A. Gould, Femtolensing of gamma-ray bursters, Sloan Digital Sky Survey-II Supernova Survey, Publ.
Astrophys. J. Lett. 386 (1992) L5. Astron. Soc. Pac. 130 (2018) 064002 [1401.3317].
[128] T. T. Nakamura, Gravitational Lensing of Gravitational [148] N. M. Law, S. R. Kulkarni, R. G. Dekany, E. O. Ofek,
Waves from Inspiraling Binaries by a Point Mass R. M. Quimby, P. E. Nugent et al., The Palomar
Lens, Phys. Rev. Lett. 80 (1998) 1138. Transient Factory: System Overview, Performance,
[129] R. Takahashi and T. Nakamura, Wave Effects in the and First Results, Publ. Astron. Soc. Pac. 121 (2009)
Gravitational Lensing of Gravitational Waves from 1395 [0906.5350].
Chirping Binaries, Astrophys. J. 595 (2003) 1039 [149] N. Suzuki, D. Rubin, C. Lidman, G. Aldering,
[astro-ph/0305055]. R. Amanullah, K. Barbary et al., The Hubble Space
[130] A. Mehrabi and S. Rahvar, Studying wave optics in the Telescope Cluster Supernova Survey. V. Improving the
light curves of exoplanet microlensing, Mon. Not. R. Dark-energy Constraints above z & 1 and Building an
Astron. Soc. 431 (2013) 1264 [1207.4034]. Early-type-hosted Supernova Sample, Astrophys. J.
[131] H. Niikura, M. Takada, N. Yasuda, R. H. Lupton, 746 (2012) 85 [1105.3470].
T. Sumi, S. More et al., Microlensing constraints on [150] D. M. Scolnic, D. O. Jones, A. Rest, Y. C. Pan,
primordial black holes with Subaru/HSC Andromeda R. Chornock, R. J. Foley et al., The Complete
observations, Nature Astronomy 3 (2019) 524 Light-curve Sample of Spectroscopically Confirmed
[1701.02151]. SNe Ia from Pan-STARRS1 and Cosmological
Strong lensing of explosive transients 36
Constraints from the Combined Pantheon Sample, [168] W. B. Atwood, A. A. Abdo, M. Ackermann,
Astrophys. J. 859 (2018) 101 [1710.00845]. W. Althouse, B. Anderson, M. Axelsson et al., The
[151] T. W.-S. Holoien, J. S. Brown, K. Z. Stanek, C. S. Large Area Telescope on the Fermi Gamma-Ray Space
Kochanek, B. J. Shappee, J. L. Prieto et al., The Telescope Mission, Astrophys. J. 697 (2009) 1071
ASAS-SN bright supernova catalogue - III. 2016, Mon. [0902.1089].
Not. R. Astron. Soc. 471 (2017) 4966 [1704.02320]. [169] C. Meegan, G. Lichti, P. N. Bhat, E. Bissaldi, M. S.
[152] D. Brout, D. Scolnic, R. Kessler, C. B. D’Andrea, T. M. Briggs, V. Connaughton et al., The Fermi Gamma-ray
Davis, R. R. Gupta et al., First Cosmology Results Burst Monitor, Astrophys. J. 702 (2009) 791
Using SNe Ia from the Dark Energy Survey: Analysis, [0908.0450].
Systematic Uncertainties, and Validation, Astrophys. [170] T. Piran, The physics of gamma-ray bursts, Reviews of
J. 874 (2019) 150 [1811.02377]. Modern Physics 76 (2004) 1143 [astro-ph/0405503].
[153] N. Yasuda, M. Tanaka, N. Tominaga, J.-a. Jiang, T. J. [171] T. Piran, Gamma-ray bursts and the fireball model, Phys.
Moriya, T. Morokuma et al., The Hyper Suprime-Cam Rep. 314 (1999) 575 [astro-ph/9810256].
SSP Transient Survey in COSMOS: Overview, arXiv [172] E. Berger, Short-Duration Gamma-Ray Bursts, Ann.
e-prints (2019) [1904.09697]. Rev. Astron. Astrophys. 52 (2014) 43 [1311.2603].
[154] J. Guillochon, J. Parrent, L. Z. Kelley and R. Margutti, [173] C. Kouveliotou, C. A. Meegan, G. J. Fishman, N. P.
An Open Catalog for Supernova Data, Astrophys. J. Bhat, M. S. Briggs, T. M. Koshut et al., Identification
835 (2017) 64 [1605.01054]. of two classes of gamma-ray bursts, Astrophys. J.
[155] A. Gal-Yam, Luminous Supernovae, Science 337 (2012) Lett. 413 (1993) L101.
927 [1208.3217]. [174] S. Savaglio, K. Glazebrook and D. Le Borgne, The
[156] T. J. Moriya, E. I. Sorokina and R. A. Chevalier, Galaxy Population Hosting Gamma-Ray Bursts,
Superluminous Supernovae, Space Sci. Rev. 214 Astrophys. J. 691 (2009) 182 [0803.2718].
(2018) 59 [1803.01875]. [175] T. J. Galama, P. M. Vreeswijk, J. van Paradijs,
[157] R. M. Quimby, G. Aldering, J. C. Wheeler, P. Höflich, C. Kouveliotou, T. Augusteijn, H. Böhnhardt et al.,
C. W. Akerlof and E. S. Rykoff, SN 2005ap: A Most An unusual supernova in the error box of the γ-ray
Brilliant Explosion, Astrophys. J. Lett. 668 (2007) burst of 25 April 1998, Nature 395 (1998) 670
L99 [0709.0302]. [astro-ph/9806175].
[158] J. Cooke, M. Sullivan, A. Gal-Yam, E. J. Barton, R. G. [176] J. S. Bloom, S. R. Kulkarni, S. G. Djorgovski, A. C.
Carlberg, E. V. Ryan-Weber et al., Superluminous Eichelberger, P. Côté, J. P. Blakeslee et al., The
supernovae at redshifts of 2.05 and 3.90, Nature 491 unusual afterglow of the γ-ray burst of 26 March 1998
(2012) 228 [1211.2003]. as evidence for a supernova connection, Nature 401
[159] T. J. Moriya, M. Tanaka, N. Yasuda, J.-a. Jiang, C.-H. (1999) 453 [astro-ph/9905301].
Lee, K. Maeda et al., First Release of High-Redshift [177] S. Nagataki, Theories of central engine for long
Superluminous Supernovae from the Subaru HIgh-Z gamma-ray bursts, Reports on Progress in Physics 81
SUpernova CAmpaign (SHIZUCA). I. Photometric (2018) 026901.
Properties, Astrophys. J. Suppl. 241 (2019) 16 [178] P. Jakobsson, A. Levan, J. P. U. Fynbo, R. Priddey,
[1801.08240]. J. Hjorth, N. Tanvir et al., A mean redshift of 2.8 for
[160] W. Li, R. Chornock, J. Leaman, A. V. Filippenko, Swift gamma-ray bursts, Astron. Astrophys. 447
D. Poznanski, X. Wang et al., Nearby supernova rates (2006) 897 [astro-ph/0509888].
from the Lick Observatory Supernova Search - III. [179] A. Cucchiara, A. J. Levan, D. B. Fox, N. R. Tanvir,
The rate-size relation, and the rates as a function of T. N. Ukwatta, E. Berger et al., A Photometric
galaxy Hubble type and colour, Mon. Not. R. Astron. Redshift of z ˜ 9.4 for GRB 090429B, Astrophys. J.
Soc. 412 (2011) 1473 [1006.4613]. 736 (2011) 7 [1105.4915].
[161] R. M. Quimby, F. Yuan, C. Akerlof and J. C. Wheeler, [180] D. Wanderman and T. Piran, The luminosity function
Rates of superluminous supernovae at z sim 0.2, Mon. and the rate of Swift’s gamma-ray bursts, Mon. Not.
Not. R. Astron. Soc. 431 (2013) 912 [1302.0911]. R. Astron. Soc. 406 (2010) 1944 [0912.0709].
[162] L.-D. Liu, B. Zhang, L.-J. Wang and Z.-G. Dai, [181] E. Berger, P. A. Price, S. B. Cenko, A. Gal-Yam, A. M.
Photospheric Radius Evolution of Homologous Soderberg, M. Kasliwal et al., The afterglow and
Explosions, Astrophys. J. 868 (2018) L24 elliptical host galaxy of the short γ-ray burst GRB
[1809.05048]. 050724, Nature 438 (2005) 988 [astro-ph/0508115].
[163] R. W. Klebesadel, I. B. Strong and R. A. Olson, [182] D. B. Fox, D. A. Frail, P. A. Price, S. R. Kulkarni,
Observations of Gamma-Ray Bursts of Cosmic E. Berger, T. Piran et al., The afterglow of GRB
Origin, Astrophys. J. Lett. 182 (1973) L85. 050709 and the nature of the short-hard γ-ray bursts,
[164] C. A. Meegan, G. J. Fishman, R. B. Wilson, W. S. Nature 437 (2005) 845 [astro-ph/0510110].
Paciesas, G. N. Pendleton, J. M. Horack et al., Spatial [183] D. Wanderman and T. Piran, The rate, luminosity
distribution of gamma-ray bursts observed by BATSE, function and time delay of non-Collapsar short GRBs,
Nature 355 (1992) 143. Mon. Not. R. Astron. Soc. 448 (2015) 3026
[165] E. Costa, F. Frontera, J. Heise, M. Feroci, J. in’t Zand, [1405.5878].
F. Fiore et al., Discovery of an X-ray afterglow [184] F. Y. Wang, Z. G. Dai and E. W. Liang, Gamma-ray
associated with the γ-ray burst of 28 February 1997, burst cosmology, New Astron. Rev. 67 (2015) 1
Nature 387 (1997) 783 [astro-ph/9706065]. [1504.00735].
[166] S. R. Kulkarni, S. G. Djorgovski, A. N. Ramaprakash, [185] E. E. Fenimore and E. Ramirez-Ruiz, Redshifts For 220
R. Goodrich, J. S. Bloom, K. L. Adelberger et al., BATSE Gamma-Ray Bursts Determined by
Identification of a host galaxy at redshift z = 3.42 for Variability and the Cosmological Consequences, arXiv
the γ-ray burst of 14 December 1997, Nature 393 Astrophysics e-prints (2000) [astro-ph/0004176].
(1998) 35. [186] L. Amati, F. Frontera, M. Tavani, J. J. M. in’t Zand ,
[167] N. Gehrels, G. Chincarini, P. Giommi, K. O. Mason, A. Antonelli, E. Costa et al., Intrinsic spectra and
J. A. Nousek, A. A. Wells et al., The Swift energetics of BeppoSAX Gamma-Ray Bursts with
Gamma-Ray Burst Mission, Astrophys. J. 611 (2004) known redshifts, Astron. Astrophys. 390 (2002) 81
1005 [astro-ph/0405233]. [astro-ph/0205230].
Strong lensing of explosive transients 37
[187] D. Yonetoku, T. Murakami, T. Nakamura, R. Yamazaki, Macquart, J. X. Prochaska, N. Tejos et al., A single
A. K. Inoue and K. Ioka, Gamma-Ray Burst fast radio burst localized to a massive galaxy at
Formation Rate Inferred from the Spectral Peak cosmological distance, arXiv e-prints (2019)
Energy-Peak Luminosity Relation, Astrophys. J. 609 arXiv:1906.11476 [1906.11476].
(2004) 935 [astro-ph/0309217]. [205] V. Ravi, M. Catha, L. D’Addario, S. G. Djorgovski,
[188] G. Ghirlanda, G. Ghisellini and D. Lazzati, The G. Hallinan, R. Hobbs et al., A fast radio burst
Collimation-corrected Gamma-Ray Burst Energies localised to a massive galaxy, arXiv e-prints (2019)
Correlate with the Peak Energy of Their νFν arXiv:1907.01542 [1907.01542].
Spectrum, Astrophys. J. 616 (2004) 331 [206] W. Lu and A. L. Piro, Implications from ASKAP Fast
[astro-ph/0405602]. Radio Burst Statistics, arXiv e-prints (2019)
[189] A. Barnacka and A. Loeb, A Size-duration Trend for arXiv:1903.00014 [1903.00014].
Gamma-Ray Burst Progenitors, Astrophys. J. 794 [207] V. Ravi, The prevalence of repeating fast radio bursts,
(2014) L8 [1409.1232]. arXiv e-prints (2019) arXiv:1907.06619 [1907.06619].
[190] V. Z. Golkhou, N. R. Butler and O. M. Littlejohns, The [208] B. Marcote, Z. Paragi, J. W. T. Hessels, A. Keimpema,
Energy Dependence of GRB Minimum Variability H. J. van Langevelde, Y. Huang et al., The Repeating
Timescales, Astrophys. J. 811 (2015) 93 [1501.05948]. Fast Radio Burst FRB 121102 as Seen on
[191] A. Katz, J. Kopp, S. Sibiryakov and W. Xue, Milliarcsecond Angular Scales, Astrophys. J. 834
Femtolensing by dark matter revisited, Journal of (2017) L8 [1701.01099].
Cosmology and Astro-Particle Physics 2018 (2018) [209] B. P. Abbott, R. Abbott, T. D. Abbott, M. R.
005 [1807.11495]. Abernathy, F. Acernese, K. Ackley et al., Observation
[192] E. Petroff, J. W. T. Hessels and D. R. Lorimer, Fast of Gravitational Waves from a Binary Black Hole
radio bursts, Astron. Astrophys. Rev. 27 (2019) 4 Merger, Phys. Rev. Lett. 116 (2016) 061102
[1904.07947]. [1602.03837].
[193] J. M. Cordes and S. Chatterjee, Fast Radio Bursts: An [210] LIGO Scientific Collaboration, J. Aasi, B. P. Abbott,
Extragalactic Enigma, arXiv e-prints (2019) R. Abbott, T. Abbott, M. R. Abernathy et al.,
arXiv:1906.05878 [1906.05878]. Advanced LIGO, Classical and Quantum Gravity 32
[194] D. R. Lorimer, M. Bailes, M. A. McLaughlin, D. J. (2015) 074001 [1411.4547].
Narkevic and F. Crawford, A Bright Millisecond Radio [211] M. Maggiore, Gravitational Waves: Volume 1: Theory
Burst of Extragalactic Origin, Science 318 (2007) 777 and Experiments. 2008,
[0709.4301]. 10.1093/acprof:oso/9780198570745.001.0001.
[195] D. Thornton, B. Stappers, M. Bailes, B. Barsdell, [212] M. Maggiore, Gravitational Waves: Volume 2:
S. Bates, N. D. R. Bhat et al., A Population of Fast Astrophysics and Cosmology. 2018,
Radio Bursts at Cosmological Distances, Science 341 10.1093/oso/9780198570899.001.0001.
(2013) 53 [1307.1628]. [213] The LIGO Scientific Collaboration, the Virgo
[196] CHIME/FRB Collaboration, M. Amiri, K. Bandura, Collaboration, B. P. Abbott, R. Abbott, T. D. Abbott,
P. Berger, M. Bhardwaj, M. M. Boyce et al., The S. Abraham et al., GWTC-1: A Gravitational-Wave
CHIME Fast Radio Burst Project: System Overview, Transient Catalog of Compact Binary Mergers
Astrophys. J. 863 (2018) 48 [1803.11235]. Observed by LIGO and Virgo during the First and
[197] A. W. Hotan, J. D. Bunton, L. Harvey-Smith, Second Observing Runs, arXiv e-prints (2018)
B. Humphreys, B. D. Jeffs, T. Shimwell et al., The [1811.12907].
Australian Square Kilometre Array Pathfinder: [214] B. P. Abbott, R. Abbott, T. D. Abbott, F. Acernese,
System Architecture and Specifications of the Boolardy K. Ackley, C. Adams et al., GW170817: Observation
Engineering Test Array, Publ. Astron. Soc. Aust. 31 of Gravitational Waves from a Binary Neutron Star
(2014) e041 [1409.1325]. Inspiral, Phys. Rev. Lett. 119 (2017) 161101
[198] R. M. Shannon, J. P. Macquart, K. W. Bannister, R. D. [1710.05832].
Ekers, C. W. James, S. Oslowski et al., The [215] B. P. Abbott, R. Abbott, T. D. Abbott, F. Acernese,
dispersion-brightness relation for fast radio bursts K. Ackley, C. Adams et al., GW170814: A
from a wide-field survey, Nature 562 (2018) 386. Three-Detector Observation of Gravitational Waves
[199] CHIME/FRB Collaboration, M. Amiri, K. Bandura, from a Binary Black Hole Coalescence, Phys. Rev.
M. Bhardwaj, P. Boubel, M. M. Boyce et al., Lett. 119 (2017) 141101 [1709.09660].
Observations of fast radio bursts at frequencies down [216] F. Acernese, M. Agathos, K. Agatsuma, D. Aisa,
to 400 megahertz, Nature 566 (2019) 230 N. Allemandou, A. Allocca et al., Advanced Virgo: a
[1901.04524]. second-generation interferometric gravitational wave
[200] L. G. Spitler, J. M. Cordes, J. W. T. Hessels, D. R. detector, Classical and Quantum Gravity 32 (2015)
Lorimer, M. A. McLaughlin, S. Chatterjee et al., Fast 024001 [1408.3978].
Radio Burst Discovered in the Arecibo Pulsar ALFA [217] Y. Aso, Y. Michimura, K. Somiya, M. Ando,
Survey, Astrophys. J. 790 (2014) 101 [1404.2934]. O. Miyakawa, T. Sekiguchi et al., Interferometer
[201] L. G. Spitler, P. Scholz, J. W. T. Hessels, S. Bogdanov, design of the KAGRA gravitational wave detector,
A. Brazier, F. Camilo et al., A repeating fast radio Phys. Rev. D 88 (2013) 043007 [1306.6747].
burst, Nature 531 (2016) 202 [1603.00581]. [218] B. P. Abbott, R. Abbott, T. D. Abbott, F. Acernese,
[202] CHIME/FRB Collaboration, M. Amiri, K. Bandura, K. Ackley, C. Adams et al., Gravitational Waves and
M. Bhardwaj, P. Boubel, M. M. Boyce et al., A second Gamma-Rays from a Binary Neutron Star Merger:
source of repeating fast radio bursts, Nature 566 GW170817 and GRB 170817A, Astrophys. J. Lett.
(2019) 235 [1901.04525]. 848 (2017) L13 [1710.05834].
[203] S. P. Tendulkar, C. G. Bassa, J. M. Cordes, G. C. Bower, [219] B. P. Abbott, R. Abbott, T. D. Abbott, F. Acernese,
C. J. Law, S. Chatterjee et al., The Host Galaxy and K. Ackley, C. Adams et al., Multi-messenger
Redshift of the Repeating Fast Radio Burst FRB Observations of a Binary Neutron Star Merger,
121102, Astrophys. J. Lett. 834 (2017) L7 Astrophys. J. Lett. 848 (2017) L12 [1710.05833].
[1701.01100]. [220] B. F. Schutz, Determining the Hubble constant from
[204] K. W. Bannister, A. T. Deller, C. Phillips, J. P. gravitational wave observations, Nature 323 (1986)
Strong lensing of explosive transients 38
for distant supernovae: breaking the degeneracy Gamma-Ray Bursts, Astrophys. J. 317 (1987) L51.
between radial mass profiles and the Hubble constant, [273] O. M. Blaes and R. L. Webster, Using Gamma-Ray
Mon. Not. R. Astron. Soc. 338 (2003) L25 Bursts to Detect a Cosmological Density of Compact
[astro-ph/0211499]. Objects, Astrophys. J. 391 (1992) L63.
[256] A. S. Bolton and S. Burles, Prospects for the [274] S. Mao, Gravitational Lensing, Time Delay, and
Determination of H0 through Observation of Multiply Gamma-Ray Bursts, Astrophys. J. 389 (1992) L41.
Imaged Supernovae in Galaxy Cluster Fields, [275] S. Mao, Gravitational microlensing of gamma-ray bursts,
Astrophys. J. 592 (2003) 17 [astro-ph/0212181]. Astrophys. J. 402 (1993) 382.
[257] T. Riehm, E. Mörtsell, A. Goobar, R. Amanullah, [276] R. J. Nemiroff and A. Gould, Probing for MACHOs of
T. Dahlén, J. Jönsson et al., Near-IR search for lensed Mass 10 -15 Msun to 10 -7 Msun with Gamma-Ray
supernovae behind galaxy clusters. III. Implications Burst Parallax Spacecraft, Astrophys. J. 452 (1995)
for cluster modeling and cosmology, Astron. L111 [astro-ph/9505019].
Astrophys. 536 (2011) A94 [1109.6351]. [277] A. Loeb and R. Perna, Microlensing of Gamma-Ray
[258] A. Zitrin, M. Redlich and T. Broadhurst, Consistent Use Burst Afterglows, Astrophys. J. 495 (1998) 597
of Type Ia Supernovae Highly Magnified by Galaxy [astro-ph/9708159].
Clusters to Constrain the Cosmological Parameters, [278] G. F. Marani, R. J. Nemiroff, J. P. Norris, K. Hurley and
Astrophys. J. 789 (2014) 51 [1311.5224]. J. T. Bonnell, Gravitationally Lensed Gamma-Ray
[259] X. Wen and K. Liao, Calibrating the standard candles Bursts as Probes of Dark Compact Objects, Astrophys.
with strong lensing, arXiv e-prints (2019) J. 512 (1999) L13 [astro-ph/9810391].
arXiv:1907.02693 [1907.02693]. [279] J. S. B. Wyithe and E. L. Turner, Gravitational
[260] M. Foxley-Marrable, T. E. Collett, G. Vernardos, D. A. microlensing of gamma-ray bursts at medium optical
Goldstein and D. Bacon, The impact of microlensing depth, Mon. Not. R. Astron. Soc. 319 (2000) 1163
on the standardization of strongly lensed Type Ia [astro-ph/0008009].
supernovae, Mon. Not. R. Astron. Soc. 478 (2018) [280] L. V. E. Koopmans and J. Wambsganss, On the
5081 [1802.07738]. probability of microlensing in gamma-ray burst
[261] L. M. Krauss and S. Tremaine, Test of the weak afterglows, Mon. Not. R. Astron. Soc. 325 (2001)
equivalence principle for neutrinos and photons, Phys. 1317 [astro-ph/0011029].
Rev. Lett. 60 (1988) 176. [281] E. A. Baltz and L. Hui, Microlensing of Gamma-Ray
[262] A. A. Abdo, M. Ackermann, M. Ajello, K. Asano, W. B. Bursts by Stars and MACHOs, Astrophys. J. 618
Atwood, M. Axelsson et al., A limit on the variation (2005) 403 [astro-ph/0311569].
of the speed of light arising from quantum gravity [282] L. Ji, E. D. Kovetz and M. Kamionkowski, Strong lensing
effects, Nature 462 (2009) 331 [0908.1832]. of gamma ray bursts as a probe of compact dark
[263] J.-J. Wei, H. Gao, X.-F. Wu and P. Mészáros, Testing matter, Phys. Rev. D 98 (2018) 123523 [1809.09627].
Einstein’s Equivalence Principle With Fast Radio [283] K. Z. Stanek, B. Paczynski and J. Goodman, Features in
Bursts, Phys. Rev. Lett. 115 (2015) 261101 the spectra of gamma-ray bursts, Astrophys. J. Lett.
[1512.07670]. 413 (1993) L7.
[264] M. Biesiada and A. Piórkowska, Gravitational lensing [284] A. Ulmer and J. Goodman, Femtolensing: Beyond the
time delays as a tool for testing Lorentz-invariance Semiclassical Approximation, Astrophys. J. 442
violation, Mon. Not. R. Astron. Soc. 396 (2009) 946 (1995) 67 [astro-ph/9406042].
[0712.0941]. [285] A. Barnacka, J. F. Glicenstein and R. Moderski, New
[265] T. E. Collett and D. Bacon, Testing the Speed of constraints on primordial black holes abundance from
Gravitational Waves over Cosmological Distances with femtolensing of gamma-ray bursts, Phys. Rev. D 86
Strong Gravitational Lensing, Phys. Rev. Lett. 118 (2012) 043001 [1204.2056].
(2017) 091101 [1602.05882]. [286] J. B. Muñoz, E. D. Kovetz, L. Dai and
[266] X.-L. Fan, K. Liao, M. Biesiada, A. Piórkowska-Kurpas M. Kamionkowski, Lensing of Fast Radio Bursts as a
and Z.-H. Zhu, Speed of Gravitational Waves from Probe of Compact Dark Matter, Phys. Rev. Lett. 117
Strongly Lensed Gravitational Waves and (2016) 091301 [1605.00008].
Electromagnetic Signals, Phys. Rev. Lett. 118 (2017) [287] Y. K. Wang and F. Y. Wang, Lensing of fast radio bursts
091102 [1612.04095]. by binaries to probe compact dark matter, Astron.
[267] T. Baker and M. Trodden, Multimessenger time delays Astrophys. 614 (2018) A50 [1801.07360].
from lensed gravitational waves, Phys. Rev. D 95 [288] R. Laha, Lensing of fast radio bursts: future constraints
(2017) 063512 [1612.02004]. on primordial black hole density with an extended
[268] H. Yu and F. Y. Wang, Testing weak equivalence mass function and a new probe of exotic compact
principle with strongly lensed cosmic transients, fermion/ boson stars, arXiv e-prints (2018)
European Physical Journal C 78 (2018) 692 arXiv:1812.11810 [1812.11810].
[1801.01257]. [289] Z. Zheng, E. O. Ofek, S. R. Kulkarni, J. D. Neill and
[269] R. Takahashi, Arrival Time Differences between M. Juric, Probing the Intergalactic Medium with Fast
Gravitational Waves and Electromagnetic Signals due Radio Bursts, Astrophys. J. 797 (2014) 71
to Gravitational Lensing, Astrophys. J. 835 (2017) [1409.3244].
103 [1606.00458]. [290] K.-H. Lai, O. A. Hannuksela, A. Herrera-Martı́n, J. M.
[270] S. Cao, J. Qi, M. Biesiada, X. Zheng, T. Xu and Z.-H. Diego, T. Broadhurst and T. G. F. Li, Discovering
Zhu, Testing the Speed of Light over Cosmological intermediate-mass black hole lenses through
Distances: The Combination of Strongly Lensed and gravitational wave lensing, Phys. Rev. D 98 (2018)
Unlensed Type Ia Supernovae, Astrophys. J. 867 083005 [1801.07840].
(2018) 50 [1810.01287]. [291] P. Christian, S. Vitale and A. Loeb, Detecting stellar
[271] M. Sasaki, T. Suyama, T. Tanaka and S. Yokoyama, lensing of gravitational waves with ground-based
Primordial black holes—perspectives in gravitational observatories, Phys. Rev. D 98 (2018) 103022
wave astronomy, Classical and Quantum Gravity 35 [1802.02586].
(2018) 063001 [1801.05235]. [292] S. Jung and C. S. Shin, Gravitational-Wave Fringes at
[272] B. Paczynski, Gravitational Microlensing and LIGO: Detecting Compact Dark Matter by
Strong lensing of explosive transients 40
Gravitational Lensing, Phys. Rev. Lett. 122 (2019) (2013) arXiv:1312.6330 [1312.6330].
041103 [1712.01396]. [311] T. Petrushevska, R. Amanullah, A. Goobar, S. Fabbro,
[293] K. Liao, M. Biesiada and X.-L. Fan, The Wave Nature of J. Johansson, T. Kjellsson et al., High-redshift
Continuous Gravitational Waves from Microlensing, supernova rates measured with the gravitational
Astrophys. J. 875 (2019) 139 [1903.06612]. telescope A 1689, Astron. Astrophys. 594 (2016) A54
[294] M. Zumalacárregui and U. Seljak, Limits on Stellar-Mass [1607.01617].
Compact Objects as Dark Matter from Gravitational [312] T. Petrushevska, T. Okamura, R. Kawamata,
Lensing of Type Ia Supernovae, Phys. Rev. Lett. 121 L. Hangard, G. Mahler and A. Goobar, Prospects for
(2018) 141101 [1712.02240]. Strongly Lensed Supernovae Behind Hubble Frontier
[295] D. Eichler, Nanolensed Fast Radio Bursts, Astrophys. J. Fields Galaxy Clusters with the James Webb Space
850 (2017) 159 [1711.04764]. Telescope, Astronomy Reports 62 (2018) 917
[296] J. Wagner, J. Liesenborgs and D. Eichler, Multiply [1901.02014].
imaged time-varying sources behind galaxy clusters. [313] T. J. Moriya, K. C. Wong, Y. Koyama, M. Tanaka,
Comparing fast radio bursts to QSOs, SNe, and M. Oguri, S. Hilbert et al., Searches for Population III
GRBs, Astron. Astrophys. 621 (2019) A91 pair-instability supernovae: Predictions for
[1811.10618]. ULTIMATE-Subaru and WFIRST, Publ. Astron. Soc.
[297] T. S. Kolatt and M. Bartelmann, Gravitational lensing of Japan 71 (2019) 59 [1903.01613].
type IA supernovae by galaxy clusters, Mon. Not. R. [314] K. C. Wong, T. J. Moriya, M. Oguri, S. Hilbert,
Astron. Soc. 296 (1998) 763 [astro-ph/9708120]. Y. Koyama and K. Nomoto, Searches for Population
[298] J. Nordin, D. Rubin, J. Richard, E. Rykoff, G. Aldering, III pair-instability supernovae: Impact of gravitational
R. Amanullah et al., Lensed Type Ia supernovae as lensing magnification, Publ. Astron. Soc. Japan 71
probes of cluster mass models, Mon. Not. R. Astron. (2019) 60 [1903.01614].
Soc. 440 (2014) 2742 [1312.2576]. [315] L. Dai, T. Venumadhav and K. Sigurdson, Effect of
[299] B. Patel, C. McCully, S. W. Jha, S. A. Rodney, D. O. lensing magnification on the apparent distribution of
Jones, O. Graur et al., Three Gravitationally Lensed black hole mergers, Phys. Rev. D 95 (2017) 044011
Supernovae behind CLASH Galaxy Clusters, [1605.09398].
Astrophys. J. 786 (2014) 9 [1312.0943]. [316] Y. Suwa, Supernova forecast with strong lensing, Mon.
[300] S. A. Rodney, B. Patel, D. Scolnic, R. J. Foley, Not. R. Astron. Soc. 474 (2018) 2612 [1711.00183].
A. Molino, G. Brammer et al., Illuminating a Dark [317] E. Babichev and V. Dokuchaev, Gravitational lensing of
Lens : A Type Ia Supernova Magnified by the Frontier relativistic fireball, Physics Letters A 265 (2000) 168
Fields Galaxy Cluster Abell 2744, Astrophys. J. 811 [astro-ph/9912383].
(2015) 70 [1505.06211]. [318] B. S. Gaudi and A. Loeb, Resolving the Image of
[301] S. Mao and P. Schneider, Evidence for substructure in Gamma-Ray Burst Afterglows with Gravitational
lens galaxies?, Mon. Not. R. Astron. Soc. 295 (1998) Microlensing, Astrophys. J. 558 (2001) 643
587 [astro-ph/9707187]. [astro-ph/0102003].
[302] C. R. Keeton and L. A. Moustakas, A New Channel for [319] K. Ioka and T. Nakamura, Microlensing of Collimated
Detecting Dark Matter Substructure in Galaxies: Gamma-Ray Burst Afterglows, Astrophys. J. 561
Gravitational Lens Time Delays, Astrophys. J. 699 (2001) 703 [astro-ph/0102028].
(2009) 1720 [0805.0309]. [320] R. Perna and C. R. Keeton, Gravitational lensing of
[303] K. Liao, X. Ding, M. Biesiada, X.-L. Fan and Z.-H. Zhu, anisotropic sources, Mon. Not. R. Astron. Soc. 397
Anomalies in Time Delays of Lensed Gravitational (2009) 1084 [0904.3935].
Waves and Dark Matter Substructures, Astrophys. J. [321] L. Dai and W. Lu, Probing Motion of Fast Radio Burst
867 (2018) 69 [1809.07079]. Sources by Timing Strongly Lensed Repeaters,
[304] L. Dai, S.-S. Li, B. Zackay, S. Mao and Y. Lu, Detecting Astrophys. J. 847 (2017) 19 [1706.06103].
lensing-induced diffraction in astrophysical [322] Y. Wang, Supernova Pencil Beam Survey, Astrophys. J.
gravitational waves, Phys. Rev. D 98 (2018) 104029 531 (2000) 676 [astro-ph/9806185].
[1810.00003]. [323] A. Goobar, E. Mörtsell, R. Amanullah and P. Nugent,
[305] D. E. Holz, M. C. Miller and J. M. Quashnock, Cosmological parameters from lensed supernovae,
Gravitational Lensing Limits on the Average Redshift Astron. Astrophys. 393 (2002) 25 [astro-ph/0207139].
of Gamma-Ray Bursts, Astrophys. J. 510 (1999) 54 [324] A. Goobar, E. Mörtsell, R. Amanullah, M. Goliath,
[astro-ph/9804271]. L. Bergström and T. Dahlén, SNOC: A Monte-Carlo
[306] C. Li and L. Li, Constraining fast radio burst progenitors simulation package for high-z supernova observations,
with gravitational lensing, Science China Physics, Astron. Astrophys. 392 (2002) 757
Mechanics, and Astronomy 57 (2014) 1390 [astro-ph/0206409].
[1403.7873]. [325] Z. Kostrzewa-Rutkowska, L. Wyrzykowski and
[307] G. Aldering, A. G. Kim, M. Kowalski, E. V. Linder and M. Jaroszyński, On rates of supernovae strongly lensed
S. Perlmutter, Snapping supernovae at z > 1.7, by galactic haloes in Millennium Simulation, Mon.
Astroparticle Physics 27 (2007) 213 Not. R. Astron. Soc. 429 (2013) 2392 [1212.1063].
[astro-ph/0607030]. [326] N. Kaiser, H. Aussel, B. E. Burke, H. Boesgaard,
[308] C.-E. Rydberg, D. J. Whalen, M. Maturi, T. Collett, K. Chambers, M. R. Chun et al., Pan-STARRS: A
M. Carrasco, M. Magg et al., Detecting strongly lensed Large Synoptic Survey Telescope Array, in Survey and
supernovae at z ˜5-7 with LSST, arXiv e-prints (2018) Other Telescope Technologies and Discoveries, J. A.
arXiv:1805.02662 [1805.02662]. Tyson and S. Wolff, eds., vol. 4836 of Society of
[309] T. Pan and A. Loeb, Finding core-collapse supernovae Photo-Optical Instrumentation Engineers (SPIE)
from the epoch of reionization behind cluster lenses., Conference Series, pp. 154–164, Dec, 2002, DOI.
Mon. Not. R. Astron. Soc. 435 (2013) L33 [327] K. C. Chambers, E. A. Magnier, N. Metcalfe, H. A.
[1303.6960]. Flewelling, M. E. Huber, C. Z. Waters et al., The
[310] D. J. Whalen, J. Smidt, C.-E. Rydberg, J. L. Johnson, Pan-STARRS1 Surveys, arXiv e-prints (2016)
D. E. Holz and M. Stiavelli, Detecting Ancient arXiv:1612.05560 [1612.05560].
Supernovae at z ˜5 - 12 with CLASH, arXiv e-prints [328] Y. Shu, A. S. Bolton, S. Mao, X. Kang, G. Li and
Strong lensing of explosive transients 41
M. Soraisam, Prediction of Supernova Rates in Known M. Oguri et al., “Refsdal” Meets Popper: Comparing
Galaxy-Galaxy Strong-lens Systems, Astrophys. J. 864 Predictions of the Re-appearance of the Multiply
(2018) 91 [1803.07569]. Imaged Supernova Behind MACSJ1149.5+2223,
[329] M. Sullivan, R. Ellis, P. Nugent, I. Smail and P. Madau, Astrophys. J. 817 (2016) 60 [1510.05750].
A strategy for finding gravitationally lensed distant [345] J. M. Diego, T. Broadhurst, C. Chen, J. Lim, A. Zitrin,
supernovae, Mon. Not. R. Astron. Soc. 319 (2000) B. Chan et al., A free-form prediction for the
549 [astro-ph/0007228]. reappearance of supernova Refsdal in the Hubble
[330] T. D. Saini, S. Raychaudhury and Y. A. Shchekinov, Frontier Fields cluster MACSJ1149.5+2223, Mon.
Observing high-redshift supernovae in lensed galaxies, Not. R. Astron. Soc. 456 (2016) 356 [1504.05953].
Astron. Astrophys. 363 (2000) 349 [346] R. Kawamata, M. Oguri, M. Ishigaki, K. Shimasaku and
[astro-ph/0002448]. M. Ouchi, Precise Strong Lensing Mass Modeling of
[331] V. Stanishev, A. Goobar, K. Paech, R. Amanullah, Four Hubble Frontier Field Clusters and a Sample of
T. Dahlén, J. Jönsson et al., Near-IR search for lensed Magnified High-redshift Galaxies, Astrophys. J. 819
supernovae behind galaxy clusters. I. Observations and (2016) 114 [1510.06400].
transient detection efficiency, Astron. Astrophys. 507 [347] M. Jauzac, J. Richard, M. Limousin, K. Knowles,
(2009) 61 [0908.4176]. G. Mahler, G. P. Smith et al., Hubble Frontier Fields:
[332] T. Petrushevska, A. Goobar, D. J. Lagattuta, predictions for the return of SN Refsdal with the
R. Amanullah, L. Hangard, S. Fabbro et al., Searching MUSE and GMOS spectrographs, Mon. Not. R.
for supernovae in the multiply-imaged galaxies behind Astron. Soc. 457 (2016) 2029 [1509.08914].
the gravitational telescope A370, Astron. Astrophys. [348] C. Grillo, W. Karman, S. H. Suyu, P. Rosati, I. Balestra,
614 (2018) A103 [1802.10525]. A. Mercurio et al., The Story of Supernova “Refsdal”
[333] A. Goobar, K. Paech, V. Stanishev, R. Amanullah, Told by Muse, Astrophys. J. 822 (2016) 78
T. Dahlén, J. Jönsson et al., Near-IR search for lensed [1511.04093].
supernovae behind galaxy clusters. II. First detection [349] S. H. Suyu and A. Halkola, The halos of satellite
and future prospects, Astron. Astrophys. 507 (2009) galaxies: the companion of the massive elliptical lens
71 [0810.4932]. SL2S J08544-0121, Astron. Astrophys. 524 (2010)
[334] R. Amanullah, A. Goobar, B. Clément, J. G. Cuby, A94 [1007.4815].
H. Dahle, T. Dahlén et al., A Highly Magnified [350] S. H. Suyu, S. W. Hensel, J. P. McKean, C. D. Fassnacht,
Supernova at z = 1.703 behind the Massive Galaxy T. Treu, A. Halkola et al., Disentangling Baryons and
Cluster A1689, Astrophys. J. 742 (2011) L7 Dark Matter in the Spiral Gravitational Lens
[1109.4740]. B1933+503, Astrophys. J. 750 (2012) 10 [1110.2536].
[335] D. Rubin, B. Hayden, X. Huang, G. Aldering, [351] M. Meneghetti, P. Natarajan, D. Coe, E. Contini, G. De
R. Amanullah, K. Barbary et al., The Discovery of a Lucia, C. Giocoli et al., The Frontier Fields lens
Gravitationally Lensed Supernova Ia at Redshift 2.22, modelling comparison project, Mon. Not. R. Astron.
Astrophys. J. 866 (2018) 65 [1707.04606]. Soc. 472 (2017) 3177 [1606.04548].
[336] R. Chornock, E. Berger, A. Rest, D. Milisavljevic, [352] S. A. Rodney, L. G. Strolger, P. L. Kelly, M. Bradač,
R. Lunnan, R. J. Foley et al., PS1-10afx at z = 1.388: G. Brammer, A. V. Filippenko et al., SN Refsdal:
Pan-STARRS1 Discovery of a New Type of Photometry and Time Delay Measurements of the
Superluminous Supernova, Astrophys. J. 767 (2013) First Einstein Cross Supernova, Astrophys. J. 820
162 [1302.0009]. (2016) 50 [1512.05734].
[337] R. M. Quimby, M. C. Werner, M. Oguri, S. More, [353] P. L. Kelly, G. Brammer, J. Selsing, R. J. Foley,
A. More, M. Tanaka et al., Extraordinary J. Hjorth, S. A. Rodney et al., SN Refsdal:
Magnification of the Ordinary Type Ia Supernova Classification as a Luminous and Blue SN 1987A-like
PS1-10afx, Astrophys. J. 768 (2013) L20 [1302.2785]. Type II Supernova, Astrophys. J. 831 (2016) 205
[338] R. M. Quimby, M. Oguri, A. More, S. More, T. J. [1512.09093].
Moriya, M. C. Werner et al., Detection of the [354] J. Vega-Ferrero, J. M. Diego, V. Miranda and G. M.
Gravitational Lens Magnifying a Type Ia Supernova, Bernstein, The Hubble Constant from SN Refsdal,
Science 344 (2014) 396 [1404.6014]. Astrophys. J. 853 (2018) L31 [1712.05800].
[339] P. L. Kelly, S. A. Rodney, T. Treu, R. J. Foley, [355] C. Grillo, P. Rosati, S. H. Suyu, I. Balestra, G. B.
G. Brammer, K. B. Schmidt et al., Multiple images of Caminha, A. Halkola et al., Measuring the Value of
a highly magnified supernova formed by an early-type the Hubble Constant “à la Refsdal”, Astrophys. J. 860
cluster galaxy lens, Science 347 (2015) 1123 (2018) 94 [1802.01584].
[1411.6009]. [356] L. L. R. Williams and J. Liesenborgs, The role of
[340] P. L. Kelly, S. A. Rodney, T. Treu, L. G. Strolger, R. J. multiple images and model priors in measuring H0
Foley, S. W. Jha et al., Deja Vu All Over Again: The from supernova Refsdal in galaxy cluster MACS
Reappearance of Supernova Refsdal, Astrophys. J. 819 J1149.5+2223, Mon. Not. R. Astron. Soc. 482 (2019)
(2016) L8 [1512.04654]. 5666 [1806.11113].
[341] T. Treu, K. B. Schmidt, G. B. Brammer, B. Vulcani, [357] A. Goobar, R. Amanullah, S. R. Kulkarni, P. E. Nugent,
X. Wang, M. Bradač et al., The Grism Lens-Amplified J. Johansson, C. Steidel et al., iPTF16geu: A multiply
Survey from Space (GLASS). I. Survey Overview and imaged, gravitationally lensed type Ia supernova,
First Data Release, Astrophys. J. 812 (2015) 114 Science 356 (2017) 291 [1611.00014].
[1509.00475]. [358] S. R. Kulkarni, The intermediate Palomar Transient
[342] M. Oguri, Predicted properties of multiple images of the Factory (iPTF) begins, The Astronomer’s Telegram
strongly lensed supernova SN Refsdal., Mon. Not. R. 4807 (2013) 1.
Astron. Soc. 449 (2015) L86 [1411.6443]. [359] A. More, S. H. Suyu, M. Oguri, S. More and C.-H. Lee,
[343] K. Sharon and T. L. Johnson, Revised Lens Model for Interpreting the Strongly Lensed Supernova
the Multiply Imaged Lensed Supernova, “SN Refsdal” iPTF16geu: Time Delay Predictions, Microlensing,
in MACS J1149+2223, Astrophys. J. 800 (2015) L26 and Lensing Rates, Astrophys. J. 835 (2017) L25
[1411.6933]. [1611.04866].
[344] T. Treu, G. Brammer, J. M. Diego, C. Grillo, P. L. Kelly, [360] D. A. Yahalomi, P. L. Schechter and J. Wambsganss, A
Strong lensing of explosive transients 42