See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225468958

The Sunrise Mission

Article  in  Solar Physics · November 2011

DOI: 10.1007/s11207-010-9662-9 · Source: arXiv

CITATIONS READS

167 380

45 authors, including:

B. Chares Werner Curdt

Max Planck Institute for Solar System Research Max Planck Institute for Solar System Research
12 PUBLICATIONS   240 CITATIONS    259 PUBLICATIONS   5,493 CITATIONS   

SEE PROFILE SEE PROFILE

Tino Riethmüller Georg Tomasch

Max Planck Institute for Solar System Research Max Planck Institute for Solar System Research
57 PUBLICATIONS   973 CITATIONS    9 PUBLICATIONS   526 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

AstroCombs View project

Daniel K Inouye Solar Telescope View project

All content following this page was uploaded by B. Chares on 04 June 2014.

The user has requested enhancement of the downloaded file.

 Submitted for publication in Solar Physics
DOI: 10.1007/•••••-•••-•••-••••-•

The Sunrise Mission

P. Barthol1 · A. Gandorfer1 · S. K. Solanki1 ·

M. Schüssler1 · B. Chares1 · W. Curdt1 ·
W. Deutsch1 · A. Feller1 · D. Germerott1 · B. Grauf1 ·
K. Heerlein1 · J. Hirzberger1 · M. Kolleck1 ·
R. Meller1 · R. Müller1 · T. L. Riethmüller1 ·
arXiv:1009.2689v1 [astro-ph.IM] 14 Sep 2010

G. Tomasch1 · M. Knölker2 · B. W. Lites2 · G. Card2 ·

D. Elmore2∗ · J. Fox2 · A. Lecinski2 · P. Nelson2 ·
R. Summers2 · A. Watt2 · V. Martı́nez Pillet3 ·
J. A. Bonet3 · W. Schmidt4 · T. Berkefeld4 ·
A. M. Title5 · V. Domingo6 · J. L. Gasent Blesa6 ·
J. C. del Toro Iniesta7 · A. López Jiménez7 ·
A. Álvarez-Herrero8 · L. Sabau-Graziati8 ·
C. Widani9 · P. Haberler9 · K. Härtel9 · D. Kampf9 ·
T. Levin9 · I. Pérez Grande10 · A. Sanz-Andrés10 ·
E. Schmidt11

c Springer ••••

Abstract
The first science flight of the balloon-borne Sunrise telescope took place in June
2009 from ESRANGE (near Kiruna/Sweden) to Somerset Island in northern Canada.

1 Max-Planck-Institut für Sonnensystemforschung,

Max-Planck-Straße 2, D-37191 Katlenburg-Lindau, Germany
email: barthol@mps.mpg.de
2 High Altitude Observatory† , P.O. Box 3000, Boulder, CO 80301,

USA
3
Instituto de Astrofı́sica de Canarias, C/Vı́a Láctea, s/n, E-38205,
La Laguna (Tenerife), Spain
4 Kiepenheuer-Institut für Sonnenphysik, Schöneckstraße 6,

D-79104 Freiburg, Germany

5 Lockheed Martin Solar and Astrophysics Laboratory, Bldg. 252,

3251 Hanover Street, Palo Alto, CA 94304, USA

6 Grupo de Astronomı́a y Ciencias del Espacio (Univ. de Valencia),

E-46980, Paterna, Valencia, Spain

7 Instituto de Astrofı́sica de Andalucı́a (CSIC), Apdo. de Correos

3004, E-18080, Granada, Spain

8 Instituto Nacional de Técnica Aerospacial, E-28850,

Torrejón de Ardoz, Spain

9 Kayser-Threde GmbH, Wolfratshauser Straße 48,

D-81379 München, Germany

10 Universidad Politécnica de Madrid, IDR/UPM, Plaza Cardenal

Cisneros 3, E-28040, Madrid, Spain

11 Ingenieurbüro für Optikentwicklung, Amalienstraße 12,

D-85737 Ismaning, Germany

 Barthol et al.

We describe the scientific aims and mission concept of the project and give an overview
and a description of the various hardware components: the 1-m main telescope with
its postfocus science instruments (the UV filter imager SuFI and the imaging vector
magnetograph IMaX) and support instruments (image stabilizing and light distribution
system ISLiD and correlating wavefront sensor CWS), the optomechanical support
structure and the instrument mounting concept, the gondola structure and the power,
pointing, and telemetry systems, and the general electronics architecture. We also ex-
plain the optimization of the structural and thermal design of the complete payload. The
preparations for the science flight are described, including AIV and ground calibration
of the instruments. The course of events during the science flight is outlined, up to the
recovery activities. Finally, the in-flight performance of the instrumentation is briefly
summarized.
Keywords: Instrumentation and Data Management, Integrated Sun Observations, Mag-
netic fields, Photosphere

1. Introduction

1.1. Solar Balloon Missions

A number of projects has been devoted in the past to observing the Sun from balloon-
borne platforms in the stratosphere. In 1957 and 1959, project Stratoscope (Schwarz-
schild, 1959) obtained white-light images of granulation and sunspots with a 30-cm
telescope (Danielson, 1961). A different instrument called Spektro-Stratoskop took sev-
eral long sequences of high quality granulation pictures with an evacuated 30-cm tele-
scope (Mehltretter, 1978). From 1966 to 1973, there were several flights of the Soviet
Stratospheric Solar Station, which operated alternatively with main mirrors of 50 cm
and 1 m diameter and took filtergrams in the visible range (e.g.Krat et al., 1972).
Smaller telescopes (5–10 cm) with broad-band filters in the visible were flown by a
Japanese group in the 1970s (Hirayama, 1978). Groups in France attempted UV imag-
ing in the wavelength range 200–300 nm with a 20-cm telescope (Herse, 1979) and
took spectra in the same wavelength range with telescopes of up to 30 cm aperture
from the mid 1960s on (Samain and Lemaire, 1967; Lemaire and Blamont, 1985).
More recently, flights of the Flare Genesis project in 1996 and 2000 provided magne-
tograms and Dopplergrams in the CaI 6122.2 line using a 80-cm telescope (Rust et al.,
1996; Bernasconi et al., 2000). The 30-cm telescope of the Solar Bolometric Imager
(Bernasconi et al., 2004) provided maps of the solar disk in integrated light between
0.28 and 2.6 µm during three flights between 2003 and 2007.
These projects demonstrated the advantages and the rich potential of balloon-borne
solar observations at stratospheric heights: (i) negligible image degradation by atmo-
spheric seeing, (ii) access to the UV range down to about 200 nm wavelength, (iii)
retrievability of the instruments, and (iv) much reduced cost in comparison to space
projects.

∗ Now at National Solar Observatory/Sacramento Peak, P.O. Box 62, Sunspot, NM 88349, USA
† HAO/NCAR is sponsored by the National Science Foundation
The Sunrise Mission

The Sunrise mission comprises the biggest and most complex payload flown so far
in a solar balloon mission: a 1 m telescope equipped with a multi-wavelength UV filter
imager, a Fabry-Pérot-based vector magnetograph, and a correlating wavefront sensor
for active alignment control and image stabilization.

1.2. Science Case

The solar photosphere is crucial for the investigation of the solar magnetic field. This
thin layer, where the plasma becomes optically thin and almost all of the radiative en-
ergy flux is emitted, represents the key interaction region: thermal, kinetic and magnetic
energy all are of the same order of magnitude and transform most easily from one form
into another. The interaction between convection, radiation, and magnetic field in the
electrically conducting solar plasma leads to the creation of a rich variety of magnetic
structure with intense (kilogauss) magnetic field concentrations on size scales reaching
well below a pressure scale height. At the same time, the photosphere appears to harbor
a stunning amount of mixed-polarity ‘turbulent’ magnetic flux, which possibly results
from small-scale local dynamo action driven by the granulation flows. The photospheric
magnetic field is in a state of constant change: ceaseless transport, stretching, emergence
and submergence, cancellation, intensification and dissipation of magnetic flux take
place down to the smallest spatial scales that can be observed so far. These processes
control the structure, dynamics, and energetics of the solar atmosphere at larger scales;
they are the source of solar variability and, ultimately, of solar influences on the Earth.
To understand these fundamental processes, we must learn how the magnetic field in-
teracts with the solar plasma and have to uncover the conversion of energy between its
mechanical, magnetic, radiative and thermal forms. Consequently, the central questions
of Sunrise science are
• What are the origin and the properties of the intermittent magnetic structure,
including the kilogauss concentrations?
• How is the magnetic flux brought to and removed from the solar surface? What is
the role played by local dynamo action and reconnection processes?
• How does the magnetic field assimilate and provide energy to heat the upper solar
atmosphere?
• How does the variable magnetic field modify the solar brightness?
In order to answer these questions, the key scientific objective of the Sunrise project
is to study the structure and dynamics of the solar magnetic field at the spatial resolution
afforded by a 1-m telescope, over extended stretches of time with constant observational
conditions free of seeing effects, with a time resolution sufficient to track rapid changes
of the magnetic field, and over a field of view large enough to provide good statistics of
relevant events and to follow the evolution of magnetic structure during all phases of its
life cycle.
The required observations include continuous, quantitative measurements of the
magnetic field vector, the plasma velocity and related atmospheric structure, together
with brightness maps in several wavelength bands. For the latter, the UV wavelengths
between 200 nm and 300 nm are of particular interest since a) no high-resolution
imaging has been done so far in this range and b) cycle-related irradiance variations
in this spectral band affect the temperature in the terrestrial stratosphere through Ozone
 Barthol et al.

photochemistry. Numerical simulations predict intensity contrasts of magnetic flux con-

centrations at these wavelengths that significantly exceed those in the visible.

1.3. Mission Concept

Sunrise has been designed as a balloon-borne stratospheric solar observatory in the

framework of NASA’s LDB (Long Duration Balloon) program. Zero pressure helium
balloons with volumes up to 1 million cubic meters lift science payloads such as
Sunrise with a mass of several tons to float altitudes of 35–37 km. This ’low cost access
to near space’ has several advantages compared to ground-based or space-borne solar
instrumentation. As with space-borne instruments, stratospheric solar observations do
not suffer from two major drawbacks encountered with ground based telescopes. Being
above 99 % of the Earth’s atmosphere, wavefront distortions due to atmospheric turbu-
lence are virtually not existent. This provides long-term and seeing-free high resolution
imaging conditions. In addition, the interesting solar UV radiation between 220 nm and
370 nm is accessible, which is absorbed in the lower stratosphere by the Hartley and
Huggins bands of Ozone. At mission termination balloon-borne instruments usually can
be recovered after landing with only moderate damage. Modifications for a future flight
and refurbishment can be accomplished very cost effectively, at far lower investment
levels compared to space instruments.
NASA LDB missions are managed and operated by the Columbia Scientific Balloon-
ing Facility (CSBF), Palestine, Texas, USA and in the past have been launched mainly
from Williams Field near McMurdo, Antarctica (77.86◦ S, 167.13◦ E), with a launch
window from December to January. NASA recently expanded their launch capabilities
by cooperating with ESRANGE (67.89◦ N, 21.10◦ E) near Kiruna, Sweden, allowing
missions to be launched during the northern summer period as well. The European
sounding rocket and balloon base provides excellent infrastructure and is accessible
with much lower logistical effort compared to McMurdo. Launching balloons during
solstice conditions close to the polar circle offers uninterrupted solar observations with-
out day/night cycles. Permanent sunlight and only small elevation changes of the Sun
form ideal conditions, so that undisturbed observation and power generation for the
instruments are guaranteed. Furthermore, thermal conditions do not vary significantly
and the balloon floats at nearly constant altitude. Relatively stable wind systems at float
altitude take the balloon and instrument on circumpolar trajectories with flight durations
of 9 to 12 days per revolution. ESRANGE was chosen for the first science flight of
Sunrise , although the flight duration currently is limited to 5–6 days due to the lack
of Russian overflight permissions for NASA balloons. The processes to be studied by
Sunrise generate highly variable structures like emerging and decaying active regions
or sunspots. The timescales involved range from seconds on small spatial scales over
minutes and hours to several days in case of sunspots. Due to their intrinsic nature these
phenomena cannot be predicted in time and location. Being at the onset of the new solar
activity cycle, the shorter flight duration was judged to be acceptable in comparison to
the higher logistical effort and risk related with an Antarctic flight.
Coordinated measurements were done with HINODE and SUMER onboard SOHO.
Parallel observations with ground-based solar telescopes such as the Swedish Solar
Telescope (SST) on La Palma, the Vacuum Tower Telescope (VTT) on Tenerife, the
IRSOL facility in Locarno (Switzerland), and the Dunn Solar Telescope on Sacramento
Peak (New Mexico) have been performed to support the target selection for Sunrise .
The Sunrise Mission

2. Instrument Description

Sunrise in full flight configuration is shown in Figure 1. The instrument can be broken
down into the following major components, which will be described in more detail
below:

• Telescope
• Postfocus instrumentation platform (PFI) with science instruments and image
stabilization system
• Gondola with photovoltaic arrays and pointing system
• Instrument electronics
• CSBF-provided ballooning equipment
• Ground-support equipment

2.1. Telescope

The Sunrise telescope is a light-weight Gregory-type reflector with nearly 25 m effec-

tive focal length. The optical path is depicted in Figure 2. A parabolic primary mirror
with 1010 mm outer diameter and 1000 mm clear aperture creates a real image of the
sun in the first focus F1 of the telescope, 2422.9 mm in front of the M1 vertex. A field
stop in F1 limits the telescope field-of-view (FoV) so that only a small fraction reaches
the elliptical secondary mirror M2. M2 has a diameter of 245 mm and a focal length of
505 mm. M2 magnifies the image in F1 by a factor of ten and refocuses the radiation
to a secondary telescope focus, located in the postfocus instrumentation package ontop
the central telescope frame. Two plane fold mirrors M3 and M4 redirect the light from
the telescope to this upper level. A further physical stop in the secondary focus limits
the telescope FoV to 180 arcsec.
The Serrurier structure (Figure 3) connects the steel central frame, the front ring
with the secondary mirror assembly and heat rejection wedge, and the rear ring with
the primary and tertiary mirrors. The central frame serves as the main structural com-
ponent. It is a riveted sheet metal construction with metallic inserts at locations with
high mechanical load. The frame carries the postfocus instrumentation (PFI) piggy-back
on top. Axles mounted left and right to interface plates at the height of the combined
telescope and PFI center-of-gravity form the elevation axis that connects to the gondola
structure. The front ring carries the housing of the secondary mirror assembly and
the heat rejection wedge including radiators. On its front side facing towards the Sun,
two Sun sensors are mounted. The LISS (Lockheed Intermediate Sun Sensor) and Full
Range Elevation Device (FRED) are part of the gondola pointing and tracking control
system (Section 2.3.3), with the LISS serving as pointing reference in flight. The back
ring (see also Figure 4a) provides a stiff platform for the primary mirror and aperture
stop. On its rear side, the tertiary plane fold mirror is mounted, as well as three blades
providing thermal control of the primary mirror. Front and rear ring are connected to the
central frame with eight struts each, forming the characteristic Serrurier mounting. This
mounting maintains parallelism of the main elements even in case of relative lateral
displacements due to changing gravity loads. The rings as well as the connecting struts
are made of carbon fiber reinforced plastic composite materials for high stiffness, low
 Barthol et al.

Figure 1. Overview of the Sunrise observatory in full flight configuration, immediately before launch. The
dimensions of the observatory can be judged by comparing with the person on the launch vehicle. In the
center of the gondola the telescope front ring can be seen, with the radiators for the heat rejecting prime
focus field stop. The box on top of the telescope is the postfocus instrumentation (PFI) platform housing
the science instruments. A flywheel below the upper bridge of the gondola structure is used for azimuth
pointing. Instrument electronics is mounted on two inclined racks on the rear side of the gondola. The white
box in the lower part of the gondola is the CSBF-provided System Instrumentation Package (SIP), needed
for commanding of balloon and payload. It uses a dedicated power system, while the main power for the
observatory is generated by the large solar arrays to both sides of the telescope.

weight and low thermal expansion. The stiffness of the struts is designed such that
front and rear ring show approximately the same lateral displacement in the presence of
gravitational loads, keeping the relative positioning of M1 and M2 within fractions of a
millimeter. The connection of the struts to the central frame was made by steel shafts,
which allowed easy and quick dismantling of the main telescope parts in the field after
recovery.
The main mirror has been manufactured by SAGEM in France. Schott grade 0 Zero-
dur is used as substrate material. The rear side of the mirror has a three-fold symmetry
with triangular honeycomb structuring, optimizing stiffness under varying gravitational
The Sunrise Mission

secondary focus F2 M4

prime focus F1

M3

primary mirror M1
secondary mirror M2

Figure 2. Optical configuration of the 1 m Gregory telescope of Sunrise .

loads (Figure 4b). Its thickness is 178 mm with a weight of less than 47 kg. This
extreme lightweighting is achieved by mechanical milling and acid etching. Three Invar
mounts connect the mirror with the rear ring of the telescope structure (Figure 4a). On
each mount, the glass-metal junction is formed by 3 Invar pads of 25 mm diameter,
which are glued to the structure behind the face sheet (Figure 4c). The facesheet of
the mirror was ground and polished in a conventional way. As final surface treatment
step, ion-beam figuring was applied. This technique removes residual quilting, which is
unavoidable when manufacturing lightweighted mirrors with facesheet thicknesses in
the order of 7 mm. To further improve the overall wavefront quality, a dedicated map-
imprint was applied during the ion-beam figuring. Based on finite-element mechanical
deformation calculations and analysis of the various thermal load cases, a compromise
was calculated providing best wavefront error (WFE) performance at 22.5◦ elevation,
with only minimal degradation in the range from 0◦ to 45◦ elevation angle, which is the
nominal elevation range of Sunrise . The measured WFE (@633 nm) at 0◦ elevation is
only 30.2 nm rms, reducing to 19.2 nm rms at 22.5◦ and 24.1 nm rms at 45◦ elevation.
The reflective coating on the front face is bare aluminum with a thickness of 100 nm. An
aperture stop with 1000 mm inner diameter directly in front of M1 defines the entrance
pupil of the system.
The primary mirror creates a real image of the sun in the primary focus F1 of the
telescope. Nearly 1 kW solar radiation is concentrated on a disk of about 22 mm diam-
eter. At this position a heat rejection wedge (HRW) with a central hole acts as field stop
(Figure 5).
 Barthol et al.

Postfocus Instrumentation
(PFI)
HRW Radiators M4 Housing

Sun sensors
LISS / FRED

M1 baffle
blades
M2 Housing
Heat rejection Gondola interface
wedge (HRW) (elevation axis)
PFI support
struts

Figure 3. Overall sketch of the telescope Serrurier structure and subassemblies

The 2.8 mm hole – and a slightly undersized stop in the conjugate image plane in
F2 – define the useable telescope field-of-view (FoV), corresponding to 130 Mm on
the solar surface. The HRW is a highly conductive cylindric aluminum block with a
wedged front face reflecting 99 % of the incoming light out of the telescope. The small
fraction of solar radiation passing through the field stop reduces the heat load on the
science instrumentation to about 10 W. Absorption of the focused energy at the HRW
is minimized by a second surface mirror, glued to the aluminum body. A thin glass
plate with vapour deposited aluminum on the rear side acts as optical solar reflector;
UV reflectivity is enhanced by a dedicated coating on its front surface. The size of
the HRW is about twice the diameter of the solar disk image, thus allowing solar limb
observations, while keeping the solar image on the HRW. Two ammonia heat pipes
connect the HRW to dedicated radiators. Equipped with optical solar reflectors – same
second surface mirrors, but without UV coating – and pointing towards the Sun, the
radiators efficiently cool the HRW to temperatures less than 25◦ C, thus avoiding any
Schlieren build-up, which could cause wavefront distortions.
The optical system of the Sunrise telescope is semi-active in order to maintain the
highest performance throughout the flight. The secondary mirror M2, polished and ion-
beam figured to a residual WFE below 6 nm rms, is isostatically mounted on bipods and
connected to a three-axis translation stage. The M2 position is fine adjustable in a range
of 1.2 mm to an accuracy of 5 µm laterally and 1 µm axially, so that the relative M1/M2
alignment can be kept constant even under varying telescope elevation and thermal
loads. A wavefront sensor located in the postfocus instrumentation (see below) monitors
the alignment status and generates control signals for M2 mirror re-positioning. The
The Sunrise Mission

a)
Carbon fiber telescope
rear ring

b)
Rear view of primary mirror
with Invar mirror mounts

c)

Glued pads seen through

the mirror surface

Figure 4. a) CAD model of the main mirror cell with the lightweighted ZERODUR mirror. The mirror
is mounted with three flex blade fixation points to the carbon fiber structure, which forms the rear ring of
the telescope. b) Backside view of main mirror after integration of the three flex blade fixation points. c)
Glass/metal junction formed by glued Invar pads

secondary mirror assembly with its housing creates a central obscuration of 324 mm
diameter for the telescope.
The folding mirrors M3 and M4 are equipped with translation stages. It was orig-
inally foreseen to use them for fine focussing. However, the adjustment accuracy of
the M2 axial translation stage proved to be sufficiently high, so that focussing could
be performed with M2. The WFE increase due to spherical aberration when axially
moving M2 is negligible. M3 and M4 are only adjusted during static alignment on
ground, determining beam height and secondary focus position within the PFI as well
as the lateral position of the pupil image at the image stabilization tip-/tilt mirror.
Thermal control of the primary mirror is essential for the performance of the tele-
scope. About 80 W solar radiation are absorbed in the coating. Dedicated baffle blades
behind the mirror with reflective front sides increase the view factor to the cold sky and
shade the mirror against Earths IR radiation and reflected sunlight from the ground (ice)
or sea below the balloon.
The energy density in the primary focus is high enough to damage structural parts in
case of uncontrolled beam wandering. A retractable curtain in the plane of the central
frame can close the rear compartment of the telescope in case of pointing loss, acting
as an aperture door. The curtain needs about 20 seconds to securely block the telescope
aperture. This is well below critical exposure times of parts in the vicinity of the first
field stop. In-flight operation of the curtain is controlled by two software flags set by
the pointing system computer. Automatic closure is initiated when the pointing error
 Barthol et al.

Radiators

Heat pipe 2

Heat pipe 1

Heat rejection
wedge HRW

Figure 5. CAD model of the primary field stop (heat rejection wedge, HRW) with the two heat pipes, which
drain the absorbed energy to the two radiators on top of the front ring. The radiators are facing direct sunlight
and are therefore covered with Optical Solar Reflectors (OSRs), thin second surface mirrors with high thermal
emissivity, as shown in the insert on the top left. The bottom left picture shows the image of the solar disk on
the HRW during ground testing.

exceeds ±15 arcmin (coarse pointing flag ’off’, solar disk image could leave heat re-
jection wedge). Re-opening is initiated if the pointing remains stable within a cone of
20 arcsec (fine pointing flag ’on’, pointing is within range of image stabilization).
The telescope has been built by Kayser-Threde, Munich, under contract of MPS.

2.2. Postfocus Instrumentation Platform

The Sunrise Post-Focus Instrumentation (PFI) rests piggy-back ontop the telescope.
The compact package consists of a rigid support structure and four instrument mod-
ules with their supporting proximity electronics (Figure 6). Two of the four instrument
modules are ”service” units: the Image Stabilization and Light Distribution (ISLiD) and
Correlation tracking and Wavefront Sensing (CWS) units. The science instrumentation
consists of the Sunrise Filter Imager (SuFI), and the Imaging Magnetograph Exeriment
(IMaX). Mechanism controllers for ISLiD and SuFI as well as proximity electronics and
power supplies for CWS and SuFI are located inside some of the PFI compartments.
The PFI structure provides room for an additional 3rd science instrument for future
flights.

2.2.1. Image Stabilization and Light Distribution System: ISLiD

The Sunrise science requirements demand precision fine pointing and simultaneous
observations by all science instruments. This is ensured by ISLiD, a panchromatic
The Sunrise Mission

CWS prox. SuFI camera

electronics electronics

ISLiD/SuFI

CWS

IMaX

Mechanism
controllers
M4

Figure 6. Semi-transparent view of the Sunrise postfocus instrumentation. Instrument covers are removed.

reimager based on dichroic beamsplitters, which provide the different wavelength bands
to the individual science branches with maximum photon flux, while preserving the
polarization of the incoming light. ISLiD is located in the center of the PFI structure
and takes up its full length (Figure 7). Reimaging of the secondary telescope focus onto
the instrument detectors is achieved with a two mirror Schwarzschild arrangement, be-
fore separating the ultraviolet below 400 nm towards SuFI. Additional refractive optics
provide image scaling and telecentricity for IMaX and CWS, which is fed by light
outside the spectral bands used by the science instruments. ISLiD contains a fast piezo-
driven tip-tilt mirror at a pupil plane of the optical system. A field lens in the telescope
secondary focus projects the aperture stop of the primary mirror onto the tip-tilt mirror.
In order to allow for the UV part of the solar spectrum to be transmitted, the field lens
is made from fused silica and uncoated. The tip-tilt mirror is part of the CWS system
(see below) and used to compensate for residual image motion due to solar rotation or
gondola shake and vibrations within the instrument. The sizing of telescope and instru-
ment FoV’s allows a capture range for the image stabilization system of ±46 arcsec
(Figure 8), still avoiding vignetting instrument FoV’s by the secondary field stop in F2.
Details of ISLiD (developed by MPS) are given by Gandorfer et al., 2010. The image
stabilization system and its damping characteristics are described in Berkefeld et al.,
2010 and in the following section.

2.2.2. Correlating Wavefront Sensor: CWS

The CWS, a Shack-Hartmann type wave front sensor, is located close to ISLiD in the
central compartment of the PFI structure. The high speed camera system has a FoV of
 Barthol et al.

SuFI CCD and filterwheel

to CWS
Tip-tilt mirror

Telescope F2
to IMaX

Telescope M4

Figure 7. The Image Stabilization and Light Distribution unit (ISLiD) is mounted into the central compart-
ment of the PFI structure. It distributes the light coming from the telescope (directions shown in red) onto the
science instruments and the CWS.

12×12 arcsec2 on the sky. The CWS is used for two purposes, for (high frequency)
precision image stabilization and guiding and to control the alignment of the tele-
scope (low frequency). A lenslet array in a pupil plane has 6 subapertures arranged
in a concentric ring, forming 6 independent images on the detector. The information
derived from the 6 independently analyzed images of the same solar scene are used to
measure the local wavefront tilt per subaperture, providing the coefficients of a Zernike
polynomial decomposition of the wavefront error up to the third radial degree. The
coefficients for tip and tilt, defocus, and Seidel coma are used as error signals. A control
loop time-integrates these error signals and converts them into actuation signals to drive
the telescope secondary mirror M2. The fast read-out of the CWS camera (>1 kHz)
allows detecting correlated image motion of the 6 separately generated images on the
detector caused by residual uncompensated gondola movements and vibration as well
as by the slow drift of solar features due to solar rotation. Fast software routines convert
the correlation signals to actuator signals for the tip-tilt mirror, performing the pointing
correction, image stabilization and guiding. The closed-loop control of the image sta-
bilization system has a bandwidth of about 60 Hz (at the 6dB level) with a sensitivity
of better than 0.003 arcsec. It provides very efficient damping of low frequencies (∼600
at 2 Hz, ∼70 at 10 Hz), where the pointing control cannot be provided solely by the
gondola. The CWS was activated whenever the gondola pointing was within its nom-
inal range (see Section 2.3.3.). The CWS unit including the tip-tilt mirror, electronics
and the control software were developed by the Kiepenheuer-Institut für Sonnenphysik,
Freiburg, Germany. Details can be found in Berkefeld et al., 2010.
The Sunrise Mission

Figure 8. Telescope and instrument fields-of-view (FoV). a) Continuum image of the solar disk. A circle
indicates the Sunrise telescope FoV of 180 arcsec diameter, corresponding to roughly 10% of the solar disk
diameter. b) Co-aligned instrument FoV’s of IMaX, SuFI and CWS within the telescope FoV. The free range
of the image stabilization system (±46 arcsec) is indicated.

2.2.3. Sunrise Filter Imager: SuFI

SuFI provides high-resolution images of photospheric and chromospheric features at

5 wavelength bands in the visible and near ultraviolet at high cadence. The optical
arrangement is a Schwarzschild 2-mirror system (part of ISLiD), magnifying the tele-
scope secondary focus by a factor of 5 onto the detector, thus resulting in an effective
focal length of 121 m. The sensor is a 2k by 2k UV enhanced fast read-out CCD. The
field-of-view of the instrument is 15×40 arcsec2 . Interference filters are employed to
isolate the observed wavelength band. They have peak transmissions at wavelengths of
214 (10) nm and 300 (5) nm together with OH and CN molecular absorption bands at
312 (1) nm and 388 (0.8) nm and the Ca II H line (singly ionized Calcium) at 397.6
(0.18) nm. The numbers in brackets are the filter FWHM. In order to achieve near
diffraction-limited imaging, a phase-diversity imaging technique is used by splitting
the image in front of the CCD detector: half of the CCD area collects the focused
image, while an optical delay line forms a second image with a defocus of approxi-
mately one wave on the other half of the CCD. Post-flight restoration of the images
free from static aberrations of the optical path can thus be achieved. The SuFI optical
elements, mechanisms and CCD are mounted side-by-side with ISLiD components on
one combined optical bench in the central compartment of the PFI structure. Details of
the SuFI instrument, which has been developed by MPS, are given by Gandorfer et al.,
2010.

2.2.4. Imaging Magnetograph eXperiment: IMaX

IMaX is an imaging vector magnetograph for observations of Doppler shifts and polar-
ization in the Zeeman-sensitive photospheric spectral line of neutral iron at 525.02 nm.
Images are taken in up to twelve wavelength bands. The instrument provides fast-
cadence, high-spatial resolution two-dimensional maps of the magnetic vector, the line-
 Barthol et al.

Figure 9. a) Integrated PFI Support Structure without instrumentation mounted on optical bench, attached
at all 7 interface points. b) Fully assembled PFI in thermal insulation ready to be integrated onto telescope.

of-sight velocity, and continuum intensity. It has a field-of-view of 50×50 arcsec2 ,

which is the largest of all instruments on Sunrise . A tunable LiNbO3 solid state Fabry-
Pérot etalon is used in double pass, thus minimizing mass and power consumption
and relaxing the requirements on passband stability. Since the free spectral range of
such a system is quite small, a narrowband interference filter (FWHM 0.1 nm) is used
in addition. Both, prefilter and etalon are thermally stabilized. Imaging is done with
two synchronized 1k×1k CCD cameras in orthogonal polarization states. Two nematic
liquid crystal retarders modulate the incoming polarization. Four switching states are
needed for full Stokes vector polarimetry, while two-states are sufficient for longitudinal
magnetometry (circular polarization). IMaX was developed by a Spanish consortium
led by the Instituto de Astrofı́sica de Canarias, La Laguna (Tenerife), in cooperation
with the Instituto de Astrofı́sica de Andalucı́a, Granada, the Instituto Nacional de Tec-
nicas Aeroespaciales, Madrid, and the Grupo de Astronomia y Ciencias del Espacio,
Valencia. The instrument is described by Martı́nez Pillet et al., 2010.

2.2.5. PFI Support Structure

The PFI structure shown in Figure 9 provides a stable platform for the science instru-
mentation. The individual instruments must be kept co-aligned to better than ±0.1 mm.
This is a demanding requirement considering instrument masses close to 50 kg (in case
of IMaX) and in view of the varying gravity and thermal loads during the mission.
In addition, the combined PFI package has to maintain the alignment with respect to
the telescope exit beam entering the PFI via M4. Similar to the telescope, carbon fiber
based composites and honeycomb structures are used for the mechanical frame. This
choice provides high stiffness and low thermal expansion, together with low weight.
The PFI structure has a width of approximately 1.4 m (compatible with the width of
the telescope) and a length of about 2 m, suitable to accommodate the instruments. The
height of 350 mm resulted from a trade-off between keeping the PFI center-of-gravity
The Sunrise Mission

(CoG) as low as possible and taking advantage of the increasing bending and torsional
stiffness with height.
Four outer panels form a box-like structure. Two inner panels spanning the full length
of the platform provide additional support for the inner instruments and for the plane
folding mirror M4, which is attached to the rear PFI panel. Each panel is a combina-
tion of a 30 mm thick aluminum honeycomb core with carbon-fiber-reinforced plastic
face sheets of 1.25 mm thickness. Dimensional stability of the panels is provided by
a (±45◦ , 0◦ /90◦ , ±45◦ ) orientation of the individual Torayca M40J fiber-layers of each
facesheet. The coefficient for thermal expansion (CTE) of the material is on the order
of −0.3×10−6 per Kelvin. To increase bending stiffness each panel is reinforced at the
two long sides by tubes with unidirectional fiber orientation. Tenax UMS2526 fibers
with an exceptionally high tensile strength of 190 GPa are used. The tubes have 30 mm
diameter and 1.5 mm wall thickness each.
Additional cross panels stiffen out the compartment for the heavy IMaX instrument.
Shear stability of the box is provided by two panels of 10 mm thickness, closing the
bottom end towards the telescope. A central carbon fiber based stiffener forms the
central mounting point to the telescope central frame. All connections are realized by
steel screws and 2-parts aluminum inserts, which are inserted into the sandwich panels
from both sides and secured by glue. The total weight of the PFI structure is 68 kg
including the mounting brackets for the instrumentation.
The center of gravity of the PFI is located directly above the steel telescope central
frame (Figure 3c). A spherical pin at the upper center of the telescope central frame is
used to fix all three translational degrees of freedom (DoF). Two flexural joints at the
outer uppermost ends of the telescope central frame fix the rotational DoF around the
Y-axis and Z-axis (direction to Sun), but allowing differential thermal expansions of
PFI and telescope structural elements in X-direction, caused by the CTE mismatch of
steel and the carbon fiber matrix.
Carbon fiber based struts support all four edges of the PFI, fixing the rotational DoF
around the X-axis (elevation axis). These struts have length compensators eliminating
any bending forces on the PFI structure in case of temperature inhomogeneities between
the telescope central frame and PFI (see Figure 3).

2.2.6. Instrument Mounting

All instrument units are individually assembled, aligned and functionally as well as
environmentally tested as separate modules before being integrated into the PFI support
structure. With the enormous effective focal length of SuFI of 121 m even a mini-
mal mechanical deformation on the order of only a few micrometers already would
transform into a considerable image shift on the SuFI detector. The required stability
cannot be provided by the 5 cm thick optical bench alone. Increasing the overall sta-
bility of this sensitive arrangement, ISLiD/SUFI makes use of the high stiffness of the
PFI support structure by being rigidly connected to the innermost side panels over its
full length. L-shaped brackets connected to the lower side of the ISLiD/SuFI optical
bench (Figure 7) provide a stiff mounting, but allowing for local deformations in the
vicinity of the attachment points, not propagating to the surface of the optical bench
or to the optical elements fixed to it. Differential thermo-mechanical expansions of the
ISLiD/SuFI optical bench relative to the PFI platform are minimized by using face sheet
material from the same lot for both applications.
 Barthol et al.

IMaX and CWS are mounted based on an isostatic concept, allowing the individual
alignment of each instrument module as such with respect to the exiting beam coming
from ISLiD. This concept has the advantage that optical tolerances in focus position
and exit beam orientation, as well as unavoidable manufacturing tolerances of the large
PFI support structure can be easily compensated. The isostatic mounting provides a
stress-free fixation of the instruments even in the case of temperature differences and
materials with different CTEs. This is especially important for IMaX, which uses an
aluminum optical bench.
The correct angular orientation of IMaX and CWS with respect to the exiting beams
of ISLiD is achieved by the use of reflective alignment reference cubes. Dedicated
observation openings in the PFI structure panels allow theodolite autocollimation mea-
surements to arcsec precision.
A 3D coordinate measurement device (Leica Laser Tracker) led to a mounting accu-
racy of better than 0.1 mm.

2.3. Gondola

The Sunrise gondola is responsible for the precision pointing of the telescope towards
the Sun. The mechanical structure serves as stable platform and protects the scientific
instrumentation during launch and landing. Photovoltaic arrays generate the required
electrical power.

2.3.1. Gondola structure

The core gondola structure consists of an aluminum/steel tube framework (Figure 1),
making it relatively lightweight, but providing the required stiffness and a sufficiently
high eigenfrequency (>10 Hz). The structure permits a telescope elevation range of -
5◦ to +50◦ , leaving some margin for pendulum motion during flight. The structure can
be split in two halves at the level of the telescope elevation axis (Figure 10), resulting
in two U-shaped components. This allows a convenient integration of the science in-
strumentation and simplifies shipment of the bulky equipment. Each of these two units
consists of two aluminum side trusses with triangular cross-section; a welded steel-
based bridge connects the two. Front and rear roll cages are attached to the framework
structure. They protect the protruding front and rear ends of the telescope (Figure 1).
Below the bottom of the framework, the crash pad assembly forms the mounting base
for the cardboard landing-shock absorbers. This frame houses the CSBF provided SIP
(System Instrumentation Package), the communication electronics needed for com-
manding to the instrument and balloon, as well as telemetry through TDRS and Iridium
satellite links.

2.3.2. Power system

Electrical power is provided by photovoltaic arrays on the front side of the gondola,
sufficiently far away that the nearly 100◦ C hot panel surfaces do not generate seeing
effects (Figure 1). An inclination angle of 22.5◦ optimizes the orientation for the ex-
pected solar elevation range (0◦ to 45◦ ) during the flight. Each array consists of five
subframes, each with 80 A-300 cells, produced by Sunpower Corporation. The panels,
The Sunrise Mission

Figure 10. Sunrise gondola main structure during integration of the science payload. The two cylinders in
the upper left side truss contain the science data storage systems.

assembled by Meer Instruments, San Diego, USA, generate approximately 1.3 kW in

total. Two lithium-ion battery packs with a capacity of 2500 Wh each act as buffers for
the operation of the observatory, before the correct pointing is achieved. The battery
packs are mounted into the lower bridge of the core framework.

2.3.3. Pointing system

The major components of the gondola pointing system are schematically displayed in
Figure 11. Actuators and their encoders are shown in pale red, solar sensors in yellow.
 Barthol et al.

“Truck plate” (cable attachment

“Ladder” (flight train) interface and launch hang point)

Precision Azimuth Sun Sensor Coarse azimuth motor & main bearing
PASS (fixed to gondola
– frame) M
Corner cell solar sensors (4)
Fine azimuth encoder (fixed to gondola frame)

E
M Fine azimuth motor
Reaction wheel (fine azimuth control)
FRED LISS Lockheed Intermediate Solar
Sensor LISS (fixed to telescope)
Full Range Elevation Detector
FRED (fixed to telescope)

SUNRISE Telescope Telescope elevation

bearings

Elevation control linkage Elevation encoder

Gondola Frame Elevation linear stage & motor

M E

Figure 11. Basic Sunrise gondola components

Actuators: Azimuthal control of the gondola is performed via a two-stage momen-

tum transfer unit at the top. The coarse azimuth drive decouples the instrument from the
rotating balloon, while the fine drive rotates a reaction wheel. Acceleration or decelera-
tion of the wheel rotates the gondola towards the desired orientation. Nominal rotation
rate of the reaction wheel is 10 rpm, helping to overcome friction effects in the wheel
bearing. The actuation between the fine azimuth motor (the reaction wheel) and the
coarse azimuth motor is divided by frequency. The coarse azimuth motor manages the
frequency range below 0.1 Hz. The reaction wheel compensates torques from ∼0.1 to
∼1.0 Hz. While the coarse azimuth motor has a very large compensation range (it can
apply about 13.5 Nm of torque for an indefinite period of time), the fine azimuth motor
can apply an equal torque to the payload only for a limited period. Tracking is kept and
the fine azimuth servo system does not saturate, as long as disturbances to the gondola
– like wind gusts – induce less or equal torque for periods of less than ∼1 second.
The elevation of the telescope is adjusted by a linear translation stage, rigidly coupled
to the central frame of the telescope by a lever arm. Slight preloading of the lever arm
minimizes hysteresis for the elevation adjustment. Several elevation drive mechanisms
were considered in the early design phase, including torque motors mounted to the
elevation axle. The linear stage was chosen because it provides high precision and
a relatively non-compliant, symmetric loading on the central frame of the telescope.
The disadvantage of this system is additional weight over a direct drive torque motor.
However, the chosen solution serves as a launch and landing lock for the telescope in
(horizontal) stow position, without the need of an additional mechanism.
Three types of sensors with different accuracy levels feed the various control loops
used to acquire correct solar pointing.
The Sunrise Mission

Coarse gondola orientation: Photovoltaic cells with wide-field, overlapping accep-

tance cones of 60◦ in elevation and 105◦ in azimuth are fixed to the gondola structure
top at four corners. The right quadrant is derived from comparison of intensity levels,
independent of solar elevation.
Medium azimuth & elevation detection: The Precision Azimuth Sun Sensor ’PASS’
is also fixed to the gondola structure top. The sensor is used as medium resolution
device, locating the Sun in azimuth with a linear capture range of ±3◦ for any given
solar elevation. For medium resolution detection of the Sun’s elevation, a similar device
is used. The Full Range Elevation Detector ’FRED’, mounted to the telecope front ring,
can find the Sun within an azimuth range of ±5◦ . This elevation sensor has a linear
range of ±15◦. For angles outside this range, it gives a saturated signal indicating that
the Sun is above or below the current telescope elevation.
High resolution solar position sensing: Highest accuracy in determining the solar
position is provided by the LISS (Lockheed Intermediate Sun Sensor), mounted to
the Sunrise telescope front ring. The LISS is used within a ±3◦ range in azimuth and
elevation, where the output signal is proportional to the off-pointing from the LISS
optical axis. The high resolution linear range is only ±15 arcmin. Electronic noise and
signal drifts with temperature are low enough, so that accuracies of 1-2 arcsec can
be achieved. The LISS is mounted on a two axis motorized tip-/tilt stage. Offsets in
telescope pointing for solar limb observations and instrument flatfielding are generated
by commanding position offsets to the LISS motors.
The medium and high resolution sensors PASS, FRED and LISS are shadow sensor
type detectors. In case of the LISS, solar radiation passes a square aperture entrance
window, falling on a set of 5 photodiodes. A central diode is used as a binary ’Sun
present’ indicator. The remaining 4 sensors are arranged in two pairs, oriented in az-
imuth and elevation direction. Each pair is connected to a transconductance amplifier,
which amplifies the difference in photocurrent. When the sensor is tipped, the amount
of light falling onto the two photocells (in a given axis) gets out of balance and produces
a signal.
Acquisition of solar pointing: The acquisition of solar pointing and tracking is com-
pletely automatic. Several software threads handle data collection, processing, actuator
stimulation and data output generation. The ’auto-pointing thread’ makes the decisions
of which Sun sensors to use and what servo control loops to engage. The ’interrupt
thread’ runs at 150 Hz update rate and does all the real time pointing and data collection.
It also records a boxcar array of sums and sums squared of all Sun sensor information
and voltages for the elevation and coarse and fine azimuth motors. A separate thread is
constantly checking thresholds to determine wether the pointing system is locked on the
Sun and can set the ’PS-lock’ flag for the CWS image stabilization system (see Section
2.2.2.).
The auto-pointing thread uses the sums and sums squared to compute running aver-
ages and standard deviations squared of these parameters. The running averages and
standard deviations are computed based on the most recent 1024 values collected,
(1024 samples covers 6.8 seconds of data). The auto-pointing thread checks every
9 seconds all instantaneous values, running averages and running standard deviations.
Then, using carefully adjusted threshold criteria, it decides which acquisition sequence
needs to be activated.
The threshold criteria and activation of servos depend on which solar sensor is get-
ting the best signal. Each pointing sequence aims to optimize the positioning for the
 Barthol et al.

next higher resolution sensor. The automated acquisition uses the following sequential
steps with dedicated control loops:

• Coarse azimuth: Coarse+fine azimuth drives to find correct azimuthal orientation

based on corner cell signals, if gondola rotation rate is less than 2 rpm
• Fine azimuth: Coarse and fine azimuth drives to position gondola based on PASS
data
• Coarse elevation: Telescope elevation is adjusted according to FRED data, until
LISS sensor indicates Sun presence
• Final tracking: Control loops closed in azimuth and elevation, based on LISS data

Final tracking is achieved by switching from ’FRED’ to ’LISS’ for elevation tracking
and from the ’PASS’ to ’LISS’ for azimuth tracking. This switching in azimuth and
elevation is done separately to accommodate which sensor gets to a good signal first. In
azimuth the pointing system uses a ’multiple-input-multiple-output’ (MIMO) approach.
The final servo control loop topology is shown in Figure 12. The filters settings applied
in this mode have a higher gain and are designed for a better tracking accuracy. The
servo loops use a combination of integrators and phase-lead compensators with up to
four filters per loop. In case of the ’AzfTrack’ servo loop (see Figure 12) a combination
of a second order low pass filter, then a first order lead filter and lastly a first order
integrator filter is applied to the LISS azimuth output, driving the reaction wheel.
The design requirement for the tracking accuracy is ±7.5 arcsec rms, to safely keep
the system within the capture range of the image stabilization system, provided by the
tip-tilt mirror and CWS (±46 arcsec).
The automated acquisition of the Sun from a non-pointing mode to ’pointing system
lock’ is typically achieved within less than 10 minutes.
Matlab models were used to try to predict the behavior of the full flight train versus
the suspension used during ground testing. The main mitigation was to not use posi-
tional feedback on the coarse azimuth drive. Such feedback gives rigid coupling to the
flight train and thus the vibrational modes of the ladder are very well coupled to the
gondola. Torque feedback was used instead (simple current to the motor), assuming
that the main rotator bearing was good enough that ladder modes would be decoupled.
Servo filter settings derived from the mathematical models were uploaded during the
commissioning phase, being fine-tuned throughout the mission.

2.3.4. Gondola mounted components/systems

On the rear side of the gondola, shaded by the solar panels from direct Sun illumi-
nation, the instrument control electronics are located on two racks. One rack houses
the instrument control unit ICU, the payload power distribution unit and the instru-
ment computers for CWS, IMaX and SuFI. The other rack carries the pointing system
computer, the gondola power distribution unit, amplifiers for the azimuth and elevation
drives, as well as the ESRANGE-provided E-Link high-speed telemetry electronics.
The racks are inclined with respect to the structure in order to minimize radiative input
from the Earth and the hot solar panels onto the electronics, while maximizing the
radiation of heat to the cold sky above the instrument. The two data storage containers
collecting data from the instruments are mounted well secured inside one of the upper
The Sunrise Mission

Motor Command Signal

M

AzfToAzc High-Pass
Input
Wheel Velocity d
dt
E
M “Matched
Filter”

FRED LISS AzfTrack Low-Pass

Input
ElTrack
LISS EL LISS AZ

SUM
M E
ElLocal
El Stage Velocity

Figure 12. Sunrise final tracking servo control loop configuration. Red lines represent signal paths and tri-
angles are signal processing (filtering). Yellow elements are solar sensors. Motors and encoders are shown in
pale red. The label ’AzfToAzc’ denotes the servo loop monitoring the rotational velocity of the fine azimuth
drive and providing feedback to the coarse azimuth motor to eventually slow down the reaction wheel.
Similarly, ’AzfTrack’ and ’ElTrack’ denote fine tracking servo loops. ’ElLocal’ transforms input values to
calibrated velocity commands for the elevation drive.

side trusses of the core framework. The truss framework provides protection, but also
allows easy access for recovery after landing. A spring-based shock protection system
ensures the mechanical integrity of the data storage containers. To avoid oscillating
masses and adverse effects on the pointing control loops, the data storage containers
are rigidly clamped during the mission. A release mechanism frees the containers at
mission termination.
The commanding and communication package provided by CSBF is located under-
neath the gondola structure (Figure 1). It has separate solar panels at all four sides
of the gondola to stay operational in case of pointing loss. The package allows com-
manding and housekeeping downlink via TDRS and Iridium satellites. Shock absorbing
cardboard crash pads at the bottom of the gondola reduce mechanical loads during
touch-down and landing. Two ballast hoppers at the center of the gondola bottom carry
about 650 kg of fine steel grains. Part of the ballast is dropped during ascent, speeding
up the balloon again after it has cooled down in the tropopause transit. The rest of the
ballast is used to compensate losses in float altitude due to the day/night cycles. Two
booms at the top of the gondola carry the satellite communication antennae.
The complete payload has dimensions of 5.5 meters in width and length and is
about 6.4 meters high. The gondola structure, the power and pointing systems have
been developed by the High Altitude Observatory (HAO), NCAR, Boulder, USA.
 Barthol et al.

2.4. Electronics Architecture

The largest part of the Sunrise electronics is located on two racks mounted left and
right of the gondola structure. Only proximity electronics such as mechanism con-
trollers or the voltage supply for the piezo-driven tip-tilt mirror are located close to
the optical modules inside the PFI. Commercial-off-the-shelf products are used as far
as possible. These products would typically not survive the environmental conditions of
a balloon flight. Critical items therefore were encapsulated in pressure vessels, modified
or specifically qualified for this type of application.
Sunrise is designed for autonomous operation similar to a spacecraft. The basic
architecture consists of an instrument control system communicating with dedicated
subsystem- or instrument-related electronics, and telemetry systems for commanding
and downlink of system status information.
The instrument control system consists of the Instrument Control Unit (ICU), two
data storage subsystems, the PFI and gondola power distribution units, and the line-of-
sight telemetry subsystem (E-Link).
The ICU is the central onboard computer. It supervises the various subsystem and
instrument control computers:
• Pointing system computer
• Gondola power distribution unit
• CWS electronics unit
• Main telescope controller
• SuFI electronics unit
• IMaX main electronics
• Payload power distribution unit
• E-Link
• PFI mechanism controllers (also used for PFI thermal control)
The instrument control system carries out the following tasks:
• interpret and route telemetry commands to subsystem and instrument computers
• acquire housekeeping data
• store housekeeping and science image data on data storage units in RAID5 format
• pre-select and route housekeeping and science image thumbnail data to downlink
telemetry subsystem
• initialize, configure and run the Sunrise instrument automatically in standard ob-
servational mode, if no observation mode is given by telecommand
• run predefined timeline-controlled observations
• control electrical power distribution to scientific instruments and subsystems
• monitor housekeeping data and take action in case of limit violations
• provide line-of-sight telemetry
The Instrument Control Unit (ICU) is housed in a pressurized container. It is based
on a NOVA 7800 P800 Pentium III single board computer with interface cards pro-
viding ethernet and serial interfaces. A 2 GB flash disk serves as boot device. Forced
convective cooling by regulated fans keeps the component temperatures within spec-
ified ranges. Onboard communication between ICU and subsystem and instrument
computers is performed via a 100base TX ethernet and serial RS232 resp. RS422 links.
The Sunrise Mission

The science data are stored onboard in two stacks with 24 (100 GByte) harddisks
each with about 3.6 Terabyte net capacity at RAID-5 functionality. This corresponds
to the expected data volume acquired in a 2 weeks mission. The harddisks are encap-
suled in two pressurized vessels (see Figure 10), maintaining the required environment
regarding temperature, pressure and humidity. The data from the ICU to the DSS are
transferred via IEEE1394 A (Fire Wire) links at 400 Mbit/s.
The commanding and telemetry from and to the ground station is handled via the
CSBF-provided System Instrumentation Package (SIP). The SIP provides simple pulse
command channels for direct instrument control. The pulse commands are used for ICU
reset control and for payload power switch off in case of emergency or to release S/W
induced deadlocks.

2.5. Telemetry

Operations control for Sunrise is performed at two different locations. All balloon sys-
tem relevant issues such as flight monitoring and tracking, ballast drops and mission
termination are handled by the Operations Control Center (OCC) in Palestine, Texas,
USA. Sunrise science operations are directly controlled and monitored from system
and instrument EGSE (electronic ground support equipment) computers at the Remote
Operations Control Center (ROCC) located at ESRANGE. All command activities at
ESRANGE are closely coordinated with CSBF personel present on site.
Commanding and downlink of telemetry data are provided through different com-
munication systems depending on the mission phase.
During the first hours after launch, while the instrument still is within the line-of-
sight, the E-Link high speed telemetry system is used. This communication system
has been developed by ESRANGE and is available on a rental basis. It operates at
frequencies around 2.4 GHz and acts as a transparent Ethernet connection to the ICU,
providing simultaneous up- and downlink rates of up to 2 Mbit/s. The range is limited
to approximately 350 km. The ground station antenna tracks the instrument position
autonomously using the GPS information provided onboard Sunrise . To enhance the
range for line-of-sight communication, a second ground station was set up in Andenes,
Norway.
‘Over the horizon’ communication with Sunrise is achieved by a Tracking and Data
Relay Satellite System (TDRSS) communication channel. Commands generated by the
Sunrise EGSE computers at ESRANGE are transferred via internet to a science relay
computer at the OCC in Palestine, Texas ( Figure 13). The OCC is connected with the
TDRSS ground station in White Sands, New Mexico, where the Sunrise commands are
uplinked to the TDRS and finally relayed to the SIP antennas onboard Sunrise. The SIP
transfers the commands via a serial link to the ICU. Data generated onboard, such as
instrument housekeeping or thumbnail images, are relayed from the instrument back to
the ROCC along the same path. A data rate of 6 kbit/s is available almost permanently
for the downlink.

2.6. Balloon System

The science goals of Sunrise profit from as high a float altitude as possible, especially
to allow observations in the UV down to 214 nm. Therefore one of the largest balloons
 Barthol et al.

TDRSS
satellite

Remote operation
control center (ROCC)
ESRANGE, Sweden
SUNRISE instrument

Internet
SUNRISE
system EGSE
Serial
cmd/data
Science
relay computer
TDRSS ground station CSBF operation control center
White Sands, NM, USA (OCC), Palestine, TX, USA

Figure 13. ’Over the horizon’ communication via TDRSS satellite link

regularly flown by CSBF has been chosen for Sunrise , an Aerostar zero-pressure bal-
loon with a volume of 34.43 million cubic feet (975.000 cubic meters) and a diameter
of close to 134 meters. In case of Sunrise this balloon type lifts a total mass of about
6 tons to stratospheric altitudes of more than 37 km. The scientific payload with a
mass of 1920 kg contributes only one third to the total weight. The rest is given by the
balloon film (2330 kg) with its helium filling ( 500 kg), auxiliary equipment as suspen-
sion, parachute, crush pads, ballast hoppers (Figure 14) and finally ballast (544 kg) for
altitude stabilization during flight.

2.7. System Analysis and Design Aspects

The optical performance of Sunrise must be guaranteed under various thermal and me-
chanical load cases. In order to stay within the required optical tolerances, which were
defined by a sensitivity and tolerance analysis on telescope and instrument level, the
Sunrise design was based on a thorough analysis of the expected thermal and mechani-
cal loadcases.

2.7.1. Structural Design

A structural analysis and iterative optimization of the complete instrument with all
components of the gondola, telescope and PFI has been performed with the aim of
maximizing stability and stiffness, while minimizing weight. The NASTRAN finite
element software package has been used. Several requirements had to be considered,
such as instrument safety, alignment and pointing stability:
The Sunrise Mission

Figure 14. Sunrise immediately before launch. The balloon is only partly filled on ground to allow expansion
with decreasing outside air pressure during ascent. At 37 km altitude it will have expanded by a factor of 300
as compared to ground. The parachute for landing is part of the flight train connecting the gondola payload
with the balloon.

- Safety aspects are important when bringing a payload of several tons to strato-
spheric heights. CSBF requests to analytically prove instrument integrity with respect
to mechanical loads induced at launch and flight termination. All components there-
fore have sufficient margin-of-safety to structural failure, for instance regarding vertical
shock loads of up to 10 g, expected during parachute opening.
- Mechanical loads induced during transport, ground handling and launch must not
lead to permanent misalignment of structural components or instrument units. The de-
sign of all components and fixation devices considers a quasistatic load of 4 g, leading
to elastic deformations only.
- In contrast to a space instrument, Sunrise is subjected to gravity when in operation.
The orientation of the telescope and instrumentation relative to the gravity vector varies
significantly with the changing elevation of the Sun. This could affect the relative align-
ment of structural components or instruments. The telescope Serrurier structure as well
as the PFI support structure provide sufficient structural stiffness to maintain optimal
instrument performance. A few residual thermo-elastic deformations cannot be con-
trolled purely by passive structural design. Therefore, the critical relative positioning of
telescope primary and secondary mirror with a sensitivity in the micro-meter range is
actively adjusted inflight, ensuring the high performance image quality of the system.
- The servo control loops providing the gondola pointing assume rigid body motions
when activating the azimuth and elevation drives. Rigid body motion is only given when
the lowest eigenfrequency of the moving parts is considerably higher than the excitation
frequencies. As the elevation and azimuth drives operate at frequencies below 10 Hz,
the gondola as largest structure has been designed to a minimum eigenfrequency of the
 Barthol et al.

order of 15 Hz. Decoupling of eigenfrequencies avoids excitation of one sub-system by

another. The PFI has been designed to a minimum eigenfrequency of 25 Hz and the
telescope to 35 Hz.

2.7.2. Thermal Design

Thermal analysis and thermal design is important to keep all instrument components
within their specified temperature ranges, under all conditions. This is especially rele-
vant for the high power dissipating commercial electronics and alignment-critical op-
tics. The environmental conditions for Sunrise at stratospheric altitudes are very differ-
ent compared to ground. Owing to the low pressure of only a few millibars, convective
coupling is no longer the dominant heat exchange mechanism and the thermal behaviour
of the instrument is controlled practically only by radiative heat exchange. Component
temperatures result from the equilibrium of dissipated, absorbed, and emitted energy.
Surface properties such as absorption and emission coefficients as well as view factors
of surface elements with respect to their neighbourhood and to heat sources and sinks
are dominant factors.
During flight, Sunrise is exposed to a changing environment. The albedo and thus
the heat input from below significantly varies when flying over sea, ice or cloud layers.
In addition, the radiative input varies with solar elevation angle.
Two extreme steady state load cases covering the wide range of conditions to be
expected were defined:
• a ’hot’ case with 45 deg solar elevation angle, a solar flux of 1397 W/m2 , an
albedo of 0.95 simulating complete ice coverage, and Earth IR radiative input of
264 W/m2
• a ’cold’ case with 0 deg solar elevation angle, a reduced solar flux of 1044 W/m2 ,
an albedo of 0.11 simulating sea water below the instrument, and Earth radiative
input of 156 W/m2 .
Sunrise with all components was modelled with several thousand nodes using the ESA-
RAD/ESATAN software package. Geometries, surface properties and radiator sizes
were determined and optimized.
Thermal Control of Instrument Electronics and Structural Components: Heat dis-
sipating elements such as computers, photovoltaic arrays etc. have been placed as far
away as possible from the telescope and instruments. The instrument electronics are
located on two aluminum honeycomb racks of 1×2m2 size, left and right on the rear
side of the gondola. The racks are shaded by, and turned away from the solar arrays,
minimizing the view factor of the electronics to the nearly 100◦ C hot photovoltaic cells.
A 20◦ inclination reduces albedo input and maximizes the view to the cold sky, while
still keeping the moments of inertia of the system low. White paint is used as sur-
face treatment for all the electronics units and passive structural components directly
exposed to the Sun, such as the tubing of the gondola and the telescope mechanical
components. Thermal filler is used underneath some of the electronics, enhancing the
conductive coupling to the top surfaces of the racks so that they serve as additional
radiating areas.
Gondola Blankets: The impact of the changing environment on telescope and in-
strumentation must be minimized to reduce thermo-mechanical deformations. Shading
The Sunrise Mission

solar radiation coming from the extended source below the observatory is provided by
large thermal blankets, which are mounted to the gondola interior sides and bottom,
including front and rear roll cage. The outer blanket layer is facing the Earth and
reduces the heat input to the system. A Mylar foil of 125 µm (5 mil) thickness with
vapour deposited aluminum on one side was chosen. The foil is used as second surface
mirror, providing similar thermal properties as white paint. The innermost layers of
the gondola thermal blankets control the heat exchange with the telescope and science
instrumentation. They are directly illuminated by the Sun during observation. Dunmore
Beta Cloth 500F is used in the front roll cage. This low-outgassing, space approved
material is tailored to form a diffusely reflecting white cavity around the front end of
the telescope, reducing the probability of structural hot spots. The interior of the rear roll
cage is covered with a highly reflective aluminized polyimide (Kapton) foil, enhancing
the radiative exchange of the primary mirror with the cold sky. Here the aluminum side
is facing outwards, maximizing the reflection properties. Heating of the foil is only
moderate, due to the grazing solar incidence on the roll cage sides and the shadowing
by the telescope and instrumentation.
PFI Thermal Design: The thermal design of the postfocus instrumentation places
emphasis on thermal stability. The overall temperature level including gradients across
the 2 meter long structure have to be kept within a small temperature range of 20±10◦C.
This ensures not only invariant alignment for non-carbon-fiber-based instruments as
IMaX, but also minimizes changes of the polarization properties of the optics.
The PFI structure is decoupled from the environment by Styrofoam insulation, a low
outgassing closed-cell material. Panels of 4 cm thickness are used on all 4 sides of the
PFI, and 1 cm panels on the bottom towards the telescope. The panels are tailored to fit
to the structure and wrapped and taped in aluminized Mylar foil of 125 µm thickness,
again used as second surface mirror. The foil wrapping minimizes particulate contami-
nation of the optics, which could occur when using cut foam pieces. The top side of the
PFI has a large view factor to the heat sink of the cold sky above the instrument. Struc-
tural components are covered with wrapped Styrofoam, similar to the PFI sides and
bottom. White painted radiator plates serve as direct link for heat-dissipating proximity
electronics, such as mechanism controllers, drive electronics for the tip-tilt mirror and
SuFI camera head, electronics and power supply (Figure 9b). Fine adjustment of the
expected electronics temperatures was achieved by partially covering the radiator areas
with wrapped foam pieces. Groups of thermostat-controlled and actively controlled foil
heaters were placed on the metallic covers of the ISLiD, SuFI, and CWS instruments
underneath the foam insulation. A heat shield on the illuminated front end shades the
PFI from direct solar radiation and minimizes temperature gradients in addition to the
foam insulation.
Wind Shields: The balloon ascent through the cold tropopause layer is considered
as a risk for the PFI instrumentation and instrument electronics. Convective cooling
in conjunction with the radiative heat exchange can lead to a critical temperature drop
during this phase (see below). Dedicated wind shields were developed to reduce the im-
pact of the cold air stream while not adversely affecting the thermal conditions at float.
A 37.5 µm (1.5 mil) Polyethylene film from the NASA ULDB program was selected
and kindly provided by Aerostar Industries free of charge. This material is optimized
in terms of mechanical durability at low temperature conditions and provides a high
transparency in the visible and the infrared. The wind shield material and mounting
concept was tested to withstand wind gusts and velocities up to 100 km/h.
 Barthol et al.

3. Mission

3.1. Subsystem Test Flights

The gondola system and large parts of the instrument control system including the main
computer (ICU), data storages and power distribution units, performed a 9-hour test
flight on Oct 3, 2007 from Ft. Sumner, New Mexico, USA. The gondola was equipped
with a small UV telescope with 25 cm aperture and a camera system with a filter
wheel, verifying the pointing capabilities of the gondola. Fully powered by batteries,
the gondola carried solar panel dummies, simulating the thermal behaviour and being
representative in terms of size, mass, inertia and aerodynamics. Pointing, onboard data
acquisition, commanding and downlink through telemetry systems were tested. The
payload was recovered largely intact on a wheat field close to Dalhart, Texas, USA, the
following day.
The test flight revealed the sensitivity of some of the electronic equipment when
exposed to the severe environmental conditions around the tropopause. Low tempera-
tures of less than -75◦C (Perez Grande et al., 2007) caused the ICU to temporarily stop
working. The ICU recovered at float altitude and performed flawlessly during nominal
flight operation. For the science flight it was decided to modify the thermal design of
the Sunrise instrument by implementing dedicated wind shields (see Section 2.7.2).
A verification of this concept was performed on a dedicated stratospheric balloon test
flight of electronics equipment in June 2008 from ESRANGE.

3.2. Instrument Integration and Mission Preparations

The two major components of Sunrise , the gondola and the telescope with its in-
strumentation package, have been assembled, integrated and tested separately at their
respective home institutions, before being shipped to ESRANGE for system integration.
The gondola had survived the 2007 Ft. Sumner test flight relatively unharmed. The
refurbishment was done at HAO. A few mechanical components needed replacement
(crash pad assembly and roll cages) and some electrical modifications were imple-
mented at HAO. Significant work was spent improving and testing the pointing control
software.
At MPS the postfocus instrumentation (PFI) including ISLiD/SuFI was assembled
and tested. The two externally provided, pre-aligned instrument units CWS and IMaX
then were integrated and co-aligned relative to ISLiD/SuFI. Final verification on PFI
level included checks of the overall wavefront quality. Polarization properties of the
integrated package as well as the closed-loop performance of the tip-/tilt mirror were
measured.
The telescope was assembled, aligned and tested at Kayser-Threde, Munich, under
contract by MPS. After delivery to MPS end of 2008, all relevant functional tests were
repeated, including wavefront measurements with an interferometric end-to-end test
and a verification of M2 movements to micrometer accuracy. The tests proved the
excellent stability of the telescope system and verified the transportation concept with
a dedicated damped telescope dolly tailored to fit into a standard sea container.
Telescope and postfocus instrumentation were mated for final instrument perfor-
mance tests at MPS. Lasertracker measurements helped to align the PFI package on
The Sunrise Mission

top of the central telescope frame to its nominal position. The four struts supporting
the PFI edges needed careful adjustment to avoid any distortion or bending of the large
structure, in order to reproduce the alignment status achieved earlier on the optical table.
Due to transport size limitations the PFI had to be dismounted from the telescope
again for shipment. Both units were packed separately in vibration damped transport
dollies and stowed in two 20ft sea containers. All equipment from HAO and MPS
arrived end of March 2009 at the launch site ESRANGE.
The Sunrise flight hardware and auxiliary equipment were set up in the integration
hangar nicknamed ’Cathedral’ with approx. 250 m2 floor space. Functional tests of all
units, alignment verification of the PFI and interferometric wavefront measurements on
the telescope proved that all systems had survived the transport without degradations.
PFI and telescope were mated on April 11, 2009, integration into the gondola was done
one week later on April 18, 2009.
Extensive in-door testing was performed with the nearly fully assembled instrument,
being suspended by the hangar crane. Parameters of the pointing system control loops
were adjusted to the real moments of inertia of the flight hardware. A 10 kW tungsten
theater light was used as artificial sun. Although not giving the radiance levels com-
parable to direct sunlight, the response of all pointing relevant sensors such as corner
cells, mid level sensors and the fine guider could be tested and the system response was
optimized. Pointing tests were continued with real sunlight from within the ’Cathedral’
the following days, further refining the control loop performance, now with higher
photon flux on the sensors.
Sunrise instrument ”first light” was achieved on April 30, 2009. Tests verified the
performance of the heat rejection wedge cooling, then first images were taken with the
science instruments. Due to the very poor ground seeing conditions and the absence of
sunspots no image stabilization was possible as no features were visible on the Sun.
However, the exposure times for SuFI, IMaX and CWS could partly be verified in
flatfield images, and the important wavelength calibration of IMaX could be performed.
Autonomous operation of the instruments needs predefined timelines of command-
ing. Those were extensively tested in May 2009, harmonizing and optimizing the vari-
ous software codes of the different instrument units.
A full flight configuration compatibility test was successfully conducted on May 30,
2009 together with CSBF and ESRANGE. Sunrise was moved from the ’Cathedral’ to
the outside, all additional flight equipment was mounted. The test demonstrated electro-
magnetic compatibility of all components and as well mechanical compatibility with
regard to the launch procedures. Sunrise declared flight readiness on June 5, 2009, after
having performed a last sun pointing test with IMaX calibrations.

3.3. Flight June 2009

Sunrise was successfully launched on June 8, 2009 at 6:27 UT (08:27h local time), on
its first launch opportunity. Balloon and instrument reached an initial float altitude of
37.2 km (122.000 ft) after about three hours ascent.
Instrument commissioning started with the gondola pointing system. During ascent,
Sunrise was in a rotary mode, providing solar illumination on virtually all parts of the
instrument and thus some heating during the critical tropopause transit. This mode how-
ever required the whole system to be run on the limited battery power. Having adjusted
 Barthol et al.

Figure 15. Flight trajectory of Sunrise.

relevant parameters of the pointing control loops to the different conditions due to the
longer flight train, the telescope aperture door was opened at 11:41:15 UT. Instrument
check-out and commissioning proceeded with the CWS, closing the loop for image
stabilization and adjusting the focus of the telescope first at 13:20 UT. At that time, first
images were taken by SuFI and IMaX. A full commissioning of the instruments was,
however, impossible due to an early failure of the line-of-sight communication. The 2
Mbit/s high speed telemetry link operated until 15:09 UT, but was unreliable towards
the end. The hand-over to the second ground station in Andenes/Norway unfortunately
failed. This additional ground station would have extended the communication range to
approximately 24 hours coverage, given the low speed and direction Sunrise was head-
ing. The TDRSS link had to be used instead, providing a downlink data rate of approxi-
mately 6 kbit/s. With this low data rate, some of the originally foreseen commissioning
tasks could not be performed as planned.
The observation program collected minimum science data at disk center on June
08 and 09, 2009. On June 10 beginning at 01:36 UT co-alignment of telescope and
sun sensor LISS was checked by searching the solar limb. Instrument flatfielding was
performed by commanding a circular movement of the telescope pointing with fixed
tip-/tilt mirror.
Zonal winds carried the balloon and instrument with almost constant speed of 30 km
per hour to Northern Canada (Figure 15). The balloon altitude varied between 37 km
and 34 km, following the sun elevation with a phase lag of approximately 3 to 4 hours.
Several ballast drops in the second half of the mission helped to regain float altitude.
Flight termination sequences were commanded on June 13, 2009 at 21:20 UT ,
followed by the balloon cut-away at 22:52 UT. Sunrise landed on Somerset Island,
Nunavut County, Northern Canada at 23:44 UT after 137 hours mission elapsed time
and a travel distance of nearly 4350 km.
The Sunrise Mission

Sunrise was completely recovered a few days after landing. All instrument parts
were flown out via helicopter first to Resolute Bay, then with airplanes to Yellowknife,
where the equipment was packed into sea containers for shipment to their home insti-
tutions. Damage to Sunrise was found to be moderate, e.g.the primary mirror survived
the landing perfectly intact. The scientific data stored on the data storage harddrives
safely arrived at MPS on June 25, 2009, being handcarried directly from Yellowknife
on commercial airplanes.

4. Instrument In-Flight Performance

The Sunrise instrument impressed with near flawless performance during its maiden
flight, although instrument commissioning could not be fully accomplished due to the
short period of line-of-sight high speed communication (see previous Section).
Power system, Electronics and Software: All on-board electronics, as power sys-
tem, instrument computers, proximity electronics and their related software, performed
nominally during most of the mission.
The sizing of battery capacity and solar panels proved to be very conservative. Al-
though the time to first Sun acquisition was about two hours longer than anticipated
(see below), the battery charge state did not drop below 77 %. Once correctly oriented,
the solar panels fully recharged the batteries within 10 hours. The system current plus
the charge current to the batteries were provided by only 3-4 out of the 10 panels. The
charge controller automatically deactivated the unused panels.
Some issues were identified concerning the handling of the data storage units. On
the second day (June 9th) the instrument control unit had to be rebooted. It was unable
to recover from a failed write operation to one of the active disks. During the data post
processing, it became obvious that the disk showed signs of malfunction already before
that, since there were files missing from the observations made during the first night.
However, all data were successfully recovered due to the applied RAID functionality.
Thermal: All component temperatures remained well within their operational design
limits and close to their predictions given by the detailed thermal models.
During ascent all components covered by windshields – PFI and electronics racks –
encountered only minimal temperature drops to about 0◦ C, when passing the cold
tropopause layer. Temperatures of unshielded components, as the gondola structure,
the solar panels or the telescope trusses, dropped to about -40◦ C at the same time. At
float, the temperatures for instrument computers mounted to the electronics racks stayed
within 0◦ C and +30◦ C, see Perez-Grande et al., 2010 and Berkefeld et al., 2010. Similar
temperatures were measured for the proximity electronics, as mechanism controllers
within the PFI structure. The optical components and mechanisms inside the PFI in-
strumentation showed temperatures between +5◦ C and +25◦ C with a diurnal variation
of only ∼8◦ C, thus perfectly maintaining the alignment as adjusted in the laboratory.
Several external components showed temperature variations of typically ±10◦ C. They
are mainly caused by the changing albedo flux onto the instrument surfaces, resulting
from changing solar elevation and terrain underneath the gondola.
The varying albedo flux also slightly influenced telescope component temperatures,
although the telescope remained nearly permanently oriented towards the Sun. Fig-
ure 16 shows a plot of three temperature sensors mounted to the Zerodur primary mirror
 Barthol et al.

Figure 16. In-flight temperatures of 3 primary mirror sensors at several positions distributed over the aperture
and of the heat rejection wedge at the primary focus (in red)

at various lateral positions as well as the temperature measured at the field stop in the
primary focal plane (heat rejection wedge). The large mirror showed only negligible
temperature gradients across the aperture of less than 10◦ C. More important, both mir-
ror and heat rejection wedge had very moderate temperatures below +22◦ C throughout
the flight, less than 30◦ C above the free air temperature at float altitude which was
measured to be around -6◦ C. This temperature difference is low enough to exclude the
risk of wavefront aberrations due to ’mirror seeing’.
Wavefront Quality: The optical performance of the instrument can be derived from
data of the CWS instrument, directly measuring the wavefront quality in-flight, as
well as from the phase diversity image analysis and reconstruction. Both data sources
prove the excellent end-to end performance and stability of the telescope and postfocus
instrumentation.
The diurnal thermo-elastic deformations and the varying orientation of the optical
system with respect to the gravity vector lead to small variations in the relative position-
ing of the primary and secondary telescope mirrors. Resulting aberrations, as defocus
and coma, are measured by the wavefront sensor CWS. They can be corrected by sec-
ondary mirror displacements. During flight only focus was adjusted in closed-loop by
continous axial repositioning of M2 in a range of about 100 micrometer. Uncorrected,
this defocusing corresponds to a wavefront error of ±0.5 λ rms. The CWS sensitivity
and the micrometer accuracy of the M2 translation stage helped to achieve focus accura-
cies below 0.01 λ rms (Berkefeld et al., 2010). Lateral displacements of M2 would have
caused a significant image displacement, deteriorating the co-alignment of telescope
and LISS sun sensor. The observed uncorrected coma values were less than λ/10, small
The Sunrise Mission

enough to be handled by the phase diversity algorithms. An additional in-flight coma

correction was therefore not performed.
Analysis of the SuFI images shows a total wavefront error in the order of λ/5 to λ/6
rms at 300 nm, very close to the expected performance, see Gandorfer et al., 2010. Main
contributors are astigmatism and trefoil, most probably originating from the primary
mirror, showing some dependance on elevation angle (Hirzberger et al., 2010a). Similar
results were obtained from phase diversity analysis of IMaX data (Martinez Pillet et al.,
2010). Differential focus between CWS and SuFI as well as CWS and IMaX was found
to be within λ/20.
Pointing Performance: The gondola pointing system worked very reliable, keeping
the instrument autonomously oriented towards the Sun. The first Sun acquisition with
aperture door opening occured 5:15 hours after launch, about two hours later than
expected. The tropopause transit had caused a temporary freeze of the fine azimuth
motor, which unfortunately wasn’t switched to rotary mode during ascent as planned
for continuos heating. Once this system was recovered, part of the commissioning time
with high speed telemetry was used to adapt gain and filter settings, optimizing the
system response on the long flight train.
The wind gusts acting on the gondola were unfortunatley strong enough to reduce
the pointing performance quite frequently. Flight data indicate that performance was
best over the free ocean between Norway and Greenland (June 9) and less good over
Greenland and the many coast lines of Northern Canada (June 13). Higher float alti-
tudes in general seem to be beneficial for the pointing stability, due to the lower air
density. The longest gondola pointing within the ±46 arcsec capture range of the image
stabilization system was 45 minutes. Fourtyseven time series with durations between
10 and 45 minutes were obtained with closed-loop image stabilization (see Berkefeld
et al., 2010). In total more than 33 hours of science observation (23 % of the total
observing time) were obtained. Details on SuFI and IMaX observing modes, number of
acquired images and and duration of time series are given in Solanki et al., 2010. The
telescope aperture door was closed 93 times due to pointing errors exceeding 15 arcmin.
No indication of structural damage to the telescope components in the vicinity of the
primary focus was found.
Post-flight analysis of the image data gives evidence that the imaging performance
of Sunrise is not limited by the wavefront error of the telescope and instrument optics,
but by a residual image smear of the order of 0.03-0.04 arcsec (Berkefeld et al., 2010),
which reduces the spatial resolution in case of SuFI to 0.1 arcsec (Gandorfer et al.,
2010) and to 0.15-0.18 arcsec for IMaX (Martinez Pillet et al., 2010).
Although the CWS system was able to dramatically improve the effective pointing
stability, not all components of the incoming excitation spectrum of the balloon/gondola
system could be damped out.
Figure 17 shows LISS Sun sensor data obtained on June 9, 2009 with 150 Hz
sampling rate. The sample is a representative part of a ∼40 minutes period, when the
gondola pointing remained within the ±46 arcsec capture range of the tip-tilt mirror and
both image stabilization and telescope focus loops were closed. The data clearly show a
slight elevation pendulum motion (pitch), as well as an azimuthal (yaw) oscillation with
a frequency of 10 Hz and several arcsec amplitude. This oscillation most likely can be
attributed to an excited resonance in the mechanical structure of the gondola (bending
 Barthol et al.

[arcsec]

0 1 2 3 4 5 6 Seconds
[arcsec]

0 1 2 3 4 5 6 Seconds

Figure 17. LISS Sun Sensor data (in arcsec) sampled with 150 Hz on June 09, 2009, 23:02:05 UT. The
length of the interval shown is 6.8 seconds, being a representative part of a ∼40 minutes period of closed loop
pointing. A 10 Hz oscillation in azimuth (yaw) can clearly be seen, as well as slight pendulation in elevation
(pitch)

mode of the solar panels), influencing the pointing system. Amplitudes varied over the
the mission and increased by up to a factor of three towards the end.
A Fourier analysis of residual image stabilization data taken on June 12 (see Berke-
feld et al., 2010, Figure 18) clearly shows this 10 Hz peak in the power spectra, but
also reveals additional components at frequencies above 40 Hz, where the image sta-
bilization is already quite limited. Both contributions immediately died out when the
azimuth motors were switched off. This can be documented by accelerometer data taken
at several positions on the gondola structure, showing higher amplitudes at the gondola
top. We therefore currently speculate that the excitation of the 10 Hz oscillation and high
frequency jitter is connected with the azimuth drives, e.g. by bearing rumble. A detailed
analysis of these phenomena is still ongoing, also regarding the relative contribution to
the observed image smear.

5. Outlook

During its 137 hours successful flight Sunrise collected a unique data set. The pay-
load remained in direct sunlight the entire flight. Seeing-free observations were pos-
sible all the time. The total observing time with image stabilization was more than
33 hours, in which SuFI collected nearly 56.000 images in the wavelength range be-
tween 210 nm to 400 nm, being above most of the ozone layer in the Earth’s atmosphere.
The Sunrise Mission

In the blue channels at 214 nm, 300 nm, and 312 nm, the solar surface could be ob-
served for the first time at high angular resolution and with unprecedented intensity
constrasts (Solanki et al., 2010; Riethmüller et al., 2010; Hirzberger et al., 2010b),
thanks to the high optical quality of the instrumentation. IMaX collected high resolution
Doppler- and magnetograms with an unprecedented combination of angular resolution
and magnetic sensitivity (Martı́nez Pillet et al., 2010).
The Sun was extremely quiet during the 2009 science flight. Although the observa-
tion of quiet solar granulation is in itself an important scientific aspect of Sunrise , a
more complete sampling of the different aspects of solar surface magnetism is needed
to finally help us understanding the underlying physical roots of solar activity. Since the
Sunrise instrumentation showed an excellent performance during flight and survived the
landing and recovery with only minor damage, a reflight of Sunrise in phases of higher
solar activity has the potential of significantly advancing our knowledge of the Sun,
overcoming the current performance limitations with relatively low additional effort.

Acknowledgements The German contribution to Sunrise is funded by the Bundesministerium für Wirt-
schaft und Technologie through Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR), Grant No.
50 OU 0401, and by the Innovationsfond of the President of the Max Planck Society (MPG). The Spanish
contribution has been funded by the Spanish MICINN under projects ESP2006-13030-C06 and AYA2009-
14105-C06 (including European FEDER funds). The HAO contribution was partly funded through NASA
grant number NNX08AH38G. We greatly appreciate the tremendous support provided by CSBF for this
mission, especially by Danny Ball. We would also like to thank all current and previous team members not
listed as co-authors for their valuable contribution to the project.

References

Bernasconi, P. N., Rust, D. M., Eaton, H.A.C., Murphy, G.A., 2000, Proc. SPIE 4014, 214
Bernasconi, P. N., Eaton, H.A.C., Foukal, P., Rust, D. M., 2004, Adv. Space Res. 33, 1746
Berkefeld, T., Schmidt, W., Soltau, D., et al., 2010, this volume
Danielson, R. E., 1961, Astrophys. J. 134, 275
Gandorfer, A., Grauf, B., Barthol, P., et al., 2010, this volume
Herse, M., 1979, Solar Phys. 63, 35
Hirayama, T., 1978, Pub. Astron. Soc. Japan 30, 337
Hirzberger, J., Feller, A., Riethmüller, T. L., et al., Astron. Astrophys., 2010, to be submitted
Hirzberger, J., Feller, A., Riethmüller, T. L., et al., 2010, Astrophys. J., submitted
Krat, V. A., Karpinsky, V. N., Pravdjuk, L. M., 1972, Solar Phys. 26, 305
Lemaire, P., Samain, D., 1984, Adv. Space Res. 4, 37
Lemaire, P., Blamont, J.E., 1967, Astrophys. J. 150, L129
Martı́nez Pillet, V., del Toro Iniesta, J. C., Álvarez Herrero, A., et al., 2010, this volume
Mehltretter, J. P., 1978, Astron. Astrophys. 62, 311
Pérez-Grande, I., Sanz-Andrés, A., Bezdenejnykh, N., Barthol, P., 2009, Applied Thermal Engineering 29,
1507
Riethmüller, T. L.,Solanki, Martı́nez Pillet, V., et al., 2010, Astrophys. J., submitted
Rust, D. M., Murphy, G., Strohbehn, K., Keller, C. U., 1996, Solar Phys. 164, 403
Schwarzschild, M., 1959, Astrophys. J. 130, 345
Solanki, S. K., et al., 2010, Astrophys. J., submitted
View publication stats