Space Science Reviews (2024) 220:11
https://doi.org/! 0.1007/sl 1214-024-01048-3
®Check for
updates
Science with a Small Two-Band UV-Photometry Mission I:
Mission Description and Follow-up Observations of Stellar
Transients
N. Werner1
• J. Řípa1
• C. Thone2
• F. Munz1
• P. Kurfurst1
• M. Jelínek2
• F. Hroch1
•
J. Benáček3
• M. Topinka4
• G. Lukes-Gerakopoulos5
• M. Zajaček1
• M. Labaj1
•
M. Prišegen1
-6
• J. Krtička1
• J. Merc7
• A. Pál8
• O. Pejcha9
• V. Daniel10
• J. Jon1 0
•
R. Šošovička1 0
• J. Gromeš1 0
• J. Václavík1 1
• L Steiger11
• J. Segiňák1 2
• E. Behar13
•
S. Tarem13
• J. Salh13
• O. Reich13
• S. Ben-Ami14
• M.F. Barschke15
• D. Berge15
-16
•
A. Tohuvavohu17
• S. Sivanandam17
• M. Bulla18
-19
-20
• S. Popov21
•
Hsiang-Kuang Chang22
Received: 27 June 2023 / Accepted: 9 January 2024 / Published online: 2 February 2024
©TheAuthor(s) 2024
Abstract
This is the first in a collection of three papers introducing the science with an ultra-violet
(UV) space telescope on an approximately 130 kg small satellite with a moderately fast
re-pointing capability and a real-time alert communication system approved for a Czech national
space mission. The mission, called Quick Ultra-Violet Kilonova surveyor—QUVIK,
will provide key follow-up capabilities to increase the discovery potential of gravitational
wave observatories and future wide-field multi-wavelength surveys. The primary objective
of the mission is the measurement of the U V brightness evolution of kilonovae, resulting
from mergers of neutron stars, to distinguish between different explosion scenarios. The
mission, which is designed to be complementary to the Ultraviolet Transient Astronomy
Satellite—ULTRAS AT, will also provide unique follow-up capabilities for other transients
both in the near- and far-UV bands. Between the observations of transients, the satellite
will target other objects described in this collection of papers, which demonstrates that
a small and relatively affordable dedicated UV-space telescope can be transformative for
many fields of astrophysics.
Keywords U V space observatory • Kilonovae • Gamma-ray bursts • Supernovae
1 Introduction
The first simultaneous detection of gravitational waves and electromagnetic radiation on
2017 August 17 (Abbott et al. 2017b,a), resulting from a coalescence of neutron stars,
marked the onset of multi-messenger astrophysics involving gravitational waves. This exciting
observation showed that neutron star mergers are of major importance for enriching
the Universe with rare heavy elements such as gold and platinum. The radioactive decay of
these heavy elements powers a thermal transient at ultra-violet/visible/infrared wavelengths
Extended author information available on the last page of the article
Springer
11 Page 2 of 41 N. Werner etal.
known as a kilonova (Li and Paczyriski 1998; Metzger et al. 2010; Metzger 2019). To make
further breakthroughs in the study of kilonovae, ultra-violet (UV) observations early after
the explosion are required (Arcavi 2018). In this collection of papers, we present an overview
of the rich science that can be achieved with a U V space telescope on a small satellite with
a moderately fast repointing capability and real-time alert communication system. We show
that the Quick Ultra-Violet Kilonova surveyor (QUVIK) two-band UV-photometry mission
(Werner et al. 2022), which we describe in Sect. 2, can provide a breakthrough in our understanding
of kilonovae as well as significantly expand our knowledge and discovery potential
in other fields of astronomy.
As we show in Sect. 3, the early measurement of the brightness evolution of kilonovae in
the U V band will allow us to distinguish between different merger scenarios. However, the
ability to distinguish between different models depends critically on the capability to point
to the target location and start monitoring the emission early, about an hour after receiving
the gravitational wave signal. In the case of GW170817, which represents the current state
of the art of kilonova observations, the kilonova AT2017gfo was discovered 11 hours after
the merger and the first U V observation was performed 15 hours after the merger with
SwiftfUVOT (Evans et al. 2017). The current scenarios known to produce UV/optical radiation
from neutron star mergers include shock-powered, nucleosynthesis-powered, and
free neutron decay-powered models, and discriminating between them is only possible by
using U V observations performed in the first hours after the merger of neutron stars (Metzger
2019; Dorsman et al. 2023). The early U V emission, in combination with optical and
near-infrared observations that can be performed from the ground, also provides a tool to
probe the properties of the merger remnant, which can be a black hole, a stable massive
neutron star, or a supramassive/hypermassive neutron star that after some time (which can
also be inferred from the U V observations) collapses into a black hole (Kasen et al. 2015;
Metzger 2019; Sarin and Lasky 2021). It also allows us to determine some of the outflow
parameters. This information determined from early U V photometry cannot be gleaned from
observations performed later or at longer wavelengths. This provides an opportunity for truly
breakthrough science, potentially allowing us to constrain the contribution of neutron star
mergers to the formation of heavy, so-called r-process, elements (Lattimer and Schramm
1974; Wanajo et al. 2014).
The small satellite, optimised for early photometry of kilonovae, will be particularly
well suited for the follow-up observations of a wide range of transient sources, such as
gamma-ray bursts (GRBs, Piran 2004; Zhang and Meszaros 2004; Meszaros 2006; Kumar
and Zhang 2015), supernovae (SNe, Bethe 1990; Smartt 2009; Nakar and Sari 2010; Rabinak
and Waxman 2011), outbursts in active galactic nuclei (Ho 2008; Padovani et al. 2017),
tidal disruption of stars by supermassive black holes (Dai et al. 2018; Gezari 2021), etc.
Gamma-ray burst science will benefit greatly from an onboard G R B detector with localisation
capabilities (GALI, Rahin et al. 2020), enabling fast U V follow-up observations that
can currently only be performed with the ageing Neil Gehrels Swift Observatory (Swift,
Gehrels et al. 2004) and soon, in the visible band, by SVOM (Atteia et al. 2022). Such an
onboard G R B detector will also enable the fastest possible follow-up of kilonovae that happen
to coincide with short GRBs. Observations of supernovae and tidal disruption events
will benefit from synergies with other major observatories, which include the Vera Rubin
Observatory (Ivezic et al. 2019) and the Square Kilometre Array (Dewdney et al. 2009). Importantly,
follow-up observations of transients in the U V band will provide opportunities for
unexpected discoveries and for the discovery of new, yet unknown, classes of transients. Between
the observations of transient sources, the satellite will have an opportunity to perform
Springer
Science with QUVIKI Page 3 of 41 11
observations of other targets of interest for the scientific community, such as stars and stellar
systems, and galactic nuclei, described in papers II (Krticka et al. 2024) and III (Zajacek
et al. 2024) of this collection.
A particularly important synergy will be provided by the Ultraviolet Transient Astronomy
Satellite (ULTRASAT, Ben-Ami et al. 2022; Shvartzvald et al. 2023), which is an Israeli mission
carrying a telescope with a 33 cm aperture and a very large FoV of 200 square degrees.
It is optimised for the 230-290 nm near-ultraviolet (NUV) band and is planned to operate in
a geostationary orbit. ULTRASAT is expected to reach a 5er sensitivity of 22.3 A B limiting
magnitude in 3 x 300 s integrations. It will operate in two modes: it will stare at two regions
for 6 months each, and it will perform Target of Opportunity (ToO) observations slewing
on a target within 15 minutes from receiving the trigger position. Follow-up observations of
ULTRASAT targets with QUVIK in complementary N U V and far-ultraviolet (FUV) bands
with a narrower point-spread function (PSF) will help to multiply the scientific return of
both missions.
2 QUVIK Mission Overview
The primary objective of QUVIK, illustrated in Fig. 1, is the early measurement of the brightness
evolution of kilonovae in the U V band to distinguish between different models of their
explosion. This objective drives the design parameters of the mission summarised in Table 1.
Our baseline mission has a moderately large FoV of at least 1.0° x 1.0°, a PSF of <2.5 arcsec
(FWHM), and a N U V photometric sensitivity of 22 A B magnitude in less than 1000
second at signal-to-noise ratio (SNR) of 5, and observation start latency for an unocculted
target (ToO response time) better than 20 minutes. Its N U V spectral range of ~260 nm to
~360 nm is the primary band for kilonova observations and provides complementarity to
ULTRASAT at longer wavelengths. The F U V band spanning from ~ 140 nm to ~ 190 nm will
give QUVIK unique capabilities as this band is currently only covered by U V I T on AstroSat
(Singh et al. 2014).1
Close coordination of the FUV-capable QUVIK with ULTRASAT, will
thus multiply the scientific returns of both missions.
Fig. 1 A render of the proposed
Quick Ultra-Violet Kilonova
surveyor (QUVIK)
On August 18 2022, NASA announced the selection of the Ultraviolet Explorer (UVEX, Kulkarni et al.
2021) as one of two MIDEX proposals for a detailed study. A final down-selection will happen in 2024 with
Springer
11 Page 4 of 41 N.Werner et a I.
Table 1 Baseline QUVIK mission parameters
QUVIK Requirement
NUV FoV
Angular resolution (FWHM)
NUV bandpass
FUV bandpass
NUV photometric sensitivity
FUV photometric sensitivity
Observation start latency (for unocculted targets)
Baseline mission duration
Target launch date
> 1.0° x 1.0°
< 2.5 arcsec
-260-360 nm
-140-190 nm
22 A B mag (SNR 5 in 1000 s)a
20 A B mag (SNR 5 in 1000 s)
<20 min
3 years
2028-2029
J
For a kilonova on the background of its host galaxy at an offset of 1.5 effective radii.
2.1 Orbit
The most cost-effective orbit, with many launch opportunities for small-satellite missions,
is a Sun-synchronous (SSO) low-Earth orbit. To minimise thermal cycling, the preferred
orbit for QUVIK is the dawn-dusk SSO (with a mean local time at the ascending node of
approximately 6:00 A M / P M ) where the spacecraft is orbiting the Earth close to its terminator
and the Sun appears approximately normal to the orbital plane. The currently baselined
orbital altitude of 550 km implies that the spacecraft is illuminated permanently for about
9 months, whereas in the remaining time of the year it is eclipsed for up to 22 minutes per
orbit.
2.2 Spacecraft
The design of the spacecraft is based on a modular small-satellite platform with the primary
telescope payload in its centre and the platform distributed along the side walls. The configuration
indicated in Fig. 2 shows the platform distributed in four modules/blocks (Modules
1-4), the spacecraft adapter/separation system at the bottom (Module 5), and the module
with the baffle and telescope front door at the top (Module 6). The telescope is mounted
via three bipods to minimise the transfer of mechanical stresses and vibrations from the
spacecraft or launcher.
The spacecraft has 3 solar panels (Modules 1, 2 and around the spacecraft adapter/separation
system on Module 5), which can generate between 64 W and 200 W, depending on
the orientation with respect to the Sun. The total power consumption of all subsystems is
up to 100 W. The observation planning will be driven by the total power, which needs to be
generated to maintain the spacecraft subsystems during the observations.
The spacecraft is stabilised in 3 axes by 4 reaction wheels in a pyramidal configuration.
The reaction wheels are continuously offloaded by 3 perpendicular magnetorquers. Attitude
knowledge is provided by the gyro-stellar estimator fusing information from two onboard
the launch expected sometime after 2028, likely around 2030. UVEX, which is a significantly larger mission
than QUVIK, would cover the same spectral bands and, if selected, the mission would most likely launch
after our proposed small satellite. QUVIK could thus serve as a pathfinder for UVEX and even if operated
simultaneously, it would provide a large discovery potential in the very rich field of UV astronomy.
Springer
Science with QUVIKI Page 5 of 41 11
Fig. 2 The modular design of Module 6
QUVIK with the telescope placed
in the centre of the spacecraft and
the platform distributed around it Module 1
in six individual blocks
GALI
Module 4
Radiator Plate
star trackers and a navigation-grade fibre-optic gyroscope. A magnetometer is planned to be
used to support commanding magnetic torquers and as a safe-mode attitude sensor. Eight
coarse sun-sensors are used in safe mode, guaranteeing high robustness, low power consumption,
and the ability to determine the Sun vector in almost any spacecraft condition. A
Global Navigation Satellite System (GNSS) receiver is intended to be used for onboard time
synchronisation and as a position sensor.
To achieve our goal for the pointing stability of 2.5 arcsec within a 60 s exposure and sufficient
agility for fast re-pointing (our goal of <20 min observation start latency requires at
least 0.4° s - 1
) , the spacecraft is designed as compact and rigid as possible, without antenna
booms or deployable solar panels.
The majority of science data processing will be performed on the ground, but the satellite
is also expected to perform several processing tasks onboard. The tasks performed onboard
include combining shorter exposures of the same field into one image with a higher SNR;
computing metrics and meta-data for image evaluation, including image orientation, level of
cosmic-ray contamination, and possible blurring of PSF due to Attitude and Orbit Control
System (AOCS) instability; creating cut-outs from the original full-frame images. These
tasks will significantly reduce the amount of data for downlink to ground and improve the
image quality.
The satellite will use an S-band radio for telemetry, tracking, and command uplink and
downlink. For redundancy, two S-band radios and two patch antennas will be used on opposite
sides of the spacecraft. A n X-band radio will be used for data downlink, allowing to
transfer 400 full images (32 M B each) twice a day. A dedicated L-band radio will be implemented
for near real-time bi-directional communication through a geostationary satellite
constellation.
The satellite will by default follow an observing schedule of non-transient targets predetermined
by the science operation centre (SOC). Once a gravitational wave or electromagnetic
transient, which fulfils the criteria for target-of-opportunity (TOO) observations,
is identified by the SOC, a command will be sent to the satellite to terminate the ongoing ob-
Springer
11 Page 6 of 41 N.Werner et a I.
Table 2 Baseline telescope
parameters Physical collection area
Effective area
635 cm2
NUV: 140-180 cm2
FUV: 20-25 cm2
NUV channel FoV 1° x 1°
NUV channel focal length
FUV channel FoV
FUV channel focal length
Optical diameter
Sun avoidance angle
Earth limb avoidance angle
Moon avoidance angle
Instantaneously accessible sky fraction
Per-orbit accessible sky fraction 0.79
0.43
20°
20°
50°
0.25° x 0.25°
330 mm
1212 mm
1064 mm
servation and slew on the TOO target. The requirement is that the satellite receives the alert
within 5 minutes after it has been issued by the SOC and will slew on target and commence
observation within 15 minutes after the alert reception. While the instantaneously accessible
sky fraction on low-Earth orbit is 43%, during the 90 minute orbit the satellite can access
nearly 80% of the sky (see Table 2).
G W observatories are expected to provide alerts with rough parameter estimates within
minutes and refined parameters within hours. ULTRASAT, with its fast slewing capability,
a considerable fraction of instantaneously accessible sky, and very large FoV, is expected
to quickly detect and localise a substantial fraction of K N e out to a distance of at least
300 Mpc and within 15 minutes distribute an alert to the astronomical community (Shvartzvald
et al. 2023). Follow-up observations with QUVIK will provide two-band photometry in
complementary spectral bands and with a higher spatial resolution. The PSF of 2.5 arcsec
will enable us to better isolate K N e on the background of their host galaxies. Following the
G W alert with the initial localisation, and before a more precise localisation is established,
QUVIK will start mosaic observations of the potential target area. The mosaic observations
will be performed in coordination with other observatories/teams to minimise unnecessary
search overlap and find the counterpart as quickly and efficiently as possible, e.g. employing
the Gravitational Wave Treasure Map tool2
(Wyatt et al. 2020).
2.3 Science Payloads
2.3.1 The UV-Telescope and Detectors
The primary payload consists of a modified Cassegrain telescope with a 33 cm diameter
primary mirror (Fig. 3). A dichroic mirror splits the light into N U V and F U V channels that
are simultaneously imaged by two focal planes. The N U V channel path is a classic 2-mirror
telescope with a field corrector and a F o V of 1° x 1° imaged by a 4k x 4k complementary
metal-oxide-semiconductor (CMOS) sensor. The sensors being considered for QUVIK are
the GSENSE4040BSI3
and the UV-detector developed for ULTRASAT (Asif et al. 2021;
Bastian-Querner et al. 2021; Liran et al. 2022). The corrector contains 4 lenses of fused
http://treasuremap.space.
https://www.gpixel.com/en/pro_details_l 1100.html.
_} Springer
Science with QUVIKI Page 7 of 41 11
200 mm
Fig. 3 The optical design of the QUVIK telescope payload. Colours represent light rays from different parts
of the FoV. The NUV focal plane is behind the primary mirror. The corrector contains 4 lenses made of fused
silica, magnesium fluoride and calcium fluoride. The corrector has an aspheric shape and the first surface of
the first lens also serves as a beam splitter. The FUV optical path is formed by the primary and secondary
mirrors with the front surface of the first lens of the corrector. The FUV focal plane is behind the secondary
mirror
silica, magnesium fluoride and calcium fluoride. It has an aspheric shape and the first surface
of the first lens also serves as a beam splitter. The F U V optical path is formed by the primary
and secondary mirrors with the front surface of the first lens of the corrector. The F U V
focal plane is behind the secondary mirror. The F o V of the F U V channel is 0.25° x 0.25°
and is imaged by a C M O S sensor. The quantum efficiency of modern backside illuminated
(BSI) C M O S detectors enables large improvements with respect to previously flown U V
missions (e.g. Nikzad et al. 2017). After accounting for losses in the optical system, taking
into account the properties of mirror and corrector coatings, as well as focal plane bandpass
filters and the quantum efficiency of the preselected C M O S sensor, the final effective area
in the N U V band will be 140-180 c m 2
. The expected effective area in the F U V band, after
accounting for all the losses, will be 20-25 c m 2
(see Table 2 for a summary of telescope
parameters).
The detectors require cooling to a temperature of —30 °C in order to reach a dark current
of 0.1 e ~ s _ 1
p x _ 1
, which is needed to reduce the noise. To maintain the required photometric
accuracy, the detectors need to be kept at a stable, selected temperature for which the pixel
gains were calibrated. The preferred cooling method is passive heat removal by a heat-strap
connected to the spacecraft radiator. The detector temperature will be stabilised using a
resistive heater.
The sensitivity, i.e. the SNR of an observed point source, has been calculated as a function
of the total accumulation time for co-added images with shorter individual exposures in the
N U V band of 260-360 nm following the standard charge-coupled device (CCD) equation
(Howell 2000; Keller et al. 2015):
where the total accumulated exposure t = Nextex is obtained by co-adding Nex number of
short exposures tex. We have assumed the effective area of the telescope, including the quantum
efficiency of the GSENSE4040BSI sensor (one of the candidates for the QUVIK detector)
and the transparency of the optical path, to be 180 c m 2
. The other parameters are:
SNR = (1)
Springer
11 Page 8 of 41 N.Werner et a I.
• n* = 1.27e~ s - 1
for 21 A B mag and 0.504 e~ s - 1
for 22 A B mag are the detected photoelectron
rates from the source in the band of 260-360 nm (accounting for the optical
throughput of the telescope and the quantum efficiency of the N U V sensor).
• n p i x = 4 (2 x 2 pixels) is the number of pixels covering the PSF assuming a pixel resolution
of 1.75 arcsec p x - 1
.
• n Z L = 0.03 e ~ s _ 1
p x _ 1
is the detected zodiacal light background as the photo-electron
rate for the 260-360 nm band and for a typical brightness of 100 Sin(V) (Leinert 1975;
Levasseur-Regourd and Dumont 1980).
• nG = 0.06 e ~ s _ 1
p x _ 1
is the detected host galaxy background as the photo-electron rate
for a surface brightness of 25 mag arcsec- 2
for the offset of 1.5 Re (Re is the effective
galaxy radius) of a kilonova from the centre of the galaxy, which is the median offset of
short GRBs from their host galaxies.
• nD = 0.1 e - s - ^ x " 1
is the dark current of the GSENSE4040BSI sensor at - 3 0 °C.
• n$ = 1.0 e ~ s _ 1
p x _ 1
is the photo-electron background rate due to stray-light.
• NR = 3.9 e" is the readout noise of the GSENSE4040BSI sensor.
For these parameters, the effective area in the N U V band is sufficient to detect a 22 A B
magnitude object with a S N R = 5 using a 900 s (15 min) exposure image stacked from
individual tex = 20 s exposures. In the case of good stability and low jitter (according to the
current estimates, the jitter will be better than 1.8 arcsec over 60 s), the individual exposures
will be extended to at least tex = 60 s, reducing the time to reach magnitude 22 A B to about
720 s (12 min). The A B magnitude of 23 will be achievable in 5460 s (91 min) by stacking
individual 20 s exposures and in 4200 s (70 min) by using tex = 60 s sub-exposures. Because
the satellite will be placed in low-Earth orbit, where most objects are only continuously
observable for a few tens of minutes, reaching the A B magnitude of 23 will require at least
two orbits of QUVIK (one orbit takes ~90 minutes).
To estimate the F U V sensitivity, we have assumed the effective area of the telescope,
including the estimate of the quantum efficiency of the GSENSE2020BSI sensor4
(Gill et al.
2022), a candidate for the F U V detector, and the transparency of the optical path, to be
20 cm2
. The other parameters are:
• n* = 0.35 e~ s - 1
for 20 A B mag is the detected photo-electron rate from the source in the
140-190 nm band (accounting for the effective area).
• n p i x = 9 (3 x 3 pixels) is the number of pixels covering the PSF assuming a pixel resolution
of 1.11 arcsec p x - 1
.
• n D = 0.1 e ~ s _ 1
p x _ 1
is the assumed dark current of the GSENSE2020BSI sensor at
- 3 0 °C.
• ns = 0.05 e ~ s _ 1
p x _ 1
is the assumed photo-electron background rate due to stray-light.
• NR = 2.67 e" is the readout noise of the GSENSE2020BSI sensor (Gill et al. 2022).
• The zodiacal light background n_L as well as the host galaxy background HQ are negligible
in the F U V band.
For these parameters, the effective area in the F U V band is sufficient to detect a 20 A B
magnitude object with a S N R = 5 using a 1000 s exposure image stacked from individual
tex = 20 s exposures.
Given the limits for the N U V stray-light contribution to the background, the pointing
direction for observations performed at nominal sensitivity will have to be > 50° from the
Sun, > 20° from the Earth limb, and > 20° from the Moon (see Table 2).
https://www.gpixel.com/en/pro_details_l 194.html.
_} Springer
Science with QUVIKI Page 9 of 41 11
2.3.2 The Gamma-Ray Burst Detector
QUVIK is expected to host a secondary payload, a gamma-ray burst (GRB) detector capable
of localising an average G R B with a few degrees accuracy. The detector concept is
the Gamma-ray burst Localizing Instrument (GALI, Rahin et al. 2020), which is developed
by the Technion (Israel). G A L I uses a novel approach of a 3-D active coded mask collecting
signals from hundreds of small scintillators. Localisation is achieved from the mutual
shadowing between scintillators. Since most scintillators are hidden inside the detector assembly,
they benefit from a low background and produce a source signal only for GRBs in
specific directions. The compact configuration is made possible by Si-PM technology. Since
the scintillators serve as both detecting units and a mask, G A L I provides relatively uniform
coverage and sensitivity of the entire observable sky.
The G A L I method is scalable and can be built to practically any size. A larger detector
will naturally be more sensitive and also enable more accurate localisation. We are simulating
several configurations and locations on the QUVIK satellite, seeking the best performance
for the mission, while also considering engineering constraints. A n engineering
model with 362 c m 3
scintillators, which is currently being tested in the laboratory, is shown
in Fig. 4. The G R B localisation accuracy will eventually depend on the burst flux and
the detector size, with a goal of ~ 1° for bright bursts and better than 10° for the faintest
ones still detected. The G R B detector will provide an onboard trigger and accurate localisation
information, allowing QUVIK to autonomously slew to the G R B location and start a
measurement sequence with the U V telescope. The observing sequence could include tiled
observations for GRBs where the localisation accuracy is not sufficient for a single pointing.
The exact set of conditions for triggering a G R B follow-up is not yet determined and will
be clarified in the following stages of the mission development.
The potential angular resolution of G A L I was simulated with the M E G A l i b tool
(Zoglauer et al. 2006), comparing two detector configurations; one with 9 layers and 362
scintillators, and another with a 2 x 5-layer configuration and 406 scintillators. The first
configuration corresponds to the 12.5 x 12.5 x 12.5 cm3
prototype shown in Fig. 4 and
the second one, occupying a volume of 25 x 12.5 x 7 cm3
, to the model in the current deFig.
4 Engineering model of
GALI proposed as the GRB
detector for QUVIK. This
detector includes 362 CsI(Tl)
scintillator crystals. The
individually wrapped scintillators
are scattered in a 3-D
configuration to exploit their
mutual shadowing for burst
localization
Springer
11 Page 10 of 41 N.Werner et al.
Table 3 Estimated directional errors for GRBs with a 10-1000 keV flux of 5 ph s _ 1
c m - 2
, averaged over the
sky based on 120 000 simulated bursts in random positions. GALI (top rows) refers to the current design of
362 scintillator cubes occupying a 12.5 x 12.5 x 12.5 crrr* volume. The 2 x 5-layer configuration refers to an
alternative design of two sets of 5 layers with 406 scintillator cubes, occupying a 25 x 12.5 x 7 cirr5
volume.
The mean error refers to the mean of the deviations of the reconstructed positions from the true locations
GALI with 362 detectors
Time[s] 1 5 10 20 40
Detected photons 444.9 2224 4449 8898 17,796
Background photons 2065 10,324 20,647 41,295 82,590
Mean error [°] 4.781 1.93 1.346 0.963 0.723
90% Range [°] 0.97-11.48 0.42^1.20 0.32-2.86 0.32-2.86 0.16-1.48
2 x 5 layers with 406 detectors
Detected photons 613 3067 6134 12,268 24,536
Background photons 2808 14,042 28,084 56,169 112,338
Mean Error [°] 4.527 1.785 1.272 0.957 0.723
90% Range [°] 0.85-11.39 0.4-3.87 0.28-2.72 0.22-2.72 0.16-1.51
sign of the satellite shown in Fig. 2. We assume that the effective area of the detector is
zero for the half of the sky blocked by the satellite. Since part of the field of view is also
blocked by the Earth, the visible area of the sky is reduced to ~ \.6TT sr. For the rest of
the sky, the effective area of the 9-layer configuration spans 56-107 c m 2
with an average
of 90 cm2
. The 2 x 5-layer detector is less uniform, but overall more sensitive, with an
effective area of 39-152 c m 2
and an average of 100 c m 2
. We simulated 120,000 bursts
in random positions using generic G R B spectra assuming a Band function with a constant
photon flux of 5 ph s - 1
c m - 2
in the 10-1000 keV band (see Rahin et al. 2020). The mean
energy in this band is 87 keV, which results in a conversion from photon flux to energy
flux of 1 ph s - 1
c m - 2
to 1.39 x 10~7
erg s - 1
c m - 2
. The bursts were simulated along with
the expected sky background considering burst durations of 1, 5, 10, 20, and 40 s. The
GRB direction was then estimated by comparing the distribution of detected photons on the
scintillators to a calibration-based sky map prepared in advance. The accuracy statistics are
presented in Table 3. We estimate that G A L I will detect around 80 GRBs per year of which
~12 will be short ones. For around 10 GRBs per year, we expect to achieve a better than
1-degree localisation.
3 Kilonovae
Mergers of binary neutron stars (BNS) result in a large amount of ejected neutron-rich material
where heavy elements with Z > 26 are synthesised via rapid neutron capture nucleosynthesis,
the so-called r-process (Lattimer and Schramm 1974; Wanajo et al. 2014).
Radioactive decay of this freshly produced material results in a transient source called kilonova
(KN, L i and Paczyhski 1998; Kulkarni 2005; Metzger et al. 2010; Metzger 2019). The
current state-of-the-art in their observational studies was provided by the detection of the
kilonova AT2017gfo, which resulted from the B N S merger that produced the gravitational
wave source GW170817 (Abbott et al. 2020). The following large observing campaign revealed
that the merger indeed resulted in a K N , the emission of which initially peaked in
N U V and over about 10 days evolved redward.
Springer
Science with QUVIKI Page 11 of 41 11
Inspiral Dynamical Accretion Remnant
Short GRB
c
_
in
Seconds Milliseconds 10 ms 1 00 ms Minutes-hours Hours-days Days-weeks Months-years
Fig. 5 The top shows the phases of a BNS merger as a function of time, showing the associated observational
GW and E M signatures. The bottom shows the physical phenomena producing the above signatures (adopted
from Fernandez and Metzger 2016)
Figure 5 from Fernandez and Metzger (2016) shows the phases of a B N S merger and
its observational signatures. The final stages of the inspiral result in a gravitational wave
signal and may also produce an electromagnetic (EM) precursor. The coalescence leads to
the ejection of 10~4
-10~2
MQ of unbound material (Metzger and Berger 2012). The ejected
matter that remains gravitationally bound to the resulting compact object, which may be a
massive neutron star or a black hole, falls back and forms an accretion disc that launches
relativistic jets, which produce a G R B . While the G R B can only be observed when viewed
close to the jet axis, the emission of the K N is nearly isotropic, with the spectral properties
possibly changing depending on the viewing angle (e.g. Kawaguchi et al. 2018; Darbha and
Kasen 2020; Korobkin et al. 2021; Collins et al. 2023; Bulla 2023). The ejecta is expected
to be rich in newly created heavier nuclei, the decay of which results in emission that turns
quickly from UV, through blue and red, to infrared. As discussed in the next section, the
source of the early U V and blue emission in the kilonova AT2017gfo is not yet clear (Arcavi
2018).
Gravitational wave (GW) observatories have recently also discovered another B N S
merger candidate G W 190425 as well as three neutron star—black hole (NS-BH) coalescences
(Abbott et al. 2023), but follow-up observations did not reveal any electromagnetic
counterparts for these events (Anand et al. 2021). Based on theoretical arguments, we would
expect that, for sufficiently small mass ratios, N S - B H mergers will also produce bright KNe
(Li and Paczyhski 1998; Kawaguchi et al. 2016). Thus, follow-up observations of these
exciting events present important discovery potential.
While GW170817 was quickly followed by a short G R B seen at an angle of 19°-25°
from the jet axis (Mooley et al. 2022), it is likely that most BNS mergers will not result in
an observable G R B . The prompt gamma-ray emission is strongly directional. It has been
estimated that only 1-2% of K N e detected by gravitational wave observatories might be
observable in gamma-rays (Metzger 2019; Colombo et al. 2022). However, recent work by
Dimitrova et al. (2023) argues that nearby short GRBs have, on average, broader jets ( @ j e t >
30°) than the more narrowly-beamed cosmological short GRBs due to detection selection
effects. Determining the fraction of gravitational wave events arising from B N S mergers
with associated short GRBs is of key importance. This fraction might be significantly higher
than previously expected.
Springer
11 Page 12 of 41 N.Werner et al.
16
MJD - 57982.529
Fig. 6 The UV-optical-NIR light curves of AT2017gfo and spherically symmetric three-component models
from Villar et al. (2017). The data were assembled from 18 different papers and 46 instruments and include
647 individual measurements obtained from 0.45 days to 29.4 days after the merger. The three-component
model includes a "blue" lanthanide-poor component (opacity component K — 0.5 cm2
g _ 1
) with ejecta
mass of Me j ~ 0.020 MQ and a velocity uej ~ 0.27c; an intermediate opacity component (K — 3 cm2
g _ 1
)
with Me j ~ 0.047 MQ and uej ~ 0.15c; and a "red" lanthanide-rich component (K — 10 cm2
g _ 1
) with
Me j ~ 0.011 MQ and i>ej ~ 0.14c. The light curves show clearly that while in NIR and optical the kilonova
was visible for weeks its U V emission dropped extremely fast
For several long GRBs (e.g., G R B 060605 and G R B 060614), which are typically connected
with massive star collapse, deep optical observations excluded an accompanying
supernova (Fynbo et al. 2006). Furthermore, GRBs with short peaked gamma-ray emission
followed by a spectrally softer extended emission (EE-SGRBs) have been proposed to originate
from mergers of compact objects (Norris 2002; Norris and Bonnell 2006; Gehrels et al.
2006). Follow-up observations of G R B 211211A with a duration of more than 30 s at a distance
of 350Mpc revealed a kilonova (Rastinejad et al. 2022; Troja et al. 2022), providing
a possible solution to the mystery of long GRBs without supernovae. Very recent observations
of the exceptionally bright, long-duration G R B 230307A also suggest the presence
of a kilonova (Levan et al. 2023; Bulla et al. 2023). Therefore, follow-up observations of
long-duration GRBs with QUVIK will be important and some of them may lead to the discovery
of new KNe. This highlights the enhanced discovery potential of the mission with an
onboard G R B detector.
The optical and U V counterparts of GW170817 were only discovered 11 hours and 15
hours, respectively, after the coalescence of neutron stars (Abbott et al. 2017c; Evans et al.
2017). Figure 6 from Villar et al. (2017) presents UV, optical, and near-infrared (NIR) light
curves along with spherically symmetric three-component models calculated for different
filters with the highest likelihood scores and their ler uncertainties. Earlier observations,
which are necessary for the understanding of the physics of KNe, remain unavailable. Here,
we argue that a UV-photometry mission with a fast-repointing capability, enabling observations
of the early emission, would result in real breakthroughs in our understanding of KNe
and their nucleosynthesis.
Springer
Science with QUVIKI Page 13 of 41 11
3.1 Localization of BNS Mergers with Gravitational-Wave Detectors
The estimated rate of mergers involving neutron stars within the distance of 200 Mpc ranges
from a few to a few tens per year (Abbott et al. 2020). With further upgrades, the sensitivities
of the Laser Interferometer Gravitational-Wave Observatory (LIGO; LIGO Scientific Collaboration
et al. 2015), Virgo (Acernese et al. 2015), and the Kamioka Gravitational Wave
Detector ( K A G R A , Akutsu et al. 2021) will continue to increase. After 2027 it is expected
that a single detector of advanced LIGO will be able to detect B N S mergers up to the distance
of 240-325 Mpc; advanced Virgo alone up to 150-260 Mpc; and K A G R A 5
alone up
to 128 Mpc. The full LIGO-Virgo-KAGRA (LVK) G W network is expected to detect B N S
mergers up to the redshift z ^ 0.3 (luminosity distance of ~1600 Mpc; Petrov et al. 2022).
Importantly, by mid-2020s, during the observation run 04, the credible region of the localisation
area of a significant number of BNS mergers is predicted to reach < 100 square
degrees (Abbott et al. 2020; Petrov et al. 2022). After 2027, during the observation run 05,
the number of well-localised BNS mergers (90% credible area < 100 square degrees) by the
full L V K G W network is expected to be 9-90 yr"1
(Petrov et al. 2022).
3.2 Detectability of KNe with QUVIK
3.2.1 Luminosity of KNe
So far, only three kilonovae have been covered well by multi-wavelength observations:
G R B 160821B (Lamb et al. 2019; Troja et al. 2019), GW170817/AT2017gfo (Abbott et al.
2017b), and G R B 211211A (Rastinejad et al. 2022). Moreover, only a few K N candidates
accompanying gamma-ray bursts have been observed: G R B 050709 (Fox et al. 2005; Hjorth
et al. 2005; Jin et al. 2016), G R B 060614 (Gal-Yam et al. 2006; Yang et al. 2015), and G R B
130603B (Tanvir et al. 2013). The luminosity distribution of K N e is thus currently not well
known and observations provided by QUVIK are expected to be of high value.
Figure 7 shows a simulated N U V band QUVIK image of a MAB =21 K N at an angular
separation from its host galaxy of 1.5 effective radii (Re) at a distance of 200 Mpc with an
exposure time of 1000 s. We simulated a K N of MAB = 21 for illustrative purpose. Most KNe
will be dimmer as seen from Fig. 9 and we expect 3.1^3 y r _ 1
K N e detectable by QUVIK
brighter than MAB = 21 to occur at a distance up to 200 Mpc. QUVIK's angular resolution
will enable it to clearly detect a K N on the background of its host, unless it occurs in the
galaxy's bright central region.
The left panel of Fig. 8 from Ascenzi et al. (2019) presents the cumulative peak u band
luminosity distribution for ten short GRBs modelled with synchrotron afterglow emission
plus the contribution of the K N in the UV/optical/NIR wavelengths powered by the radioactive
decay of r-process elements. The peak absolute magnitude ranges from —10.0 to —17.3.
Recent work by Sagues Carracedo et al. (2021) presents the distribution of nucleosynthesispowered
K N light curves in the Z T F gri filter system (Bellm et al. 2019) using the 3D
Monte Carlo (MC) radiative transfer code POS S I S 6
(Polarization Spectral Synthesis In Supernovae;
Bulla 2019). Time-dependent spectral energy distributions (SEDs) are computed
for different viewing angles and used to construct multi-band light curves. We employed
the new K N model grid (Anand et al. 2023) of SEDs pre-computed with the improved
https://observing.docs.ligo.org/plan/.
https://github.com/mbulla/kilonova_models/.
Springer
11 Page 14 of 41 N.Werner et al.
Fig. 7 A simulated N U V band
QUVIK image (pixel resolution
of 1.75 arcsec p x _ 1
, FoV cutout
of 1.75 arcmin x 1.75 arcmin) of
a kilonova of m^g =21 at an
angular separation of 1.5
effective radii (7?e) from the
centre of its host galaxy at a
distance of 200 Mpc with an
exposure time of 1000 s. The
simulation assumed the star field
and the host galaxy of
AT2017gfo (see Soares-Santos
et al. 2017)
version of the code, P O S S I S 2 . 0 (Bulla 2023). The model describes axially symmetric
two-component ejecta (high-velocity lanthanide-rich dynamical ejecta component close
to the merger's equatorial plane, lanthanide-poor material closer to the orbital axis, and
a disk-wind component at lower velocities) and from these SEDs, we obtained the simulated
distribution of the K N light curves in the SDSS u filter (York et al. 2000). In this
improved P O S S I S 2 . 0 code, the nuclear heating rates, thermalisation efficiencies, and
wavelength-dependent state-of-the-art opacities taken from Tanaka et al. (2020) depend on
the local properties of the ejecta and time. The two-component ejecta model is simulated
for 11 different viewing angles @0 bs (equally spaced in cos(@0 bs) from a face-on/jet axis to
the edge-on/merger plane) and is characterised by five parameters: the mass of the dynamical
ejecta Me j,d yn = [0.001, 0.005, 0.010]MG , the averaged velocity of the dynamical ejecta
^ej.dyn = [0.15,0.20,0.25]c, the averaged electron fraction of the dynamical ejecta 7e,dyn =
[0.15, 0.20, 0.25], the mass of the disk-wind ejecta M e j j W i n d = [0.01,0.05,0.09,0.13]MG ,
and the averaged velocity of the disk-wind ejecta vej,wind = [0.05, 0.10, 0.15]c (for details
see Anand et al. 2023). The median absolute peak A B magnitude is —17.4 (90% confidence
interval is between —14.9 and —18.6 mag).
Note that the opacities from Tanaka et al. (2020) are computed up to the ionisation stage
of IV, which is valid for ejecta temperatures roughly below 20 000 K. At early times (earlier
than about 0.5-1 days after the merger), the temperatures in the ejecta are higher and the
material can be ionised to higher stages. The lack of opacity contribution from elements in
higher ionisation stages in the modelling might result in an overestimated brightness in the
early U V light curves (Bulla 2023) by ~ 1 mag when compared to models including contributions
up to ionisation stage X I (Banerjee et al. 2023). Note, however, that this model does
not include possible additional brightening from the free neutron beta decay and the shockcooling
of the material surrounding the merger remnant, which, as discussed in Sect. 3.3,
might increase the early U V luminosity of KNe by as much as 2 magnitudes.
Springer
Science with QUVIK I Page 15 of 41 11
-10 -11 -12 -13 -14 -15 -16 -17 0 100 200 300 400 500
Peak Absolute Magnitude Luminosity distance (Mpc)
Fig. 8 Left panel: Cumulative peak luminosity distribution in u filter of kilonova GW170817 / AT2017gfo,
KNe candidates GRB 130603B, GRB 050709, GRB 060614 including GRB 150101B, GRB 050724A, GRB
061201, GRB 080905A, GRB 150424A, and GRB 160821B regarded as KNe events, i.e. the normalized
number of events with peak luminosity lower than the given value. The blue solid, dashed, and dotted lines
mark the median, the upper, and the lower limits of the distribution. The vertical black dotted line marks
the QUVIK limiting A B magnitude of 22 of a K N placed at the distance of 200 Mpc (adpoted with modification
from Ascenzi et al. 2019). Right panel: Apparent magnitude of KNe vs distance in u filter. The red
curve marks the range of the peak apparent magnitude from kilonova GW170817/AT2017gfo as modelled by
Nicholl et al. (2021) in the M . Blanco DECam u filter (Flaugher et al. 2015) if the K N is placed to a different
distance. The yellow curve marks the range of confirmed KNe and KNe candidates using the lower and upper
limit of the peak absolute magnitudes shown in the left panel adopted from Ascenzi et al. (2019). The black
curve denotes the median and the 90% CI of the simulated apparent AB magnitudes in the SDSS u filter (York
et al. 2000) obtained from the nucleosynthesis-powered KNe spectral energy distributions pre-computed by
the M C radiative transfer code POSSIS 2 . 0 (Anand et al. 2023; Bulla 2023)
Having estimates for the peak absolute magnitude distribution of the K N emission, one
can plot the apparent magnitudes for different distances and compare them with a given
detection threshold. This is presented in the right panel of Fig. 8.
The left panel of Fig. 9 shows a similar simulation of nucleosynthesis-powered K N light
curves with the P O S S I S 2 . 0 code for the 260-360 nm band assuming the preliminary
mean optical throughput of the telescope in the N U V band (the exact coatings and filters
and hence the throughput of the telescope as a function of wavelength will be determined
in the future development phases) and the quantum efficiency QE(v) function of the Gpixel
GSENSE4040BSI sensor, which is a candidate for QUVIFs N U V detector (see Sect. 2.3.1).
For this band and sensor, the median absolute peak A B magnitude is —17.3, and the 90%
CI is between -14.4 and -18.6.
3.2.2 Detectable Kilonovae Rate
There is considerable uncertainty in the estimated B N S merger rate in the local Universe.
Recent estimate of BNS coalescence rate density is 320^40 G p c - 3
y r - 1
(Abbott et al. 2021),
which corresponds to a coalescence rate of 1 l ^ g 6
y r - 1
up to a distance of 200 Mpc. The BNS
merger detection rate expected during the fifth observing run (05) of the full L V K G W network
is 190+tao
yr 1
(90% credible interval, Petrov et al. 2022). To provide an estimate
for the number of KNe detectable by QUVIK, we assume purely nucleosynthesis-powered
KNe and the predictions for BNS mergers observed by the full L V K G W network in the 05
observing run. We simulated ~ 100 million K N e uniformly distributed in a sphere with a
radius of 1 600 Mpc (viewing angles were equally spaced in cosine and the ejecta model parameters
were sampled as described in Sect. 3.2.1). Then, we used the distribution of ~ 3600
Springer
11 Page 16 of 41 N.Werner et al.
0 1 2 3 4 5 0 1 2 3 4 5
t-tmerger (days) t - t m e r g e r (days)
Fig. 9 Left panel: Around 3600 simulated KNe light curves (blue curves) viewed by the QUVIK telescope in
the 260-360 nm band in absolute magnitudes in AB magnitude system using the quantum efficiency QE(v)
of the Gpixel GSENSE4040BSI sensor. Light curves were obtained from the nucleosynthesis-powered KNe
spectral energy distributions pre-computed by the 3D M C radiative transfer code POSSIS 2 . 0 (Anand et al.
2023; Bulla 2023). No other components, such as heating from the free neutron decay or cocoon emission,
were added. Right panel: A sample of about 90,000 simulated KNe light curves (blue curves) displayed in
apparent AB magnitudes randomly drawn from the distribution shown on the left panel and randomly placed
(following uniform number density) in the volume up to the distance of 1600 Mpc. A limiting AB magnitude
of 22 is marked by the dotted line. The simulations reveal that the fraction of detectable nucleosynthesispowered
KNe for the limiting AB magnitude of 22 (23) is 16% (44%)
absolute peak A B magnitudes of simulated K N e with the P O S S I S 2 . 0 code as shown in
Fig. 9 to randomly assign to each of the ~ 100 million K N e an absolute peak A B magnitude.
Next, for a given limiting magnitude threshold, we calculated the fraction of simulated
KNe /det.EM for which the E M counterpart would be detectable by QUVIK. The simulations
were performed in the 260-360 nm band employing the quantum efficiency QE(v) of the
Gpixel GSENSE4040BSI sensor. Next, we assumed the B N S merger detection rate foreseen
for the 0 5 run by the full L V K G W network 190^ yr"1
(90% credible interval) as
mentioned above by Petrov et al. (2022). One has to consider that only a fraction / B N S . K N
of BNS mergers will produce a K N ; for the remaining fraction, the merger will result in a
prompt black hole collapse without a disk or ejecta. We adopt the value of / B N S / K N = 0.78
(Colombo et al. 2022). The last two factors are the duty cycle fa of the telescope and the
fraction of the sky visible by QUVIK. We assume a 50% duty cycle. As for the sky visibility
fraction, we assume / s k y = 0.8. Thus, the predicted number of detectable K N e is calculated
as AfjOxKyr-1
) = A^BNs(yr _ 1
) /det.EM / B N S / K N / d / s k y
Our simulations reveal that the fraction of detectable nucleosynthesis-powered K N e for
the limiting A B magnitude of 22 is 16% in the 260-360 nm band. This fraction increases to
44% if we target the limiting A B magnitude of 23, which will require longer exposures and
at least two orbits with QUVIK (see Sect. 2.3.1). The right panel of Fig. 9, displays ~ 3600 of
simulated nucleosynthesis-powered K N e light curves in apparent A B magnitudes randomly
distributed in a volume up to the distance of 1600 Mpc. The plot indicates that more efficient
probing of nucleosynthesis-powered kilonova light curves will require deeper multi-orbit exposures,
targeting the A B magnitude of 23. We estimate that for the limiting A B magnitude
of 22, the expected rate of detectable K N e is 9.3_j?°4 y r - 1
. Similarly, by using the B N S coalescence
rate of l l _ g 6
y r - 1
, as mentioned at the beginning of this subsection, we expect
the rate of detectable K N e up to the distance of 200 Mpc to be 3.2+24 yr"1
(3.1+23 yr"1
)
for the limiting A B magnitude of 22 (21). The estimates are summarised in Table 4, which
provides the expected detectable KNe rates for a number of limiting magnitudes.
Springer
Science with QUVIKI Page 17 of 41 11
Table 4 Predicted detectable
KNe rate by QUVIK in the
260-360 nm band following GW
events from BNS mergers during
the L V K 05 observing run
(Petrov et al. 2022) for different
limiting magnitudes
Limiting A B magnitude
22.0
21.5
21.0
Rate ( y r - 1
)
22.5
23.0
Note that the simulated U V light curves may be overestimated in brightness by ~ 1 mag
in the early time as mentioned above; however, also note that the light curves did not account
for the additional emission from the free neutron decay, which may increase the N U V
luminosity during the first few hours by a magnitude or more (Metzger 2019). The simulations
did also not include possible additional brightening from the shock-cooling of the
material surrounding the merger remnant, which can be an important source of early emission
in N U V and F U V The expected constraints on these additional emission components
are discussed in the next section. Given that no early emission of a K N has been observed
since GW170817, every observation by QUVIK will be highly valuable.
3.3 Constraints from UV Light Curves of Kilonovae
While the red and infrared emission components of the kilonova AT2017gfo were most
likely produced by the radioactive decay of newly produced r-process nuclei, there is a debate
about the observed U V and blue emission, which was dominant in the first ~1.5 days.
The observed early blue emission is well-explained by radioactive material with a relatively
low opacity ("blue" K N ; Metzger et al. 2010). However, the implied large quantity of lowopacity
fast material is not predicted by the simulations of dynamical or disk ejecta. A possible
explanation is that the merger did not immediately result in a black hole but produced
a short-lived magnetar, which is responsible for the large amounts of fast ejecta with a high
electron fraction (Metzger et al. 2018). This has recently also been supported by general relativistic
magnetohydrodynamic numerical simulations (e.g. Curtis et al. 2024; Combi and
Siegel 2023). Very early U V emission could also arise from shock interaction, so-called cocoon
emission (Nakar and Piran 2017a), or the beta decay of free neutrons (Kulkarni 2005).
U V observations obtained early after the neutron star coalescence will be able to distinguish
between the various models.
Piro and Kollmeier (2018) suggested that the power-law evolution of the luminosity during
the early time of AT2017gfo could be explained by the cooling of shock-heated material
around the neutron star merger. This heating could be the result of the interaction of the
gamma-ray burst jet with the merger debris, the so-called cocoon emission theoretically
predicted by Nakar and Piran (2017a) and Gottlieb et al. (2018). Possible additional brightening
of a K N from the shock-cooling of the material surrounding the merger remnant can
be an important source of emission in N U V and FUV. The peak luminosity can be higher
by 2 mag (Kulkarni et al. 2021) compared to the purely nucleosynthesis-powered K N and
would be important in the first several hours after the merger. It was suggested that the cocoon
emission is likely the first UV/optical signal that can be seen when the jet responsible
for the short G R B , which follows the BNS merger, is viewed off-axis (Kulkarni et al. 2021).
Free neutrons, if present in the outer ejecta layers, would decay as n° -> p+
+ W~ —»
p+
+ e~ + ve with a half-life of ~15 minutes, providing extra heating and enhancing the
early K N emission (Kulkarni 2005; Metzger et al. 2015) which, as indicated in Fig. 10,
Springer
11 Page 18 of 41 N.Werner et al.
Fig. 10 Left panel: Model of a red K N (lanthanide-rich) including the emission from free neutron decay
("neutron precursor" emission, solid curve) from the outer layers of the ejecta containing the mass of neutrons
M n = 1 0 - 4
MQ . Right panel: Similar to the left panel, but for a blue K N (lanthanide-poor). Both models were
calculated for the same total ejecta mass M — 1 0 - 2
M© and velocity VQ — 0.1c. For comparison, the dashed
curves are the models without free neutron decay emission. From Metzger (2019)
would increase the N U V luminosity during the first few hours by a magnitude or more
(Metzger 2019). In the F U V band, the free neutron decay would increase the luminosity of
KNe by as much as two magnitudes (Kulkarni et al. 2021). The predicted free neutron decay
in a K N explosion has not yet been observed and early photometry by QUVIK, performed
less than about 6 hours after the merger, would thus provide critically important data.
Early observations in N U V and F U V with QUVIK, in the first few hours after the B N S
merger, will be crucial for distinguishing between models. Dorsman et al. (2023) performed
a Bayesian analysis to determine whether a U V satellite could distinguish between the physical
processes driving the early blue emission component. They show that if the satellite
starts collecting data early, within a couple of hours, it will be able to distinguish between
early radiation models. In particular, probing the beta decay of free neutrons requires light
curves taken less than 6 hours after the B N S merger (see also Kulkarni et al. 2021). Dorsman
et al. (2023) also show that having simultaneously taken U V and optical data improves
the constraints on models significantly. Given the planned extensive ground-based follow-up
efforts, the availability of complementary optical and NIR data is likely.
The ratios between the K N fluxes obtained in the UV, optical, and NIR bands will allow
us to identify and constrain the properties of the different ejecta (Metzger 2019). In
particular, the U V observations performed in the first few hours of the K N will probe the
fastest ejecta's mass, composition, and thermal content and allow us to constrain its geometry,
quantity, and kinematics. Figure 11 shows that early multi-wavelength observations
will also allow us to determine the nature of the merger product, which can be a hypermassive
neutron star that quickly collapses into a black hole or a stable, rapidly spinning,
highly magnetised neutron star. Alternatively, a BNS merger can result in a direct collapse
into a black hole. Each of these outcomes produces different ratios of observed UV, optical,
and NIR fluxes after the merger (Kasen et al. 2015; Fernandez and Metzger 2016) and if
a long-lived stable magnetar is formed, the K N signal might be drastically different (e.g.
Bucciantini et al. 2012; Yu et al. 2013; Metzger and Piro 2014). Importantly, since the gravitational
wave observations allow us to obtain accurate measurements of the mass of the
binary, by ascertaining the outcome of the merger from the early electromagnetic signal, we
can, in principle, constrain the equation of state of neutron stars (e.g. Bauswein et al. 2013;
Margalit and Metzger 2019).
Springer
Science with QUVIKI Page19of41 11
UV (n-precursor) optical (disk wind) infrared (disk wind + dynamical)
bo
CD
c
Fig. 11 Illustration of various scenarios of the NS-NS or NS-BH merger (indicated in the top panel), the resultant
remnant (middle panel) and the produced relative amount of UV/blue emission from a neutron precursor
(purple), optical emission from the lanthanide-free material (blue) and IR emission from the lanthanide-rich
ejecta (red), t-^s denotes the time after which the hypermassive neutron star (HMNS) merger remnant collapses
to a black hole. The Greek letter £2 denotes a spinning magnetized neutron star and a rapidly spinning
black hole remnant (from Kasen et al. 2015)
4 Gamma-Ray Bursts
4.1 GRB Physics
Gamma-ray bursts (Piran 2004; Zhang and Meszaros 2004; Meszaros 2006; Kumar and
Zhang 2015) are some of the most luminous explosions in the Universe. They are traditionally
divided into two classes, defined by their observed duration in soft y-rays: long GRBs
with T 9 0 >2 s and short GRBs with T 9 0 <2 s, where T 9 0 is the time in which 90% of the soft
/-rays are emitted. Although this division is somewhat arbitrary, the two classes are indeed
connected to two different progenitor systems: Most long GRBs are due to the gravitational
collapse of stars with masses reaching tens of MQ and are accompanied by broad-line Ic
SNe (e.g. Hjorth et al. 2003; Cano et al. 2017). Short GRBs, as has been discussed in the
previous section (Sect. 3), originate from the merger of neutron stars and are followed by
the K N emission. However, there are also cases of merger progenitors for seemingly long
GRBs (Rastinejad et al. 2022) and collapsar progenitors for short GRBs (Ahumada et al.
2021; Rossi et al. 2022).
The progenitors of both short and long GRBs produce well-collimated ultra-relativistic
jets, where collisions of shells with different Lorentz factors are responsible for the so-called
"prompt" y-ray emission (see Fig. 12). When the jet interacts with the interstellar medium, it
slows down and produces the GRB afterglow emission detectable in a wide range of frequencies
from radio to high-energy y-rays. The afterglow has a simple synchrotron spectrum
characterised by a three-fold broken power law that breaks at characteristic frequencies: at
the self-absorption frequency va, the typical frequency v{ and the cooling frequency vc (Sari
et al. 1998). The later afterglow is located in the "slow cooling regime", where v{ < vc and
the NIR-optical-UV bands are usually located on the same slope above or below the cooling
frequency. Figure 13 shows a schematic spectrum of a G R B afterglow and the location of
different wavebands, including the bands envisioned for QUVIK and ULTRASAT.
Springer
11 Page 20 of 41 N.Werner et al.
Fig. 12 Sketch of the processes leading to a GRB, its prompt emission and afterglow (from Gehrels et al.
2002)
Fig. 13 Model of a GRB
afterglow in the slow cooling
regime and four observing
spectral regions: purple denoting
the QUVIK FUV band; blue and
light blue indicating the QUVIK
and ULTRASAT NUV bands, and
green for the range 400-1000 nm
observable by ground-based
telescopes. Based on Sari et al.
(1998)
io4
b
1
F
i
\ V - ( P - 1 K 2
1
1 •
slow cooling
»«0 .
• f
X3
LL
10°
' / E
t°
V
H
io"2
•
V
a
. i ^
V
m
1
V
c
10" 10' 10" 10"
v(Hz)
10' 10'
The study of slopes, break frequencies, and their evolution depends on the micro- and
macro-physics of the explosion and its environment. These simple power-law spectra are
altered by the material within the host galaxy and the material between the host and observer.
In particular, dust extinction can introduce a curvature in the spectrum and significantly
attenuate the U V emission.
Long G R B afterglows are on average more luminous than those of short GRBs (Kann
et al. 2011). While short GRBs have only been detected up to a redshift of 2.2 (Seising et al.
2018), long GRBs have been found out to z ~ 9.4 (Cucchiara et al. 2011). However, above
z ~ 2 they are no longer observable in U V due to the L y a dropout, hence G R B science in
the U V band focuses predominantly on GRBs below redshift ~ 2.
Springer
Science with QUVIKI Page 21 of 41 11
4.2 The Science Potential of UV Observations
4.2.1 UV Data of GRB Afterglows
So far, essentially all U V data of G R B afterglows have come from the Swift satellite (Gehrels
et al. 2004). Swift is equipped with the Burst Alert Telescope (BAT; Barthelmy et al. 2005),
X-Ray Telescope (XRT; Burrows et al. 2005a), and the U V and Optical Telescope (UVOT;
Roming et al. 2005). U V O T is a 30 cm telescope with six different filters ranging from 170
to 600 nm (uvw2, uvm2, uvwl, u, b, v). When a G R B is detected by BAT, the satellite
slews so that the burst gets in the F o V of X R T and UVOT. The nominal slewing time is
~ 100 s and most GRBs are observed within the first 150 s, but the reaction time can be
as fast as 40 s (Roming et al. 2009). U V O T is also prepared to receive G R B triggers from
other missions and initiate follow-up. Other U V observatories in orbit are U V I T on AstroSat
and WFC3/STIS on HST (Woodgate et al. 1998; Kimble et al. 2008). However, their fast
response capability is limited.
G R B observations by U V O T have been published in two catalogues: Roming et al. (2009)
and Roming et al. (2017) which include data between 2004 and 2010. About half of the
observations show a simple power-law behaviour for the temporal slope, but ~ 20% show
clear flares or unusual deviations from a (broken) power-law. Oates et al. (2012) found a
correlation between the peak magnitude in U V and the temporal decay slope where GRBs
with higher peak luminosities decay faster than those with lower luminosities. The most
likely interpretation is that this is not an intrinsic property, but a viewing angle effect, since
a similar correlation is also observed in X-rays (Oates et al. 2015). U V O T observations
have been an integral part of the G R B afterglow data for getting very early observations
and contributing to the broad-band spectral energy distributions across the electromagnetic
spectrum.
4.2.2 Early GRB Data: Flares and the Onset of the Afterglow Probing Jet Physics
U V O T has been vital for many years to observe GRBs at very early times, sometimes even
during the prompt emission, observing the rise to maximum light and subsequent brightness
decay. In recent years, rapidly slewing robotic optical telescopes have also been able to
observe the transition between the prompt and the afterglow emission (Ror et al. 2023;
Greiner et al. 2009; Rykoff et al. 2009; Racusin et al. 2008; Vestrand et al. 2005).
At early times, the afterglow evolution often differs from a simple power-law decay
showing features such as bright flares, which can be brighter than the prompt emission,
breaks in the decay, or rebrightenings at later times (see also Fig. 14). Flares have been observed
in X-rays (Margutti et al. 2011; Y i et al. 2016) and U V by U V O T (Zaninoni et al.
2013; Swenson et al. 2013; Y i et al. 2017). Their origin is still not fully understood; models
include a revived central engine activity (Burrows et al. 2005b), interactions with the
interstellar medium (ISM) which might be stratified from previous mass ejections of the
progenitor star (Ayache et al. 2020), or reverse shocks that are created when the shock front
hits the I S M (Lamberts and Daigne 2017). These processes might all be acting at different
times, e.g. I S M interaction might explain later flaring, while the central engine activity
might cause the earlier flares.
Thanks to modern C M O S BSI sensors, QUVIK will have better sensitivity than U V O T
and it will also observe in the F U V (see Sect. 4.3). This will allow us to sample the temporal
evolution of the colour index of flares, brakes, and rebrightenings. The colour index allows
to discriminate between different models for the origin of flares from either central engine
Springer
11 Page 22 of 41 N. Werner etal.
time from trig |s|
FRAM Unfiltered
FRAM R-band
BOOTES-3 Unfiltered
BOOTES-5 Unfiltered
GCN
pulsel
pulse2
pulsel + 2
- 17
0.01
1000 10000
Time post trigger [s]
Fig. 14 Left: UVOT light curves in two filters for two representative GRBs on the bright and faint ends of the
distribution (Roming et al. 2017, data taken from the second UVOT catalogue). Right: Light curve of GRB
190819 showing two prominent flares at early times (from Jelínek et al. 2022)
activity or the ambient medium (Gao et al. 2013). We will also be able to build up a larger
sample of features at different times in simultaneous bands to probe the different scenarios.
The understanding of the physics of this emission requires close coordination with longer
wavelength observations in the optical and NIR. Ground-based facilities such as G R O N D at
the 2.2 m telescope in L a Silla, a 7-channel simultaneous camera (Greiner et al. 2008), have
proven extremely important to obtain dense, simultaneous light curves and thus probe the
evolution across different bands (Greiner 2019).
Only a subset of models is able to explain colour changes (variations of colour index):
mostly those employing central engine variability. A case of X R F 071031 is dicussed in
detail in Kruhler et al. (2009).
Observing the very early phases of GRBs, possibly including flares related to the prompt
emission, would be extremely beneficial. ULTRASAT, covering ~ 0.5% of the entire sky in
a single pointing, may detect one G R B per year directly within its FoV just by coincidence.
Due to its smaller FoV, accidental detection by QUVIK is unlikely. Still, an early simultaneous
two-band observation of a G R B afterglow will benefit greatly from the presence of the
onboard G R B detector with localisation capabilities (see Sect. 2.3.2), which will reduce the
reaction time to a few minutes. Furthermore, QUVIK will also have the ability to perform
rapid follow-up observations of GRBs detected and localised by other observatories.
4.2.3 Redshift Estimates
Astrophysical sources get absorbed below the Lyman limit at 912A, the shortest wavelength
of the Lyman forest at which the electron of a hydrogen atom in the ground state gets ionised.
This can be used to determine photometric redshifts, as with increasing redshift blue bands
subsequently do not detect the object, a technique also called "Lyman dropout" used in a
search for high redshift galaxies. This has also frequently been used for GRBs, in particular
for fast U V O T detections, giving a rough estimate of the redshift very early on.
Considering the current baseline design with the N U V band starting at 260 nm, galaxies
at z > 1.8 would show a loss in flux. A non-detection with QUVIK could imply z > 3.3.
To determine photometric redshifts, usually more bands are required (see e.g. Kruhler et al.
2011); ideally, one needs a further band in the optical/NIR range to derive a secure photometric
redshift, because a decrease in flux might also be due to extinction from the host
Springer
Science with QUVIKI Page 23 of 41 11
galaxy. This is observed to occur for faint bursts at low redshifts. For bright bursts, with a
high S/N, a dropout in one band results in a sharp feature, which cannot be mimicked by
dust. Redshift estimates will clearly benefit from 2-band observations.
4.2.4 Host Galaxy Observations
U V observations of G R B host galaxies are still scarce. To date, only a few G R B hosts have
been observed in U V bands using either SwiftfUVOT, AstroSatfUVTT, or HST. Massive stars
emit a considerable part of their light in the UV, in contrast to older and redder stars and
hence directly probe the star-formation rate (SFR) in galaxies. Particularly for long G R B
hosts as star-forming galaxies, U V data are one of the methods to determine the unobscured
ongoing SFR. The corresponding relation between U V flux in the 150-280 nm rest frame
and the unobscured SFR is described in Kennicutt (1998).
At higher redshifts, the U V restframe gets shifted into the visible band, allowing to probe
the restframe U V SFR. However, for low redshift GRBs (z < 1), U V observations are important
to directly probe the SFR and the U V luminosity function of G R B hosts. To date, this
has only been done by extrapolating S E D fits to galaxy templates obtained by observations
in optical to NIR bands (see also Fig. 15 or Schulze et al. 2015).
A n important goal for a U V space observatory, such as QUVIK, is to build up a sample
of low-redshift G R B hosts, something currently lacking in the field. Up to z ~ 0.5 this is
feasible for a significant part of the hosts, as can be seen from the QUVIK limits indicated
in Fig. 18. For fainter galaxies, deeper observations, reaching 1-2 mag fainter, would be
beneficial. For the brighter galaxies and/or very nearby GRBs, late observations of the host
will be needed in any case to subtract the host galaxy background (see Sect. 4.4.1).
4.2.5 Cocoon Emission of Long GRBs
The ejection of material from the interior of the star, swept up by the G R B jet, is responsible
for the so-called cocoon emission (Nakar and Piran 2017b). This emission is produced
when the shock breaks out from the star, interacts with the surrounding medium and deposits
energy. The exact strength depends on the mixing between the shock and the external
medium.
This phenomenon was detected for the first time in G R B 171205A (Izzo et al. 2019, see
also Fig. 15), which was a low luminosity burst with a relatively weak afterglow. This G R B
occurred at an unusually small distance of only 163 Mpc, which was crucial to observe this
feature. The cocoon was observed less than a day after the G R B as the additional blackbody
emission rapidly cooled and faded. Spectroscopic observations revealed material at
very high speeds (1/3 of the speed of light) and rich in Fe and Ni, indicating it had to come
from the very inside of the star.
The cocoon emission is expected to be about two magnitudes brighter in U V compared
to optical wavelengths (Nakar and Piran 2017b). It can also be observed at larger angles
from the axis of the G R B jet, which would allow large F o V missions such as ULTRASAT
(Sagiv et al. 2014) to detect such signals even in the absence of a G R B . QUVIK will be able
to follow up ULTRASAT detections in two complementary bands with a narrower PSF and
look for re-brightenings in the U V within a day after a G R B , which could be indicative of a
cocoon emission.
Springer
11 Page 24 of 41 N.Werner et al.
Best model for GRB171205A-Host
(z=0.0368, reduced x2
=2.0)
0-01 01 10 10.0 lew 1*15 lel« lel7 lel8
Time from GRB trigger (days) Frequency (Hz)
Fig. 15 Top: SED fit of the host of GRB 171205A (distance 163 Mpc) including observed U V data from
HST at the blue end (from de Ugarte Postigo et al. subm.). Bottom: (a) Evolution of the multi-wavelength
light curve of the transient following GRB 171205A. (b) Color image of the host galaxy of GRB 171205A
with the GRB/SN present, (c) Emission from the cocoon in the long GRB 171205A visible as an additional
blue component at t ~ l day (from Izzo et al. 2019)
Springer
Science with QUVIK I Page 25 of 41 11
enery {kcVj
Fig. 16 Left: Sample of GRBs with observations in the optical-UV and X-ray wavelengths, and fit to the
spectral slopes for the two regimes (from Schady 2017). Right: SED evolution of GRB 100418A from X-ray
through UV and optical bands. The cooling break is shifting from the FUV towards the optical bands during
the observations (from Nardini et al. 2014)
4.3 FUV Observations of GRBs
To date, there have been no F U V observations of GRBs due to the lack of F U V observatories
that allow for ToO observations; hence there is a gap of unknown SED between U V and X rays
(see Fig. 16). In the slow cooling regime, described in Sari et al. (1998), the cooling
frequency can lie either in the X-ray regime or in the gap and is expected to evolve towards
lower frequencies (longer wavelengths) as t~0 5
(see Fig. 12 and Sari et al. 1998). Due to
the Lyman limit, the exact evolution of this gap is difficult to observe and follow, and its
presence is inferred by a mismatch of the spectral slope between X-rays and UV-opticalNIR
(see Fig. 16). However, these studies are limited to GRBs below z ~ 0.5, above which
they will start missing flux due to Lyman dropout.
The F U V channel onboard QUVIK will thus have an impact on several of the abovementioned
science areas:
• Improve the determination and evolution of the cooling break for low-redshift GRBs by
extending the optical-UV SED further into the FUV.
• Observe early afterglow features of low-redshift GRBs such as flares or re-brightenings
and better determine their behaviour in different bands; look for colour changes or a
gradual shift towards shorter wavelengths.
• Improve the photometric redshift estimates, which is especially crucial at low redshifts
(see e.g. Kruhler et al. 2011) and disentangle the effect of dust from the lack of flux due
to the Lyman limit.
• Observations of galaxy hosts in F U V do not exist so far but GALEX observations of
nearby field galaxies have proven to be valuable as indicators of unobscured star formation
(Morrissey 2006). The F U V to N U V slope can also give an independent value for the
extinction in case the stellar population of the galaxy is sufficiently known.
• Allow for very early observations of the cocoon emission from the interaction of the jet
and the surrounding medium.
Springer
11 Page 26 of 41 N. Werner etal.
70
to
50
>
50
& 40
-i
g 30
r
jj 20
10
c
5 point LC
single detect.
24 23 22 21 20 13 1S 17 16
60
£ 50
>•
•S 30
e
2 20
10
5 point LC
single detect.
24 23 22 21 20 19 18 17 16
12.5 15.0 17.5 20 0 22.5 25.0 27.5 30.0
UVW1 mag
Fig. 17 Top: Detection rate of GRBs in Swift/UVOT u and uvwl bands. Bottom: Brightness distribution for
the average of the first five detections by UVOT split into the brightest, average, and faintest detections. All
data are based on the second Swift data release (Roming et al. 2017)
4.4 Detectability of GRBs with QUVIK
We estimate the number of afterglows that could be observed by QUVIK based on data from
Swift/UVOT published in the second U V O T catalogue (Roming et al. 2017). This catalogue
contains data and fitted parameters for 538 bursts over the first 6 years of Swift observations.
To estimate the sensitivity of QUVIK, we use 2 bands based on U V O T filters:7
u filter centred
at 346 nm and uvwl filter centred at 260 nm. Figure 17 shows the average number of bursts
per year where either one observation or five data points reached a specific magnitude limit
based on simple statistical calculations. About 6% of the light curves in the catalogue have
more than 40 points with better than 20% precision in the uvwl band.
Compared to Swift/UVOT, QUVIK will have the advantage of simultaneous observations
in two bands. The on-target slew time will most likely be longer than for Swift, however, the
onboard G R B detector would reduce the reaction time significantly. For U V O T data in the
uvwl band, 35% of light curves started < 15 minutes and 55% in < 1 hour after the trigger.
Due to a typical power-law decay, the usual observation strategy of U V O T (similar to
many ground-based follow-ups) is to increase the length of exposures as the afterglow becomes
fainter. Grouping available data by exposure times from the two catalogues, we can
derive the detection limits for the u, uvwl and uvw2 filters: half of the objects detected in
a 500 s exposure lies above 22.0, 21.8 and 22.3 mag, respectively; changing the percentile
to 90% of detections (still 500 s exposure), the limits change to 23.8, 23.6 and 24.4 mag,
respectively. For a 20-minute exposure, the median detected magnitude is 22.6, 22.5 and
23.0, respectively; the faintest 10% of the detections fall below 24 mag for all filters.
7
https://www.swift.ac.uk/analysis/uvot/filters.php.
Spring
Science with QUVIKI Page 27 of 41 11
Absolute magnitude Absolute magnitude
Fig. 18 Global host galaxy absolute magnitudes extrapolated from SED fits based on optical to M R data.
Dashed lines show the absolute magnitude corresponding to an observed magnitude of 22 in NUV and 19 in
the FUV band at different redshifts
Based on the Swift/UVOT sample, our expected detection rate is ~ 50 events per year
with several data points on the light curve. For pure detection, we expect a rate of ~ 70 per
year, assuming a limiting magnitude of 22.
4.4.1 Host Galaxy Contamination
Contamination by the host galaxy can be an important issue for G R B afterglows. Long GRBs
have a very small offset from their host galaxy with an average offset of 0.6 Re and 80%
being within \Re (Lyman et al. 2017). The average effective radius is only 1.7 ± 0.2 kpc.
Short GRBs have larger offsets of ~ 1.5Re (Fong et al. 2022) but also larger hosts with an
average radius of 3.3 kpc.
To estimate possible contamination, we used SED fitting of a sample of short and long
G R B hosts with data in the optical and NIR and extrapolate the result to the N U V and F U V
bands. The N U V and F U V filters are based on the passbands of the GALEX U V satellite,
where the F U V covers the 134-180 nm range and the N U V spans the 170-300 nm band.
The data of the SED fits are based on the S H O A L S sample for long G R B hosts (Perley et al.
2016) and the samples of Fong et al. (2022) and Nugent et al. (2022) for short GRBs (Agui
Fernandez et al. in prep.).
The distributions for N U V and F U V are shown in Fig. 18. As expected, long G R B hosts
are on average brighter in the U V than short G R B hosts since they are actively forming
stars and the average offset of a long G R B from its host (0.6Re) is relatively small. With
the larger offset for short GRBs, contamination might be less of an issue even at lower
redshifts; however, short G R B afterglows are also on average fainter (Kann et al. 2011) and
hence more often close to our detection limit. These results show that for most low-redshift
GRBs, we will need an additional image of the host galaxy, obtained at a later time, for
image subtraction.
5 The UV Emission of Fast Radio Bursts and Magnetars
Fast radio bursts (FRBs) are extraordinarily bright millisecond transients first observed in
the radio band (Lorimer et al. 2007; Zhang 2020). Some FRBs were found to be repeating
Springer
11 Page 28 of 41 N.Werner et al.
and their association with distant galaxies confirmed their extra-galactic origin. Importantly,
their repeating nature indicates that their progenitors are non-cataclysmic, although it is not
yet clear whether all FRBs are repeating and whether they all have the same progenitors
(Kramer 2023). To date, more than 1000 individual ("one-off) FRBs events and about 50
repeating sources have been detected, but their exact origin remains a mystery (Zhang 2020).
Even though it is not clear if all F R B sources come from a unique population of objects, the
mainstream set of models link FRBs with magnetar activity.
Recently, the soft gamma-ray repeater S G R 1935 + 2154 was observed to be associated
with both an F R B (although its radio luminosity is orders of magnitude smaller than that
of extragalactic FRBs; 2020; Bochenek et al. 2020) and a simultaneous X-ray flare (Ridnaia
et al. 2021; Tavani et al. 2021; Mereghetti et al. 2020; L i et al. 2021), becoming the
first F R B detected in the Milky Way. Soft gamma-ray repeaters are magnetars, neutron stars
with extremely strong magnetic fields (101 3
—101 5
G), showing X-ray and soft gamma-ray
bursts. This observation suggests that they emit the radio and X-ray emissions simultaneously
(CHIME/FRB Collaboration et al. 2022). The F R B emission model of a magnetar
is shown in Fig. 19. Nonetheless, there are also many other models for F R B sources, for
example, coherent radio bursts of relativistic shocks (Platts et al. 2019; Zhang 2020).
Many theoretical models have been proposed to interpret FRBs in various spectral bands
(Platts et al. 2019); however, the exact emission mechanism is so far unknown. U V measurements
can open a unique spectral window to place constraints on the emission properties of
magnetars producing FRBs and to refine theories of their emission mechanisms. Assuming
that FRBs are indeed produced by magnetars, an observation performed during or just after
a soft gamma-ray/X-ray flare could, in principle, place constraints on their physics. Such
a measurement of a U V flux of an F R B counterpart has a major breakthrough potential
because no simultaneous X - r a y - U V or radio-UV detection has been obtained so far.
A potential candidate for observation by QUVIK is the Galactic magnetar S G R 1935 +
2154, which has previously been associated with the above-mentioned low luminosity F R B
(Fig. 20, Ridnaia et al. 2021). This soft gamma-ray repeater in a flaring state will present a
perfect opportunity for long-term monitoring. Even though an F R B is a millisecond event
in radio, magnetar bursts, outbursts, and giant flares can last milliseconds to days (Kaspi
and Beloborodov 2017) in X-rays and can be followed by an afterglow. Thus, the emission
model also predicts a longer-lasting increase in the U V emission. In addition, there
is a known delay in X-ray and radio observations of several milliseconds for the Galactic
magnetar S G R 1935 + 2154 and the associated burst F R B 200428 (Ge et al. 2023).
Springer
Science with QUVIK I Page 29 of 41 11
v(Hz)
1 0 9
1 0 1 0
1 0 n
1 0 l !
1 0 1 3
1 0 " 1 0 l s
1 0 1 6
1 0 " 1 0 1 8
1 0 1 B
1 0 2
°
""I • 1 1 1 1 1 1 1 1
1 1 1 1
1 1 "I I 1 1 0 2
r STARE2
.& - *
j • t J
1 0 ° 1 0 " 1 0 ' 1 0 ° 1 0 s
1 0 " 1 0 s
1 0 s
10' 10° 101
1 0 2
1 0 3
Energy (keV)
Fig. 20 Average spectral distribution of SGR 1935 + 2154, the proposed source of FRB 200428 observed
by the Konus-Wind (KW) instrument. The detected FRB fluxes are shown in observations of CHIME/FRB
(green, orange) and STARE2 (red) (Bochenek et al. 2020). The yellow region denotes the probable detection
interval, assuming the fitting model of the X-ray spectrum. The blue arrows indicate the 5er detection
threshold of a single exposure for exposure times 0.1 s, 1 s, and 10 s. From Ridnaia et al. (2021)
As can be estimated from the spectral fit in Fig. 20, observations could reach intensities
^ 5000 [xJy, ^ 500 [xJy, and ^ 50 [xJy for 0.1 s, 1 s, and 10 s exposure times, providing important
constraints on the emission model. In addition, the magnetar activity time window
can be much longer than the assumed exposure times. For the long-time activity windows,
sources brighter than ~23.5 mag can be detected in up to several hour long exposure times.
Nonetheless, in this case, also other U V satellites such as HST or AstroSat could be used.
Simultaneous U V and radio observations of nearby repeating F R B sources could provide
valuable measurements or upper limits on their U V fluxes. The repeating sources could
be the first targets of F R B observations because of their high repetition level. The nearest
known repeating F R B source in a globular cluster of the M81 galaxy at a distance of 3.6 Mpc
demonstrated remarkable activity when it emitted 53 bursts within only 40 min (Nimmo
et al. 2023).
A U V detection of a magnetar before, during, or after an F R B will help to distinguish
between different emission models. Combining the U V data with those in X-rays allows for
spectral fitting of various emission models because a wider frequency range is covered. For
example, Ridnaia et al. (2021) used a cutoff power law and a sum of two blackbody functions
for fitting the X-ray spectrum. However, more complex profiles may appear when the U V
band is included. Such broadband coverage could reveal whether the radiation is produced
by a thermal or nonthermal emission mechanism or their combination. While blackbody
radiation can be expected to be connected to the neutron star surface (Riley et al. 2019),
nonthermal spectra can originate from magnetospheric plasma (Cerutti et al. 2016). The
study of the emission of magnetars in their quiet state is described in Paper II (Krticka et al.
2024).
6 UV Emission of Supernovae
The physics of explosions of massive stars as supernovae (SNe) remains one of the outstanding
unsolved problems in astrophysics. Early U V observations of SNe present a very effec-
Springer
11 Page 30 of 41 N.Werner et al.
tive tool for studying the radius, surface composition, as well as the surrounding circumstellar
medium (CSM) of massive stars that eventually explode as SNe (e.g. Branch and
Wheeler 2017), providing key initial conditions for computer simulations of the explosive
process. Here, we focus on observations of the initial shock breakout, as well as on type Iln
and super-luminous SNe (SLSNe), which may provide some key ingredients to S N study.
6.1 Supernova Shock Breakout
Observations of the so-called shock breakout, which is usually regarded as the start time of
an SN, require measurements in X-ray and U V wavelengths obtained in the first hours and
up to a day after the explosion and could be triggered by ULTRASAT or by ground-based
wide field surveys. The S N shock breakout emission may provide invaluable information
about the progenitor size and properties of the surrounding C S M , including its asymmetry,
inhomogeneities, asphericity (e.g. Bayless et al. 2022), and chemical composition (e.g.
Nishikida et al. 2006; Yan et al. 2017).
The shock breakout is defined as the moment when the internal S N shock wave emerges
from the progenitor's photosphere (e.g. Nakar and Sari 2010; Waxman and Katz 2017). After
escaping the stellar body and entering the C S M surrounding the progenitor star, the shock
wave accelerates and produces X-rays accompanied by a longer-lasting afterglow in the UV.
As demonstrated by e.g. Chevalier and Fransson (2008) and Soderberg et al. (2008) for S N
2008D, X-ray and U V observations of the early characteristics can be used to determine
the radius of the S N progenitor. A n equally important goal is to map the morphology and
properties of the transition zone between the star and the surrounding C S M . The density and
velocity profiles of this region remain not well-understood. The same applies to the time
variability of these quantities, even if we assume only spherically symmetric explosions.
We expect that U V photometry will contribute significantly to our understanding of the nonstationary,
eruptive pre-explosive ejections and enhanced mass loss.
The current lack of X-ray/UV measurements does not allow us to determine the characteristic
chemical composition and plasma parameters of S N explosions with the required
accuracy. To a large extent, it prevents the proper classification of SNe and the study of the
origin of elements (Waxman and Katz 2017). Namely, in FUV, thousands of spectral lines
of iron group elements are the source of significant absorption. Therefore, the F U V excess
may help us to reveal the abundance of the freshly synthesised iron-peak group elements in
the outer ejecta and relate it to the overall metal abundance of the S N progenitor (Yan et al.
2017).
However, the approach described so far does not include the geometric and physical
complexity imprinted in this early-phase emission (e.g. Nakar and Sari 2010). Namely, it
is becoming increasingly evident that the asphericity of the expanding SNe envelope, the
surrounding C S M , and the star-CSM transition region play a crucial role in the investigation
of the pre-SN and S N phenomena (e.g. Couch et al. 2011; Couch 2017; Janka 2017). Asymmetries
may induce a great number of secondary shocks of various strengths and directions
and thus significantly alter the timescale of the shock breakout and the related high-energy
afterglows (e.g. Matzner et al. 2013; Fryer et al. 2020; Irwin et al. 2021). Moreover, the multiplicity
in time of such shocks can dramatically change the overall observational picture of
SNe. The shocks may thus help to model and understand the conditions and processes in the
progenitor star that may have led to the overall mass, temporal variability, and spatial complexity
of the C S M ejection. The complexity of the measurement is, of course, enhanced by
the fact that the emission resulting from the S N - C S M interaction takes place in many wavebands.
Thus, a thorough examination and comparison of observations performed in different
wavelengths will help us to decipher the processes behind this emission.
Springer
Science with QUVIKI Page 31 of 41 11
Fig. 21 Illustration of early UV emission models of SN compared with observations. Models of Sapir et al.
(2013)—red supergiant (RSG; dashed red line) and blue supergiant (BSG; solid blue line); Nakar and Sari
(2010)—BSG, dashed blue line; Rabinak and Waxman (2011)—RSG, solid red line, Wolf-Rayet (WR) star
with He, solid yellow line, and C/O dominated WR star, dashed yellow line. Observations from G A L E X / N U V
('supernova legacy survey' — SNLS-04D2dc, type II-R Schawinski et al. 2008; Gezari et al. 2008, red
circles) and Swift/UVOT (SN 2008D, type lb, Soderberg et al. 2008, black circles). Inspired/adapted from
Ganotetal. (2016)
For these reasons, the exact behaviour of U V light curves is somewhat heterogeneous;
their detailed time evolution and luminosity depend both on the above-mentioned characteristics
and on the particular S N type. To determine the common features of the U V luminosity
behaviour, we can take, for example, the models of Rabinak and Waxman (2011), Nakar and
Sari (2010), and others, together with observations of, e.g. Schawinski et al. (2008), Gezari
et al. (2008), Soderberg et al. (2008), shown in Fig. 21. F U V light curves show an initial
rise over a duration R+/c (where R+ is the radius of a progenitor and c is the speed of light),
which corresponds to about 2000 s after the shock breakout for a red supergiant (RSG) explosion
and correspondingly less for a blue supergiant (BSG) progenitor. However, this fast
rise is followed by a slow decay up to typically 14 hrs for an R S G and 0.5-1 hr for a B S G
(assuming that the breakouts from RSGs and most of BSGs are in thermal equilibrium). This
decrease is followed by a second peak, as the "observed temperature" falls into the given
frequency range, typically at 2 days for RSGs, and slightly less than 1 day for BSGs, while
it occurs at a yet earlier time for more compact progenitors (see, e.g. Fig. 4 in Nakar and
Sari 2010). The peaks and profiles of N U V light curves evolve more slowly, on the order of
days after a shock breakout (see Fig. 21). Considering QUVIK's expected slew time of about
20 minutes, observing these light curves after receiving an alert from a wide-field transient
survey (for example, the anticipated alert time of ULTRASAT is 15 minutes; Shvartzvald
et al. 2023) appears feasible.
6.2 Early-to-lntermediate Time Emission of Super-Luminous Supernovae
Important targets for QUVIK are type Iln SNe (Schlegel 1990; Filippenko 1997; Smith
2014; Gal-Yam 2017), which show particularly strong and narrow-to-intermediate Balmer
emission lines. The narrow lines, usually superimposed on much broader lines that manifest
as wings, are formed in the photoionized, relatively slow unshocked material surrounding
the S N before the explosion; they reveal SNe that expand into a dense pre-explosion C S M
(e.g. Chevalier 1982; Moriya et al. 2013; Dessart et al. 2016; Kurfürst et al. 2020). Figure 22
Springer
11 Page 32 of 41 N. Werner etal.
06gyO
10 100 1000
Expansion speed 1/CSM (km s )
Fig. 22 Schematic plot of average mass-loss rate of relevant types of stars with high mass-loss rates versus the
velocity of their winds or eruptive mass-losing events. The coloured regions highlight the values of selected
types of stars or their configurations, that is (from bottom left) of asymptotic giant branch (AGB) and superAGBs
(sAGB), red supergiant (RSG) and extreme RSG (eRSG) stars, yellow supergiants (YSG), yellow
hypergiants (YHG), luminous blue variable (LBV) winds, W N stars with H (WNH), H-free WR stars, binary
Roche-lobe overflow (RLOF), and L B V giant eruptions. Several famous stars with extremely high mass-loss
rates are plotted with larger black solid circles, while the smaller empty circles localize the other well-known
examples of SNe Iln. The two diagonal lines increasing to the upper right are the wind density parameters
(w — Mj VCSMX denoting the typical (solid line) and the lower (dashed line) values required for producing
SNe Iln. Some specific stars may 'sit' below these limits in case of significantly aspherical or clumpy winds,
etc. Inspired by Smith (2017)
illustrates the required mass-loss rates of stellar types that are considered to supply the C S M
by a relevant amount of steady or eruptive mass ejection before the S N event, demonstrating
the importance of the pre-explosion C S M asphericity.
It is still difficult to judge whether S N Iln will manifest as super-luminous supernovae
(SLSNe). The class of S L S N defined by Gal-Yam (2012) exhibits extremely bright and
long-lasting light curves with peak absolute magnitudes of —21 mag or above in the visible
band. Their duration may be hundreds of days or even more than a year. This may also be
accompanied by various bumps or rebrightenings in the bolometric light curve as well as
in particular photometric bands (e.g. Fig. 10 in Kurfiirst et al. 2020). SLSNe are divided
into types I and II according to the presence of hydrogen in the spectrum. Light curves of
SLSNe I usually evolve relatively quickly, with a rise time of several weeks to more than
a month, which is somewhat shorter than in S L S N II, followed by a relatively fast decline.
S L S N I are of significant interest because of their association with L G R B s and KNe, as
well as their large U V luminosities (Branch and Wheeler 2017, see also Fig. 23). Their
extreme luminosities and long durations (in particular for SLSNe II) are mostly interpreted
as a result of interaction with extended and inhomogeneous C S M . Still, other engines, such
as magnetars, might also play a role (Metzger et al. 2017). According to the extent, total
mass, morphology, density distribution, or chemical composition of the C S M surrounding
the explosion, the characteristic narrow Balmer line occurrence may last for days, weeks,
Springer
Science with QUVIKI Page 33 of 41 11
or even years (Soumagnac et al. 2020). Well-documented examples of the long-term objects
of this type may be the prominent SNe: SN1988Z (Williams et al. 2002; Schlegel and Petre
2006; Smith et al. 2017), SN1998S (Fassia et al. 2000, 2001), SN2005gl (Gal-Yam et al.
2007), SN2010mc (Ofek et al. 2013), or SN2010J1 (Gall et al. 2014; Ofek et al. 2014).
Magnetar-driven shocks, which may be important in SLSNe, may differ from the shock
breakout of a 'normal' S N accompanied by a brief X-ray burst (e.g. Klein and Chevalier
1978; Matzner and McKee 1999; Kasen et al. 2016). Since the magnetar-infiated nebula
strongly accelerates the surrounding S N ejecta (Metzger et al. 2017), a weakened
magnetar-driven shock will radiate at longer wavelengths (FUV/NUVVoptical), and its emission
will last longer. The double-peaked early light curve will also indicate the presence of
a magnetar-driven S N due to the persisting injection of energy into the shell. That will produce
an unusually fast shock that can radiate outside the bulk of the S N ejecta before most
of the centrally thermalised energy has had time to diffuse out (Kasen and Bildsten 2010;
Kasen et al. 2016).
A n emerging class of SN-like explosions with luminosities > 104 4
erg s - 1
, which may
even exceed SLNSe, are the "Fast Blue Optical Transients" (FBOTs; Drout et al. 2014;
Prentice et al. 2018; Ho et al. 2020; Metzger 2022). The prototypical member of this class
is AT2018cow, the optical emission of which rose over a few days to a luminosity of L
4 x 104 4
erg s - 1
. The physical nature of this phenomenon, especially in the case of its
exceptionally "luminous" subclass (LFBOTs), could be associated with the tidal disruption
and hyper-accretion of a massive hydrogen-depleted Wolf-Rayet (WR) star by a black hole
or neutron star binary companion. The optical/UV flux tends to be stable for these types of
transients; for example, for AT2018cow, it showed a change of only about 0.1 mag over 1000
days (Metzger 2022). The study of this new class of fast, UV-bright transients is expected
to benefit strongly from the quick 2-band photometry provided by QUVIK.
To summarise, potential science objectives of a small two-band NUV7FUV space telescope
observing the early-to-intermediate time S N emissions include:
• The shocks that propagate in the C S M after the shock breakout (Ofek et al. 2010) are
predicted to radiate mostly in the U V and X-rays (Katz et al. 2011; Murase et al. 2011,
2014; Chevalier and Irwin 2012; Soumagnac et al. 2020). Observation of the explosion
and subsequent S N - C S M interaction at these wavelengths may reveal valuable information
regarding the explosion mechanism, its morphology and chemical composition, and
C S M global properties (e.g. Ofek et al. 2013). Measurement of the luminosity in the U V
band will contribute to a much more precise estimate of the overall bolometric luminosity.
• U V observations will help to determine the C S M geometry. Studies of well-known potential
S N Iln progenitors (e.g. r] Car, Davidson and Humphreys 1997, 2012; Gal-Yam
and Leonard 2009) indicate that S N and C S M asymmetries must be taken into account
to achieve realistic models of observed characteristics in the case of such physically and
geometrically complex processes.
• F U V measurements will provide deep insight into the very early light curves that may
indicate weaker shocks due to the presence of magnetars or reveal a rapidly dimming
phase of the radiative tail closely following the X-ray breakout flash (Kasen et al. 2016).
Furthermore, F U V observations may also help to reveal the chemical composition of S N
ejecta (Yan et al. 2017).
A realistic objective of a small, two-band N U V / F U V space observatory is to provide U V
photometry of SNe in host galaxies within z < 0.1. A real-time alert can be provided by soft
X-ray and U V instruments with very large fields of view, such as the M A X I X-ray monitor
Springer
11 Page 34 of 41 N.Werner et al.
Days from V-band maximum (MJD=57178.5) Days from V-band maximum
Fig. 23 Left panel: Example of late-time SN emission: UVOT light curves of ASASSN-151h (major features
characterise it as SLSN I) in A B magnitudes. The late time rebrightening is brighter than the M < —21
mag cut-off for SLSNe. The readjusted figure is adapted from Brown et al. (2016). Right panel: U V and
V-band light curves of SLSN I Gaialöapd (circles) plotted together with synthetic light curves from models
of SN-CSM interactions (solid lines), reconstructed from Tolstov et al. (2017)
on the ISS8
and ULTRASAT, as well as by ground-based wide-field optical transient surveys
and neutrino observatories.9
6.3 Late Phase of Type II Super-Luminous Supernovae
A n essential characteristic of SLSNe is their high U V luminosity (see Fig. 23). Therefore,
their U V observations are crucial for determining their total luminosities and temperatures.
U V observations of SLSNe II (and SNe Iln, which usually precede them and are related to
them) at later times are important for probing the complex C S M distribution in the wider
surroundings of exploding massive stars. Figure 10 in Kurfürst et al. (2020) compares the
models of asymmetric S N - C S M interactions with the observed bolometric light curves of
selected SNe (Kurfürst et al. 2014; Kurfürst and Krticka 2017; Kurfürst et al. 2018).
Most of the SNe are of type Iln and show a steep initial decline followed by a long slow
decline. Undulations and bumps in their light curves are rare but have been observed in a
few cases (Nyholm et al. 2017), and their exact nature is still debated. The interaction of S N
ejecta with a clumpy C S M (Calderön et al. 2016, 2020) is also expected to produce bumps
in the light curves.
7 Summary
Recent progress in detector and coating technologies, the advent of high-quality off-theshelf
micro/small-satellite platform sub-systems, and the availability of affordable launch
opportunities enable, for the first time in history, also smaller countries and organisations to
deploy sensitive and highly responsive space observatories. The QUVIK mission, described
in Sect. 2, will enable significant progress in the study of explosive transients, as well as in
the field of stellar astronomy (see companion paper II, Krticka et al. 2024) and in the study
of active galactic nuclei (see companion paper III, Zajacek et al. 2024).
https://iss.jaxa.jp/en/kiboexp/ef/maxi/.
https://gcn.nasa.gov/notices.
Springer
Science with QUVIKI Page 35 of 41 11
The mission's primary objective is the follow-up of BNS and N S - B H mergers detected
by gravitational wave observatories and wide-field surveys and their photometry in the U V
band to discriminate between different kilonova scenarios. Their photometry obtained less
than ~6 hours after the merger will provide important constraints on the beta-decay of free
neutrons. Early two-band photometry will also provide key data on shock interactions. The
presence of early emission from beta-decay and shock interactions may increase the early
U V brightness of kilonovae by as much as 2 magnitudes, potentially significantly boosting
their detectable number. The measurements performed by QUVIK will also provide constraints
on the structure and composition of the ejecta, thus probing the contribution of BNS
and BH-NS mergers to r-process nucleosynthesis.
Simultaneous N U V / F U V photometry of G R B afterglows will enable to study and determine
the nature of flares, rebrightenings and plateaus at the onset of the afterglow emission.
The two U V bands will help to better understand the jet physics and the emission mechanism
by searching for colour changes. Two-band U V photometry of the rare cocoon emission feature
from the interaction between the jet and the material of the star predicted by Nakar and
Piran (2017b) and observed by Izzo et al. (2019) will be particularly valuable.
Kilonova and G R B science will greatly benefit from a G R B detector with a fast onboard
localisation capability. Such a detector will enable much-needed early follow-up of G R B
afterglows and, in the case of some long GRBs, the detection of the prompt emission in two
simultaneous U V bands.
A sample of U V observations of kilonova and low-redshift G R B host galaxies, which is
currently lacking, will be valuable for determining the ancestors of the progenitors.
For supernovae, early observations of the shock breakout will provide information about
the progenitor stars and their surrounding C S M . For super-luminous supernovae, later observations
will also probe the C S M and the wider surroundings of massive exploding stars.
Wide-field transient surveys operating or coming online in various wavebands are also
expected to discover new classes of supernovae and other transients. The two-band U V photometry
mission thus provides opportunities for discoveries and follow-up observations
of new types of yet unknown objects and phenomena.
The upcoming democratisation of space activities is expected to result in a growing number
of increasingly capable small satellites complementing large missions. The relative affordability
and potential economies of scale may even result in greater numbers of small
space observatories covering complementary parts of the electromagnetic spectrum or dedicated
to particular targets or classes of objects. They are expected to be especially valuable
for transient astronomy, but other fields will also benefit from the progress. QUVIK may
thus help pave the way for a future multi-wavelength constellation of space observatories
monitoring the energetic universe.
Acknowledgements We thank the Czech Ministry of Transportation and the European Space Agency for
their support of the QUVIK project within the Czech Ambitious Mission Programme. J.B. acknowledges the
support of the German Science Foundation (DFG) project BE 7886/2-1. G L . - G has been supported by the
fellowship Lumina Quaeruntur No. LQ100032102 of the Czech Academy of Sciences. The research of OP
has been supported by Horizon 2020 ERC Starting Grant 'Cat-In-hAT (grant agreement no. 803158). We
thank B. D. Metzger for his valuable comments.
Funding Open access publishing supported by the National Technical Library in Prague.
Declarations
Competing Interests The authors declare no competing interests.
Springer
11 Page 36 of 41 N.Werner et al.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence,
and indicate if changes were made. The images or other third party material in this article are included in the
article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is
not included in the article's Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this licence, visit http://creativecommons.Org/licenses/by/4.0/.
References
Abbott BP, Abbott R, Abbott TD et al (2020) Living Rev Relativ 23:3
Abbott BP, Abbott R, Abbott TD et al (2017a) Astrophys J Lett 848:L13
Abbott BP, Abbott R, Abbott TD et al (2017b) Astrophys J Lett 848:L12
Abbott BP, Abbott R, Abbott TD et al (2017c) Phys Rev Lett 119:161101
Abbott R, Abbott TD, Abraham S et al (2021) Astrophys J Lett 915:L5
Abbott R, Abbott TD, Acernese F et al (2023) Phys Rev X 13:011048
Acernese F, Agathos M , Agatsuma K et al (2015) Class Quantum Gravity 32:024001
Ahumada T, Singer LP, Anand S et al (2021) Nat Astron 5:917
Akutsu T, Ando M , Arai K et al (2021) Prog Theor Exp Phys 2021(5):05A101
Anand S, Coughlin MW, Kasliwal M M et al (2021) Nat Astron 5:46
Anand S et al (2023) submitted to MNRAS 2307.11080
Arcavi I (2018) Astrophys J Lett 855:L23
Ascenzi S, Coughlin MW, Dietrich T et al (2019) Mon Not R Astron Soc 486:672
Asif A, Barschke M , Bastian-Querner B et al (2021) in Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series, Vol. 11821, UV, X-ray, and Gamma-Ray Space Instrumentation for Astronomy
XXII, ed. O. H. Siegmund, 118210U
Atteia JL, Cordier B, Wei J (2022) Int J Mod Phys D 31:2230008
Ayache EH, van Eerten HJ, Daigne F (2020) Mon Not R Astron Soc 495:2979
Banerjee S, Tanaka M , Kato D, Gaigalas G (2023) e-prints, 2304.05810
Barthelmy SD, Barbier L M , Cummings JR et al (2005) Space Sci Rev 120:143
Bastian-Querner B, Kaipachery N, Kiister D et al (2021) In Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series, Vol. 11819, UV/Optical/IR Space Telescopes and Instruments: Innovative
Technologies and Concepts X, ed. A . A. Barto, J. B. Breckinridge, & H. P. Stahl, 118190F
Bauswein A, Goriely S, Janka HT (2013) Astrophys J 773:78
Bayless AJ, Fryer C, Brown PJ et al (2022) Astrophys J 931:15
Bellm EC, Kulkarni SR, Graham MJ et al (2019) Publ Astron Soc Pac 131:018002
Ben-Ami S, Shvartzvald Y, Waxman E et al. (2022) In Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series, Vol. 12181, Space Telescopes and Instrumentation 2022: Ultraviolet
to Gamma Ray, den Herder J-WA, Nikzad S, Nakazawa K (eds), 1218105
Bethe H A (1990) Rev Mod Phys 62:801
Bochenek CD, Ravi V, Belov K V et al (2020) Nature 587:59-62
Branch D, Wheeler JC (2017) Supernova explosions
Brown PJ, Yang Y, Cooke J et al (2016) Astrophys J 828:3
Bucciantini N, Metzger BD, Thompson TA, Quataert E (2012) Mon Not R Astron Soc 419:1537
Bulla M (2019) Mon Not R Astron Soc 489:5037
Bulla M (2023) Mon Not R Astron Soc 520:2558
Bulla M , Camisasca AE, Guidorzi C et al (2023) GRB Coordinates Network 33578:1
Burrows DN, Hill JE, Nousek JA et al (2005a) Space Sci Rev 120:165
Burrows DN, Romano P, Falcone A et al (2005b) Science 309:1833-1835
Calderon D, Ballone A, Cuadra J et al (2016) Mon Not R Astron Soc 455:4388
Calderon D, Cuadra J, Schartmann M et al (2020) Mon Not R Astron Soc 493:447
Cano Z, Wang S-Q, Dai Z-G, Wu X-F (2017) Adv Astron 2017:8929054
Cerutti B, Philippov A A , Spitkovsky A (2016) Mon Not R Astron Soc 457:2401
Chevalier R A (1982) Astrophys J 258:790
Chevalier RA, Fransson C (2008) Astrophys J Lett 683 :L 135
Chevalier RA, Irwin C M (2012) Astrophys J Lett 747:L17
CHIME/FRB Collaboration (2020) Nature 587:54
CHIME/FRB Collaboration, Andersen BC, Bandura K et al (2022) Nature 607:256-259
Springer
Science with QUVIKI Page 37 of 41 11
Collins CE, Bauswein A, Sim SA et al (2023) Mon Not R Astron Soc 521:1858
Colombo A, Salafia OS, Gabrielli F et al (2022) Astrophys J 937:79
Combi L, Siegel D M (2023) Phys Rev Lett 131:231402
Couch SM (2017) In: Alsabti AW, Murdin P (eds) Handbook of Supernovae, p 1791
Couch SM, Pooley D, Wheeler JC, Milosavljevič M (2011) Astrophys J 727:104
Cucchiara A, Levan AJ, Fox DB et al (2011) Astrophys J 736:7
Curtis S, Bosch P, Mösta P et al (2024) Astrophys J Lett 96LL26
Dai L, McKinney JC, Roth N, Ramirez-Ruiz E, Miller MC (2018) Astrophys J Lett 859:L20
Darbha S, Kasen D (2020) Astrophys J 897:150
Davidson K, Humphreys R M (1997) Annu Rev Astron Astrophys 35:1
Davidson K, Humphreys R M (2012) Nature 486:E1
Dessart L, Hillier DJ, Audit E, Livne E, Waldman R (2016) Mon Not R Astron Soc 458:2094
Dewdney PE, Hall PJ, Schilizzi RT, Lazio TJLW (2009) Proc IEEE 97:1482
Dimitrova TA, Butler NR, Ravi S (2023) Astrophys J 949:15
Dorsman B, Raaijmakers G, Cenko SB et al (2023) Astrophys J 944:126
Drout MR, Chornock R, Soderberg A M et al (2014) Astrophys J 794:23
Evans PA, Cenko SB, Kennea JA et al (2017) Science 358:1565
Fassia A, Meikle WPS, Chugai N et al (2001) Mon Not R Astron Soc 325:907
Fassia A, Meikle WPS, Vacca W D et al (2000) Mon Not R Astron Soc 318:1093
Fernandez R, Metzger BD (2016) Annu Rev Nucl Part Sei 66:23
Filippenko AV (1997) Annu Rev Astron Astrophys 35:309
Flaugher B, Diehl HT, Honscheid K et al (2015) Astron J 150:150
Fong W-f, Nugent AE, Dong Y et al (2022) Astrophys J 940:56
Fox DB, Frail DA, Price PA et al (2005) Nature 437:845
Fryer CL, Fontes CJ, Warsa JS et al (2020) Astrophys J 898:123
Fynbo JPU, Watson D, Thöne CC et al (2006) Nature 444:1047
Gal-Yam A (2012) Science 337:927
Gal-Yam A (2017) In: Alsabti AW, Murdin P (eds) Handbook of Supernovae, p 195
Gal-Yam A, Fox DB, Price PA et al (2006) Nature 444:1053
Gal-Yam A, Leonard DC (2009) Nature 458:865
Gal-Yam A, Leonard DC, Fox DB et al (2007) Astrophys J 656:372
Gall C, Hjorth J, Watson D et al (2014) Nature 511:326
Ganot N, Gal-Yam A, Ofek EO et al (2016) Astrophys J 820:57
Gao H, Lei W-H, Zou Y-C, Wu X-F, Zhang B (2013) New Astron Rev 57:141
Ge MY, Liu CZ, Zhang SN et al (2023) Astrophys J 953:67
Gehreis N, Chincarini G, Giommi P et al (2004) Astrophys J 611:1005
Gehreis N, Norris JP, Barthelmy SD et al (2006) Nature 444:1044
Gehreis N, Piro L, Leonard PJT (2002) Sei Am 287:84
Gezari S (2021) Annu Rev Astron Astrophys 59:21
Gezari S, Dessart L, Basa S et al (2008) Astrophys J Lett 683:L131
Gill AS, Shaaban M M , Tohuvavohu A et al (2022) In Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series, Vol. 12191, X-Ray, Optical, and Infrared Detectors for Astronomy X , Holland,
A D & Beletic, J (eds), 1219114
Gottlieb O, Nakar E, Piran T (2018) Mon Not R Astron Soc 473:576
Greiner J (2019) Publ Astron Soc Pac 131:015002
Greiner J, Bornemann W, Clemens C et al (2008) Publ Astron Soc Pac 120:405
Greiner J, Krühler T, McBreen S et al (2009) Astrophys J 693:1912
Hjorth J, Sollerman J, M0ller P et al (2003) Nature 423:847
Hjorth J, Watson D, Fynbo JPU et al (2005) Nature 437:859
Ho AYQ, Perley DA, Kulkami SR et al (2020) Astrophys J 895:49
Ho L C (2008) Annu Rev Astron Astrophys 46:475
Howell SB (2000) Handbook of CCD astronomy
Irwin C M , Linial I, Nakar E, Piran T, Sari R (2021) Mon Not R Astron Soc 508:5766
Ivezič Ž, Kahn SM, Tyson JA et al (2019) Astrophys J 873:111
Izzo L, de Ugarte Postigo A, Maeda K et al (2019) Nature 565:324
Janka H-T (2017) Astrophys J 837:84
Jelínek M , Topinka M , Karpov S et al (2022) Astron Astrophys 662:A126
Jin Z-P, Hotokezaka K, Li X et al (2016) Nat Commun 7:12898
Kann DA, Klose S, Zhang B et al (2011) Astrophys J 734:96
Kasen D, Bildsten L (2010) Astrophys J 717:245
Kasen D, Fernandez R, Metzger BD (2015) Mon Not R Astron Soc 450:1777
Springer
11 Page 38 of 41 N.Werner et al.
Kasen D, Metzger BD, Bildsten L (2016) Astrophys J 821:36
Kaspi V M , Beloborodov A M (2017) Annu Rev Astron Astrophys 55:261-301
Katz B, Sapir N, Waxman E (2011) Proc IAU 7 7(S279):274-281
Kawaguchi K, Kyutoku K, Shibata M , Tanaka M (2016) Astrophys J 825:52
Kawaguchi K, Shibata M , Tanaka M (2018) Astrophys J Lett 865:L21
Keller C, Navarro R, Brandl B (2015) Field guide to astronomical instrumentation, Field Guides (SPIE)
Kennicutt RC Jr (1998) Annu Rev Astron Astrophys 36:189
Kimble RA, MacKenty JW, O'Connell RW, Townsend JA (2008) In Society of Photo-Optical Instrumentation
Engineers (SPIE) Conference Series, Vol. 7010, Space Telescopes and Instrumentation 2008: Optical,
Infrared, and Millimeter Oschmann J, Jacobus M , de Graauw M W M , MacEwen HA (eds), 70101E
Klein RI, Chevalier RA (1978) Astrophys J Lett 223:L109
Korobkin O, Wollaeger RT, Fryer CL et al (2021) Astrophys J 910:116
Kramer M (2023) IAU Symposium 363:72
Krtička J, Benáček J, Budaj J et al (2024) Space Sei Rev 220
Krühler T, Schady P, Greiner J et al (2011) Astron Astrophys 526:A153
Krühler T, Greiner J, McBreen S et al (2009) Astrophys J 697:758
Kulkarni SR (2005) e-prints, astro-ph/0510256
Kulkarni SR, Harrison FA, Grefenstette BW et al (2021) e-prints, 2111.15608
Kumar P, Zhang B (2015) Phys Rep 561:1
Kurfürst P, Feldmeier A, Krtička J (2014) Astron Astrophys 569:A23
Kurfürst P, Feldmeier A, Krtička J (2018) Astron Astrophys 613:A75
Kurfürst P, Krtička J (2017) Appl Math 62:633
Kurfürst P, Pejcha O, Krtička J (2020) Astron Astrophys 642:A214
Lamb GR Tanvir NR, Levan AJ et al (2019) Astrophys J 883:48
Lamberts A, Daigne F (2017) Mon Not R Astron Soc 474:2813
Lattimer JM, Schramm D N (1974) Astrophys J Lett 192:L145
Leinert C (1975) Space Sei Rev 18:281
Levan AJ, Gompertz BP, Malesani DB et al (2023) GRB Coordinates Network 33569:1
Levasseur-Regourd AC, Dumont R (1980) Astron Astrophys 84:277
Li CK, Lin L, Xiong SL et al (2021) Nat Astron 5:378
Li L-X, Paczyňski B (1998) Astrophys J Lett 507:L59
LIGO Scientific Collaboration, Aasi J, Abbott BP et al (2015) Class Quantum Gravity 32:074001
Liran T, Shvartzvald Y, Lapid O et al (2022) In Society of Photo-Optical Instrumentation Engineers (SPIE)
Conference Series, Vol. 12181, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference
Series, den Herder, J-WA, Nikzad, S & Nakazawa, K (eds), 121812X
Lorimer DR, Bailes M , McLaughlin M A , Narkevic DJ, Crawford F (2007) Science 318:777
Lyman JD, Levan AJ, Tanvir NR et al (2017) Mon Not R Astron Soc 467:1795
Margalit B, Metzger BD (2019) Astrophys J Lett 880:L15
Margutti R, Bernardini G, Barniol Duran R et al (2011) Mon Not R Astron Soc 410:1064
Matzner CD, Levin Y, Ro S (2013) Astrophys J 779:60
Matzner CD, McKee CF (1999) Astrophys J 510:379
Mereghetti S, Savchenko V, Ferrigno C et al (2020) Astrophys J Lett 898:L29
Mészáros P (2006) Rep Prog Phys 69:2259
Metzger BD (2019) Living Rev Relativ 23:1
Metzger BD (2022) Astrophys J 932:84
Metzger BD, Bauswein A, Goriely S, Kasen D (2015) Mon Not R Astron Soc 446:1115
Metzger BD, Berger E (2012) Astrophys J 746:48
Metzger BD, Berger E, Margalit B (2017) Astrophys J 841:14
Metzger BD, Martinez-Pinedo G, Darbha S et al (2010) Mon Not R Astron Soc 406:2650
Metzger BD, Piro A L (2014) Mon Not R Astron Soc 439:3916
Metzger BD, Thompson TA, Quataert E (2018) Astrophys J 856:101
Mooley KP, Anderson J, Lu W (2022) Nature 610:273
Moriya TJ, Blinnikov SI, Tominaga N et al (2013) Mon Not R Astron Soc 428:1020
Morrissey P (2006) In Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol.
6266, Space Telescopes and Instrumentation II: Ultraviolet to Gamma Ray, Turner, MJL & Hasinger, G
(eds), 62660Y
Murase K, Thompson TA, Lacki BC, Beacom JF (2011) Phys Rev D 84:043003
Murase K, Thompson TA, Ofek EO (2014) Mon Not R Astron Soc 440:2528
Nakar E, Piran T (2017a) Astrophys J 834:28
Nakar E, Piran T (2017b) Astrophys J 834:28
Nakar E, Sari R (2010) Astrophys J 725:904
Springer
Science with QUVIKI Page 39 of 41 11
Nardini M , Elliott J, Filgas R et al (2014) Astron Astrophys 562:A29
Nicholl M , Margalit B, Schmidt P et al (2021) Mon Not R Astron Soc 505:3016
Nikzad S, Jewell AD, Hoenk M E et al (2017) J Astron Telesc Instrum Syst 3:036002
Nimmo K, Hessels JWT, Snelders M P et al (2023) Mon Not R Astron Soc 520:2281
Nishikida K, Edelstein J, Korpela EJ et al (2006) Astrophys J Lett 644:L171
Norris JP (2002) Astrophys J 579:386
Norris JP, Bonnell JT (2006) Astrophys J 643:266
Nugent AE, Fong W-f, Dong Y et al (2022) Astrophys J 940:57
Nyholm A, Sollerman J, Taddia F et al (2017) Astron Astrophys 605:A6
Oates SR, Page MJ, De Pasquale M et al (2012) Mon Not R Astron Soc 426:L86
Oates SR, Racusin JL, De Pasquale M et al (2015) Mon Not R Astron Soc 453:4121
Ofek EO, Fox D, Cenko SB et al (2013) Astrophys J 763:42
Ofek EO, Rabinak I, Neill JD et al (2010) Astrophys J 724:1396
Ofek EO, Sullivan M , Cenko SB et al (2013) Nature 494:65
Ofek EO, Zoglauer A, Boggs SE et al (2014) Astrophys J 781:42
Padovani P, Alexander DM, Assef RJ et al (2017) Astron Astrophys Rev 25:2
Perley DA, Kruhler T, Schulze S et al (2016) Astrophys J 817:7
Petrov P Singer LP, Coughlin M W et al (2022) Astrophys J 924:54
Piran T (2004) Rev Mod Phys 76:1143
Piro AL, Kollmeier JA (2018) Astrophys J 855:103
Platts E, Weltman A, Walters A et al (2019) Phys Rep 821:1
Prentice SJ, Maguire K, Smartt SJ et al (2018) Astrophys J Lett 865:L3
Rabinak I, Waxman E (2011) Astrophys J 728:63
Racusin JL, Karpov SV, Sokolowski M et al (2008) Nature 455:183
Rahin R, Moleri L, Vdovin A et al (2020) In Society of Photo-Optical Instrumentation Engineers (SPIE)
Conference Series, Vol. 11444, Space Telescopes and Instrumentation 2020: Ultraviolet to Gamma Ray,
den Herder J-WA, Nikzad KS (eds), 114446E
Rastinejad JC, Gompertz BP, Levan AJ et al (2022) Nature 612:223
Ridnaia A, Svinkin D, Frederiks D et al (2021) Nat Astron 5:372-377
Riley TE, Watts AL, Bogdanov S et al (2019) Astrophys J Lett 887:L21
Roming PWA, Kennedy TE, Mason KO et al (2005) Space Sci Rev 120:95
Roming PWA, Koch TS, Oates SR et al (2017) Astrophys J Suppl Ser 228:13
Roming PWA, Koch TS, Oates SR et al (2009) Astrophys J 690:163
Ror AK, Gupta R, Jelínek M et al (2023) Astrophys J 942:34
Rossi A, Rothberg B, Palazzi E et al (2022) Astrophys J 932:1
Rykoff ES, Aharonian F, Akerlof CW et al (2009) Astrophys J 702:489
Sagiv I, Gal-Yam A, Ofek EO et al (2014) Astron J 147:79
Sagués Carracedo A, Bulla M , Feindt U, Goobar A (2021) Mon Not R Astron Soc 504:1294
Sapir N, Katz B, Waxman E (2013) Astrophys J 774:79
Sari R, Piran T, Narayan R (1998) Astrophys J Lett 497:L17
Sarin N, Lasky PD (2021) Gen Relativ Gravit 53:59
Schady P (2017) R Soc Open Sci 4:170304
Schawinski K, Justham S, Wolf C et al (2008) Science 321:223
Schlegel E M (1990) Mon Not R Astron Soc 244:269
Schlegel EM, Petre R (2006) Astrophys J 646:378
Schulze S, Chapman R, Hjorth J et al (2015) Astrophys J 808:73
Seising J, Kruhler T, Malesani D et al (2018) Astron Astrophys 616:A48
Shvartzvald Y, Waxman E, Gal-Yam A et al (2023) e-prints, 2304.14482
Singh KP, Tandon SN, Agrawal PC et al (2014) In Society of Photo-Optical Instrumentation Engineers (SPIE)
Conference Series, Vol. 9144, Space Telescopes and Instrumentation 2014: Ultraviolet to Gamma Ray,
Takahashi T, den Herder J-WA, Bautz M (eds), 91441S
Smartt SJ (2009) Annu Rev Astron Astrophys 47:63
Smith N (2014) Annu Rev Astron Astrophys 52:487
Smith N (2017) In: Alsabti AW, Murdin P (eds) Handbook of supernovae, p 403
Smith N, Kilpatrick CD, Mauerhan JC et al (2017) Mon Not R Astron Soc 466:3021
Soares-Santos M , Holz DE, Annis J et al (2017) Astrophys J Lett 848:L16
Soderberg A M , Berger E, Page K L et al (2008) Nature 453:469
Soumagnac MT, Ofek EO, Liang J et al (2020) Astrophys J 899:51
Swenson CA, Roming PWA, De Pasquale M , Oates SR (2013) Astrophys J 774:2
Tanaka M , Kato D, Gaigalas G, Kawaguchi K (2020) Mon Not R Astron Soc 496:1369
Tanvir NR, Levan AJ, Fruchter AS et al (2013) Nature 500:547
Springer
11 Page 40 of 41 N.Werner et al.
Tavani M , Casentini C, Ursi A et al (2021) Nat Astron 5:401
Tolstov A, Zhiglo A, Nomoto K et al (2017) Astrophys J Lett 845:L2
Trója E, Castro-Tirado AJ, Becerra Gonzalez J et al (2019) Mon Not R Astron Soc 489:2104
Trója E, Fryer CL, O'Connor B et al (2022) Nature 612:228
Vestrand WT, Wozniak PR, Wren JA et al (2005) Nature 435:178
Villar VA, Guillochon J, Berger E et al (2017) Astrophys J Lett 85LL21
Wanajo S, Sekiguchi Y, Nishimura N et al (2014) Astrophys J Lett 789:L39
Waxman E, Katz B (2017) In: Alsabti AW, Murdin P (eds) Handbook of Supernovae, p 967
Werner N, Řípa J, Münz F et al (2022) In Society of Photo-Optical Instrumentation Engineers (SPIE) Conference
Series, Vol. 12181, Space Telescopes and Instrumentation 2022: Ultraviolet to Gamma Ray, den
Herder J-WA, Nikzad KS (eds), 121810B
Williams CL, Panagia N, Van Dyk SD et al (2002) Astrophys J 581:396
Woodgate BE, Kimble RA, Bowers C W et al (1998) Publ Astron Soc Pac 110:1183
Wyatt SD, Tohuvavohu A, Arcavi I et al (2020) Astrophys J 894:127
Yan L, Quimby R, Gal-Yam A et al (2017) Astrophys J 840:57
Yang B, Jin Z-P, Li X et al (2015) Nat Commun 6:7323
Yi S-X, X i S-Q, Yu H et al (2016) Astrophys J Suppl Ser 224:20
Yi S-X, Yu H, Wang FY, Dai Z-G (2017) Astrophys J 844:79
York DG, Adelman J, Anderson JE Jr et al (2000) Astron J 120:1579
Yu Y-W, Zhang B, Gao H (2013) Astrophys J Lett 776:L40
Zajaček M , Czerny B, Jaiswal V K et al (2024) Space Sei Rev 220
Zaninoni E, Bernardini MG, Margutti R, Oates S, Chincarini G (2013) Astron Astrophys 557:A12
Zhang B (2020) Nature 587:45
Zhang B, Mészáros P (2004) Int J Mod Phys A 19:2385
Zoglauer A, Andritschke R, Schopper F (2006) New Astron Rev 50:629
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Authors and Affiliations
N. Werner1
• J. Řípa1
• C. Thöne2
• F. Münz1
• P. Kurfürst1
• M. Jelínek2
• F. Hroch1
•
J. Benáček3
• M. Topinka4
• G. Lukes-Gerakopoulos5
• M. Zajaček1
• M. Labaj1
•
M. Prišegen1
-6
• J. Krtička1
• J. Merc7
• A. Pál8
• O. Pejcha9
• V. Daniel10
• J. Jon1 0
•
R. Šošovička1 0
• J. Gromeš1 0
• J. Václavík1 1
• L Steiger11
• J. Segiňák1 2
• E. Behar13
•
S. Tarem13
• J. Salh13
• O. Reich13
• S. Ben-Ami14
• M.F. Barschke15
• D. Berge15
-16
•
A. Tohuvavohu17
• S. Sivanandam17
• M. Bulla18
-19
-20
• S. Popov21
•
Hsiang-Kuang Chang22
EE3 N.Werner
werner @physics .muni.cz
[3] J. Řípa
ripa.jakub @ gmail.com
1
Department of Theoretical Physics and Astrophysics, Faculty of Science, Masaryk University,
Kotlářská 2, Brno, 611 37, Czech Republic
2
Astronomical Institute, Czech Academy of Sciences (ASU CAS), Ondřejov, Czech Republic
^ Institute for Physics and Astronomy, University of Potsdam, D-14476 Potsdam, Germany
4
INAF - Osservatorio Astronomico di Cagliari, via della Scienza 5, Selargius (CA), 09047, Italy
5
Astronomical Institute, Czech Academy of Sciences (ASU CAS), Boční II 1401/la, 141 00,
Prague, Czech Republic
Springer
Science with QUVIKI Page 41 of 41 11
10
Advanced Technologies Research Institute, Faculty of Materials Science and Technology in
Trnava, Slovak University of Technology in Bratislava, Bottova 25, Trnava, 917 24, Slovakia
Astronomical Institute, Faculty of Mathematics and Physics, Charles University, V Holesovickach
2, Praha, 180 00, Czech Republic
Research Centre for Astronomy and Earth Science, Konkoly Observatory, K Konkoly-Thege M . ut
15-17, Budapest, H-1121, Hungary
Institute of Theoretical Physics, Faculty of Mathematics and Physics, Charles University, V
Holesovickach 2, Prague 180 00, Czech Republic
Czech Aerospace Research Centre (VZLU), Prague, 199 05, Czech Republic
1 1
Research Centre for Special Optics and Optoelectronic Systems (TOPTEC), Institute of Plasma
Physics, Czech Academy of Sciences, Za Slovankou 1782/3, 182 00, Prague, Czech Republic
1 2
PEKASAT SE, Brno, Czech Republic
^ Department of Physics, Technion, Haifa 32000, Israel
14 Department of Particle Physics and Astrophysics, Weizmann Institute of Science, Herzl St 234,
Rehovot, Israel
1 5
Deutsches Elektronen-Synchrotron DESY, Platanenallee 6, 15738 Zeuthen, Germany
1 6
Institut für Physik, Humboldt-Universität zu Berlin, Newtonstrasse 15, 12489 Berlin, Germany
17 Department of Astronomy & Astrophysics, University of Toronto, 50 St. George Street, Toronto,
Ontario, M5S 3H4 Canada
^ Department of Physics and Earth Science, University of Ferrara, via Saragat 1,1-44122, Ferrara,
Italy
1 9
Sezione di Ferrara, INFN, via Saragat 1,1-44122, Ferrara, Italy
2
^ Osservatorio Astronomico d'Abruzzo, INAF, via Mentore Maggini sne, 1-64100, Teramo, Italy
21 ICTP - Abdus Salam International Center for Theoretical Physics, Strada Costiera 11,1-34151
Trieste, Italy
99
Institute of Astronomy, National Tsing Hua University, 101 Sect. 2 Kuang-Fu Rd., Hsinchu,
300044, Taiwan, Republic of China
Springer