Contrib. Astron. Obs. Skalnaté Pleso 47, 200 – 207, (2017)
Ultraviolet astronomy with small space
telescopes
N. Brosch1, Ana I. Goméz De Castro2, J. Murthy3, E.N. Ribak4,
E. Behar4 and V. Balabanov5
1 Tel Aviv University, Tel Aviv, Israel (E-mail: noah@wise.tau.ac.il)
2 Complutense University, Madrid, Spain
3 Indian Institute of Astrophysics, Bangalore, India
4 Dept of Physics, Technion, Haifa, Israel
5 Asher Space Research Institute, Technion, Haifa, Israel
Received: May 15, 2017; Accepted: June 15, 2017
Abstract. After describing the present situation with astronomy in the ultra-
violet (UV), reviewing a few past and proposed future missions, we present
options to develop space missions that have been realized for modest costs. In
this context, we bring together a few aspects of different missions and projects
that, when combined, might result in a low-cost mission for imaging or low-
resolution spectroscopy in the UV.
Key words: cubesat – ultraviolet – space astronomy
1. Five decades of UV astronomy
The UV astronomy started with rocket flights that offered only short observation
windows. The first major advance was the survey mission by the TD-1 satellite
from 1972 to 1974 that surveyed the entire sky to the relatively bright limit-
ing magnitude of 8.8 (Boksenberg et al., 1974). The TD-1 catalog comprised
31,215 sources measured in a single spectral band centered on 275 nm. TD-1
also had modest spectroscopic capabilities since it used also a spectrometer for
the 130-255 nm region for the brighter sources. The magnitude system used is
“monochromatic” such that
mλ = − 2.5 log(fλ) − 21.175 (1)
−1
where fλ is the source flux density in erg sec
−1 cm−2 Å at wavelength λ.
Following the first all-sky survey by TD-1 the next all-sky survey was started
only in 2003 by the GALEX satellite (Martin et al., 2005). GALEX, a NASA
“small” (Small Explorer=SMEX) mission, operated until 2012 albeit with grad-
ually degrading efficiency. Originally it observed in two spectral bands, the Near
UV (NUV: 175-280 nanometers) and the Far UV (FUV: 135-174 nanometers).
The FUV detector failed in May 2009 due to overcurrent and for the rest of the
mission observations were performed only with the NUV channel. The GALEX
Ultraviolet astronomy with small space telescopes 201
all-sky catalog contains some 65.3 million sources to a depth of about 19.9/20.8
FUV/NUV (ABmag). The AB magnitudes are defined as
mAB = − 2.5 log(fν) − 48.600 (2)
The GALEX mission’s life-cycle cost to NASA was $150.6 million. A sizable
fraction of the sky was not surveyed at all in order to avoid bright stars that
could have damaged the detectors and the emphasis was put on high galactic
latitude regions to sample as many galaxies as possible.
The long time span between TD-1 and GALEX, and the lack of other survey
missions since the decommissioning of GALEX, make it difficult to retain re-
taining scientists and engineers specialized in UV technology and data analysis.
The lack of a clear vision of future missions has become a major problem to
retain the know-how in the field. This paper point to a possible solution using
the now-popular CubeSat philosophy and offers a preliminary concept for such
a missions.
2. Small missions
In this section we describe a few small missions, selecting specifically a few in
the UV, and pointing out that significant science can be done with small optics
and small payloads.
2.1. Chang’e-3
The China National Space Administration (CNSA) landed on the Moon, on
14 December 2013, a mission called Chang’e-3. The spacecraft carried a small
rover and a number of experiments operating on the lander. One of these, called
Lunar-based Ultraviolet Telescope(LUT), was a 15-cm telescope equipped with
a orientable folding mirror to allow observations of sources in different sky di-
rections. The operation of LUT during the first 18 months was described by
Wang et al. (2015).
LUT observes in a band rather close to the NUV one of GALEX (∼ 270 nm)
and can access sources in the vicinity of the lunar north pole in a sky area of
∼3000 square degrees. By observing UV standards, Wang et al. determined the
zero point of the LUT photometry, showed the relative photometric stability, and
determined that lunar dust does not affect observations as much as previously
feared.
2.2. PISAT-UV
The PES Institute of Technology in Bangalore, India developed a series of small
satellites of which the first was launched in a sun-synchronous orbit on 26
September 2016. PISAT-1 is a three-axis stabilized structure with dimensions
25×25×18 cm with a mass of 5.3 kg carrying an Earth-imaging payload (a
202 N. Brosch, Ana I. Goméz De Castro, J. Murthy, E. Ribak, E. Behar and V. Balabanov
CMOS color camera in the visible for Earth observation experimentation man-
ufactured by GomSpace). Each image covers an area of ∼ 165 × 125 km with a
spatial resolution of 80-m.
One of the future PISATs will carry an ultraviolet imaging telescope that will
be designed and built at the Indian Institute of Astrophysics (IIA) in Bangalore.
The preliminary design is for a 10.8-cm aperture Ritchey-Chrétien design that
will operate in the NUV to cover a three-degree wide field of view with an angular
resolution of 13-arcsec (Ambily et al., 2016). The detector will be a solar-blind
multi-channel plate optically coupled to a CMOS detector read out with a FPGA
based data acquisition board. The entire telescope masses 2.5-kg and requires
less than 5W to operate (excluding thermal control). Its development costed
less than 10,000 US$.
2.3. Duchifat, Hoopooe, and ABIR
The Herzliya Space Laboratory (HSL) is a student satellite building laboratory.
The students are upper high-school, primarily from the Herzlyia (Israel) area,
but now HSL is collaborating with same-level students from at least five other
high schools, including some from Arab, Druse and Bedouin communities. In
building space hardware, the students enjoy specialist help from experienced
mentors from the space industry. Duchifat was the first Israeli satellite to be
designed, built and constructed by high school students, and the first Israeli
CubeSat (see below on CubeSats). It is a 1U CubeSat used for store-and-forward
communications, it was launched on July 19, 2014, and is still operating at the
time of writing this paper.
The second nano-satellite built by HSL is Hoopooe, a 2U nano-sat with a
2-kg mass that was launched as part of the QB50 multi-satellite mission to
study upper atmosphere electricity. HSL is now building its third nano-sat,
ABIR (Knight), a 3U nano-sat that will carry an Earth-imaging payload, rather
similar to the one described on PISAT-1. ABIR will be three-axis stabilized and
it, as do most of the nano-sats constructed or planned, relies heavily on off-
the-shelf subsystems now available from a large variety of vendors. Although
not operating in the UV, these nano-satellites demonstrate that it is possible
to have successful space experiments without extremely high costs and very
sophisticated infrastructure.
3. The CubeSat revolution
Space research was strongly dependent since its inception by the fabrication of
the platforms used to operate scientific experiments. These were mostly one-of-
a-kind constructions, requiring expensive testing to ensure a reasonable degree
of success. Yet satellite and spacecraft platforms have a number of common
requirements that drive a need for similar properties. The platforms have to be
sturdy to survive the rigors of launch and operation in space (i.e., power supply,
Ultraviolet astronomy with small space telescopes 203
thermal and attitude control, etc.). In many cases these requirements could be
served by standardized components; these could be fabricated in relatively large
numbers, reducing the cost-per-unit due to volume production, and if one would
be qualified for space flight further products of the same kind would not require
the expensive qualification process.
In 1999 the California Polytechnic State University (Cal Poly) and Stanford
University developed the CubeSat specifications to ease the design, manufacture,
and testing of small satellites intended for low Earth orbit (LEO). The standard
CubeSat is a 10×10×11.35 cm unit (1U) that provides 10×10×10 cm clear space
(one liter of useful volume) while weighing no more than 1.33 kg. More units
can be stacked to offer larger volumes, forming 2U, 3U, 6U or larger platforms.
Since many CubeSats are 10×10 cm (regardless of length) they can be launched
and deployed using a common deployment system called a Poly-PicoSatellite
Orbital Deployer (P-POD), developed and built by Cal Poly. Since the P-PODs
and the CubeSats they deploy in orbit are relatively light-weight, they do not
require dedicated launchers but can be deployed as secondary payloads whenever
a larger and much more massive satellite is orbited. On the other hand, there
is almost no control of the orbit, which will be close to that of the primary
payload.
The standard 10×10 cm cross-section implies that electronic boards and
special sub-system, such as mechanical structure, power distribution, on-board
computer, attitude and control systems can be fabricated on standard (PCI-
104 form factor) boards and used for many missions. This caused the rise of
specialized industries that design, fabricate, test and qualify sub-systems, which
are then available to interested experimenters as custom-off-the-shelf (COTS)
products. All these measures, and a philosophy of one-string design (foregoing
the redundancy requirements) and extensive testing of different models of the
experiment (structural, electrical/electronical, qualification, etc.) work together
to reduce considerably the cost of a space experiment to as low as a few tens of
thousands of US$ for a 1U CubeSat. This has been pointed out clearly by the
Committee on Achieving Science Goals with CubeSats; Space Studies Board;
Division on Engineering and Physical Sciences; National Academies of Sciences,
Engineering, and Medicine (Zurbuchen & Lal, 2017).
Advances in CubeSat design allow now the use of deployable solar panels
(more available power and even deployable structures with final dimensions
much larger than those of the CubeSat as launched; e.g. the RainCube 6U
platform by JPL, which houses a ∼50-cm parabolic antenna for a Ka-band (35.75
GHz) radar payload. Enlisting this capability of deploying larger structures from
CubeSats for a possible astronomy mission implies the capability of orbiting
relatively large collecting optics that would make a CubeSat mission almost
equivalent to a “small” sized NASA mission such as GALEX.
204 N. Brosch, Ana I. Goméz De Castro, J. Murthy, E. Ribak, E. Behar and V. Balabanov
4. Segmented optics
Classical telescope optics follow approximately a ratio of 1/10 between the di-
ameter and the primary mirror (PM) thickness. This is the result of the need
to maintain the reflecting surface as accurately as possible, at all orientations
of the mirror relative to the gravity vector. Thus high-capability optics suffer
from a relatively large volume and, given the thickness-to-the-diameter ratio
and the density of the materials used to fabricate the mirror, from an associ-
ated large mass. All these make the launching of a classical telescope to space
a prohibitively expensive undertaking.
There are means of reducing the mass of the PM without affecting its optical
performance, such as selectively machining away large fractions of the glass
(“light-weighting”) while leaving sufficient material to form structural ribs and
bosses for securing the finished mirror to its support structure. Extreme light-
weighting can reduce the mass of a mirror to 20% or less of its initial, not
light-weighted mass. Other means to produce lightweight mirrors require the
use of exotic materials, such as beryllium or silicon-impregnated carbon (C-
SiC). Also, it is possible to divide the primary mirror into thin segments, saving
mass and relying on a support structure to maintain the optical surface. This
is the method chosen for the PM of the James Webb Space Telescope (JWST),
where a complicated mechanism deploys the PM stack of 18 beryllium segments,
each weighting about 20-kg, and aligns them to form the 6.5-m diameter almost
diffraction-limited (in the near-infrared at 2 µm) PM.
To reach the required optical quality, the segments of the JWST’s PM can
be adjusted in space to bring them perfectly in-phase with actuators applying
force to the backs of the segments. This allows not only adjusting the tip and tilt
of each segment, but also of slightly changing its curvature. It is expected that
after JWST reaches is operational orbit at the second Lagrangian point (L2)
of the Sun-Earth system, some two months will be spent aligning the optics.
The “folding” of the JWST PM also reduces the launched volume, allowing the
6.5-m diameter mirror to fit in the nose cone of the Ariane-5 launcher, which
allows a maximum 4.57 meter payload diameter. Segmented mirrors reduce thus
both the mass and the volume of the optical systems to be orbited.
5. Segmented optics at the Technion
The Asher Space Research Institute (ASRI) and the Department of Physics at
the Technion in Israel run a program to develop segmented mirrors that can be
deployed and aligned in space. The preliminary design of such a device has each
segment (called here “petal”) made of the reflecting surface that has the shape
and finish of the final mirror and is supported by a stiff mounting surface that
deploys, together with all its attachments, from the main structure. Between
the mounting surface and the back of the mirror segment there are a number
Ultraviolet astronomy with small space telescopes 205
of nano-actuators that can push or pull the segment relative to the mounting
surface. This way, each segment has independent capability to “piston”, “tip”
or “tilt”, thus being able to adjust as required.
The nano-actuators are piezo-electric devices of high reliability and accuracy,
thus the motions are well-controlled and repeatable, while not requiring power
when not actuated. As with many segmented telescopes, the major issue in
forming the final accurate reflecting surface is the phasing of all the segments.
As mentioned above, this phase is estimated to require two full months for
the JWST. As of now, there is no direct method of measuring the phase of
propagating light and the best one can do is to measure intensity. Using this
method, laboratory tests with a bench-top system done at ASRI succeeded
to phase a four segment mirror system in a few hours by using a hardware
“simulated annealing” algorithm that is iterative and uses a feedback loop to
correct the phase errors (Paykin et al., 2015). Success is determined by the
achievement of the sharpest image, and the method is efficient since the phasing
sensor is the same as the imaging sensor of the payload and the source against
which the optimization procedure is performed is a star (simulated in the lab
by a point source).
6. Putting it all together
With the elements described above, it is possible to propose a revolutionary
platform which, with the form factor of a 3U or 4U CubeSat could perform
UV science almost at par with what GALEX yielded for a mission cost well
below 1 M$, i.e. less than 1% of the total GALEX budget. A generic view of
the experiment is shown in Fig.1.
Figure 1. A possible 3U configuration of the proposed spacecraft. The left panel shows
a view from the sky with the four petal extended. The right panel shows a side view
of the open petals and the secondary assembly. The incident light is focused on the
detector located in the bottom unit of the 3U configuration.
206 N. Brosch, Ana I. Goméz De Castro, J. Murthy, E. Ribak, E. Behar and V. Balabanov
As Fig.1 shows, the top two units of the 3U configuration carry the four
“petals” of the segmented PM and the secondary mirror assembly. Once in space,
the petals deploy and form an unphased PM of some 800 cm2 unobstructed
collecting area (smaller than the ∼ 5000 cm2 of GALEX, but still significant).
The phasing is done via the simulated annealing algorithm using the detector,
which could be a variant of the one developed for PISAT-UV. The back-sides of
the support elements of each petal carry solar panels and are therefore directed
sunward after the petal deployment, which puts the telescope itself in an anti-
sunward direction.
Between the deployed petals it will be necessary to deploy some thermal
blankets. These would ensure that the optics are always in shadow and stay
cool during the orbital day. The bottom unit of the 3U configuration would
carry all the necessary subsystems for the proper operation of the experiment;
if necessary, this can be extended by another unit for a final configuration of 4U.
Power is supplied from the solar panels on the bottoms of the petals and on all
sides of the bottom unit(s) of the configuration, where the telemetry antennas
are also located.
There are numerous science goals that can be achieved with variants of this
experiment. To mention just a few, it would be possible to fill-in areas and topics
not covered by GALEX such as observing in the FUV, or to monitor transient
sources in the UV, or to perform a survey of the 2174Å interstellar extinction
feature by adding a grism in the converging beam that would allow measuring
the equivalent width of the feature in many stars.
7. Conclusions
We have described a possible mission for the space UV based on the CubeSat
architecture, with most components commercially available, and where the sci-
ence payload is based on work on segmented mirrors and their phasing done
at the Technion, and on the detector package for small satellites developed at
the IIA. We estimate that such a mission could be realized within two years,
would cost less than 1M$, and could provide unique science in the UV. More-
over, if realized such a mission could provide the pathfinder of future missions
using essentially the same hardware with minor modification to attack different
problems in the field of ultraviolet astronomy.
Acknowledgements. NB acknowledges support from the Czech Academy of Sci-
ences to attend the IBWS 2017 meeting. AIG acknowledges Research Grants from
Spain’s Ministry of Economy, Industry and Competitiveness: ESP2015-68908-R, and
AYA2015-71864-REDT.
References
Ambily, S., Mathew, J., Sarpotdar, M., et al. 2016, in SPIE proceedings, Vol. 9905,
Ultraviolet astronomy with small space telescopes 207
Space Telescopes and Instrumentation 2016: Ultraviolet to Gamma Ray, 990539
Boksenberg, A., Evans, R. G., Fowler, R. G., et al. 1974, in Reports of New Ultraviolet
Observations, C8.1
Martin, D. C., Fanson, J., Schiminovich, D., et al. 2005, The Astrophysical Journal,
619, L1
Paykin, I., Yacobi, L., Adler, J., & Ribak, E. N. 2015, Physical Review E, 91, 023302
Wang, J., Meng, X. M., Han, X. H., et al. 2015, Astrophysics and Space Science, 360,
10
Zurbuchen, T. H. & Lal, B. 2017, in American Astronomical Society Meeting Ab-
stracts, Vol. 229, American Astronomical Society Meeting Abstracts, 328.01