APEX: Five Years of Operations
Andreas Lundgrena , David Rabanusa , Rolf Güstenb , Karl Mentenb , Tim de Zeeuwc , Hans
Olofssond , Andreas Kaufera , Lars-Åke Nymana , Per Bergmand , Carlos de Breuckc , Friedrich
Wyrowskib , Claudio Agurtoa , Francisco Azagraa , Michael Dumkea , Felipe Mac-Auliffea ,
Mauricio Martineza , Francisco Montenegroa , Dirk Mudersb , Vincent Reverete , Christophe
Risacherf , Rodrigo Parraa , Giorgio Siringoa and Gundolf Wiechinga .
a European
Southern Observatory, Alonso de Córdova 3107, Vitacura, Santiago, Chile;
für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany;
c European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany;
d Onsala Space Observatory, Chalmers University of Technology, 43992 Onsala, Sweden;
e CEA, IRFU/SAp, 91191 Gif-Sur-Yvette / Lab. AIM, UMR, CEA, CNRS, UP7, France;
f SRON, Postbus 800, 9700 AV Groningen, The Netherlands
b Max-Planck-Institut
ABSTRACT
APEX, the Atacama Pathfinder EXperiment, is being operated successfully, now for five years, on Llano de Chajnantor at 5107 m altitude in the Chilean High Andes. This location is considered one of the worlds outstanding
sites for submillimeter astronomy, which the results described in this contribution are underlining. The primary
reflector with 12 m diameter is cautiously being maintained at about 15 µm by means of holography. This
allows to access all atmospheric submillimeter windows accessible from the ground, up to 200 µm. Telescope and
instrument performance, operational experiences and a selection of scientific results are given in this publication.
Keywords: telescope, submillimeter
1. INTRODUCTION
APEX, the Atacama Pathfinder EXperiment,1 is a single dish telescope for millimeter and submillimeter astronomy, and operates since its inauguration in September 2005. As its name suggests, the project is set to pioneer
scientific exploration of the millimeter and submillimeter wavelength range, which will be followed up by other
projects currently in planning or construction at the same site, especially by the Atacama Large Millimeter
Array (ALMA).2 The full description of the telescope is given in the SPIE paper by Güsten et al. (2006).3 This
follow-up paper will summarize the basic information of the project, and present changes and results from the
first five years of the operations. This paper is structured as follows: Section 2 describes the telescope and its
performance. The experimental character of the project is on purpose, facilitating a versatile instrumentation
strategy, which is outlined in section 3. Operations involve new approaches to staff management and safety
measures, as discussed in section 4. We give a summary of the statistics in section 5 and list selected science
results in section 6. Finally we give an outlook in section 7 on what the future has for us.
APEX is a collaboration between the Max-Planck-Institut für Radioastronomie (MPIfR), the Onsala Space
Observatory (OSO) and the European Southern Observatory (ESO). The observing time (45% MPIfR, 24%
ESO, 21% OSO) is shared along the Partners investments, with 10% allocated to the host nation. The operation
of the observatory has been entrusted to ESO (Chile).
The team comprises 19 staff members (astronomers, telescope and instrument operators, electronics/mechnical
engineers, and technicians), contracted through ESO, and support visitors from the partner institutes. A group
of ten contractors is in charge of the facility maintenance and canteen operations.
Further author information: (Send correspondence to A.L.). A.L.: E-mail: alundgre@eso.org; D.R.: E-mail: drabanus@eso.org
!"s$%&a()%y !,$%a(-).s/ 0(%a($1-$s2 3%)4$ss$s2 a.d 0ys($ms 7772 $d-($d "y
Da&-d R. 0-;&a2 <;-s). =. 3$4>2 =. ?@)mas 0)-A$%2 3%)4. )A 037B V);. DDED2 DDED0G
H 20J0 037B K LLL 4)d$/ 02DDMDGNO/J0/QJG K d)-/ J0.JJJD/J2.G5NSG5
3%)4. )A 037B V);. DDED DDED0GMJ
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
1.1 Facilities
The facilities of APEX comprise two locations: the antenna site on Llano de Chajnantor, at 5107 m a.s.l., and
the base camp in Séquitor, an Ayllu∗ of San Pedro de Atacama. The Chajnantor site offers an oxygen-enhanced
control room, a server room, laboratory space, storage for spare parts and other maintenance related devices,
a kitchen and sanitary compartment, as well as a power generation house, containing three redundant Diesel
generators of 250 kVA each† . In Séquitor, there are dormitories for staff and visitors, a control room for remote
operations, a server room for data mirroring and offline data reduction, electronics and mechanical workshops
for smaller repairs, developments and fleet maintenance, storage spaces, a canteen, and rooms for meetings and
entertainment. A 36 Mbit/s microwave link connects high site and base camp for remote operation and data
relay.
A fleet of five double–traction vehicles allows commuting between base camp and high site, and a mini van
facilitates commuting to the Calama airport for staff exchange.
1.2 Observatory Site
The APEX site, located at an elevation of 5107 m in the main cordillera of the Andes, is accessed through the
service road that the ALMA project has constructed. It passes through the ALMA basecamp, about 30 km
distant from San Pedro de Atacama, and continues another 35 km to the Chajnantor site. There, both air
pressure and humidity are unique: At a pressure of about 555 hPa and frequent relative humidities of below 15%,
the atmospheric opacity permits to access the millimeter and submillimeter windows up to 200 µm (1.5 THz).
2. TELESCOPE
The APEX telescope was designed and constructed by VERTEX Antennentechnik GmbH, Germany, and is a
modified copy of a prototype for the ALMA project. It accommodates two additional Nasmyth cabins to host a
larger number of instruments, a container for supplementary equipment and a compressor compartment for the
closed–cycle cryocoolers.
The 12 m diameter main dish consists of 264 aluminium panels in 8 rings fixed on a carbon fiber reinforced
plastic (CFRP) backup structure of 24 sandwich backup structure (bus) segments. Each panel is supported by five
vertical (four corners and center) and three lateral adjustment elements. The panels, which have been chemically
etched to scatter visible and IR radiation and thus allow daytime observations, have been manufactured to an
average surface accuracy of 8 µm rms. The bus is supported by an INVAR cone, which is attached to the top of
the Cassegrain cabin. The total mass of the modified antenna is ≈125 tons.
2.1 Surface quality
The surface accuracy is routinely measured using near-field holography, which is described in Güsten et al. (2006).3
If it is found that the surface accuracy has degraded to more than 15 µm rms during the altiplanic summer season, the surface is adjusted and the result is verified in the beginning if the altiplanic winter season. A typical
surface error pattern of the APEX main reflector after a holography run is presented in Figure 1.
2.2 Pointing
Since early 2010, we are using an extended pointing model, which consists of 28 pointing parameters. Most
parameters do not change with time, and are therefore kept fixed. Some parameters show long term variations
(∼month), and are recalculated each pointing run. A few parameters vary on time scales <24h, and are updated
regularly using pointing checks or other sources of information. All pointing modelling is done in TPOINT.4
A master pointing model is created using the optical telescope. Each instrument is associated with a delta
pointing model that accounts for the difference with respect to the master model.
∗
A former settlement for Atacamenian family clans, nowadays forming a sub–village in the outskirts of the San Pedro
de Atacama village
†
At an altitude of 5107 m, this power is de–rated. At sea level, the generators can produce 400 kVA each
3%)4. )A 037B V);. DDED DDED0GM2
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
20100426_cgdc_avelO.Epr rms = 13.66 urn
30
24
18
12
-
0-6
-12
-18
-24
-6
-4
0
-2
X
2
4
6
30
Em]
Figure 1. A surface error pattern of the APEX main reflector after the April 2010 holography run. The map is an average
of 10 individual maps and the dish is optimized for observations at 50 degrees elevation (i.e. pre-loading has been applied
to the surface). Masking the inner 1 m and the projected area of the quadripod legs, but without radial tapering, the
surface accuracy is calculated to 13.7 µm rms (including pre-loading).
Each master model pointing run consists of 100-150 stars, which takes 1 hour to observe. All pointing runs
are added to a data base and to fit the pointing model, we use pointing results from about 20 pointing runs.
The strategy that is followed for a master model calculation is:
• 18 parameters do not show variations with time (HASA, HASA2, HACA2, HACA3, HESE, HECE2,
HECE6, HESA2, HECA2, HESA3, HECA3, HESA4, HESA5, HSCA, HSCA2, HSSA3, HSCA5, and NPAE)
and are therefore fitted overall in TPOINT. The typical error in determining the value is 0.05"" (1σ), but
it can for certain parameters be higher (e.g. NPAE 0.3"" , HESE 0.6"" ).
• 6 parameters (IA, IE, HECE, CA, AN, AW) are fitted per pointing run. Typical errors are 0.6"" for IA, IE,
HECE, CA and 0.2"" for AN and AW.
The typical pointing rms for a master model run (2000-3000 stars) is 1.5"" (dR, 1σ), see Fig. 2 for an example
of the residuals. The best sky rms one could obtain using the old model was 1.9"" ,3 but in most cases the sky
rms was 2.0 − 2.2"" .
For the delta model, we freeze the parameters that were fitted overall in the master model calculation, fit
some of the other parameters overall, and let a few parameters vary from run to run. In order to keep the
free parameters to a minimum, the antenna tilt parameters (AN and AW) are treated separately; They are
derived using a method we call ”tiltmeter run”, where the yoke tiltmeters are read while the antenna is rotated
∼400◦ in azimuth. Each pointing run is kept to be less than 2 hours, and we usually get about 15-20 pointing
measurements in that time. After a new tiltmeter run, the pointing measurements can continue if neccessary.
The data set are coadded to the data base of previous pointing runs for the same instrument.
A typical delta model is calculated following the strategy:
• IA, HECE, HSCA are fitted overall.
3%)4. )A 037B V);. DDED DDED0GME
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
+5
?
;c
dV
dE
dS
+5
+20
1
FS
F
'
5
5
20
dZ
dS
AM
+5
+5
+20
I!uiiIiII
:°
0
5
5
20
U
X
_.II.
16
arcsec
I
z
-
Li-.
R
I
10"
_.II.
s
L
N
D
w
?
-
10
arcsec
bins
APEX OPT 2010
Figure 2. The output of the command .a9 in TPOINT. The upper three panels show the optical pointing residuals (in
arcsec) in azimuth (dS), elevation (dE), and Az/El nonperpendicularity (dV) as a function of the azimuth. In the middle
row, the two left panels show the residuals (in arcsec) in azimuth (dS) and zenith distance (dZ) as a function of the
zenith distance, and the right panel shows the axis misalignment (AM). The lower row shows from left to right: pointing
residuals as a scatter diagram, distributions of the pointing residuals, and finally a map of the pointing residuals as error
projected on sky. In total this image consists of 2371 pointing measurements, and the sky rms is 1.55"" .
3%)4. )A 037B V);. DDED DDED0GM4
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
• For the instruments in the Nasmyth cabins, NRX and NRY are determined overall, and then kept fixed.
These parameters are closely coupled to other parameters, mainly IA and HECE, and in addition NRY is
coupled to CA.
• AN and AW are fixed to the result from the tiltmeter run carried out most closely in time.
• Only two parameters, CA and IE are free to vary from run to run.
The typical rms for a delta model run is about 2"" (1σ). A normal pointing data set consists of 100-200
pointing measurements in 5-10 pointing runs obtained during typically a few months time. The value is slightly
higher for the lower frequency receiver (APEX-1) and lower for the higher frequency receivers (like APEX-3).
Both the master model and the delta model rms have been verified on sky.
2.3 Chopping Secondary
The chopping secondary was installed in late 2007. The practical frequency range for observations is 0.5-1.5 Hz,
and the chopping amplitude can be up to 150”, which means that the OFF-beam can be placed up to 300” away
from the ON-beam.
The secondary, lightweight mirror is made out of carbon fiber with an aluminum coating. It was observed
that the corrosive environment, most probably caused by the abandoned Purico sulfur mine upstream of the
prevailing wind direction, deteriorates the metallization significantly. Therefore, the aluminum layer of the spare
subreflector was coated with silicon monoxide as passivation layer. So far no degradation was observed.
2.4 Observing and Control Software
The APEX telescope is controlled by the APEX Control System (APECS).5 APECS is based on the ALMA
Common Software (ACS)6 and the (ALMA) Test Interferometer Control Software (TICS).7 ACS provides the
CORBA-based8 middleware communication layer to interface the hardware components to the control system.
TICS provides the basic CORBA objects for antenna control in horizontal and equatorial coordinates. In addition
to that, there are utilities to record several kinds of time stamped monitor points into a database (IBM DB2)
and to perform optical pointing runs.
The ACS and TICS packages fulfill the requirements of common network communication, automatic monitoring, real-time tracking and remote observing. The overarching software to use all hardware devices in a
coordinated way necessary for astronomical observations was developed by the APEX software development
group.5 This included defining the instrument and device interfaces9–11 and the raw data format interface
MBFITS.12
After its initial deployment in 2003, APECS has been continuously developed and improved. In 2005 the
system was fully usable for standard observing modes. From 2006 on more complex modes such as drift scans,
spiral and Lissajous patterns have been added. APECS now supports array receivers including active field
derotation. Complex IF processing setups and sideband separating receivers can be handled automatically.
Until the end of the year 2008 APECS was based on a very early version of ACS and TICS which had been
used to test and characterize the ALMA prototype antennas at the VLA site in New Mexico, USA. This ACS
version exhibited problems in some of the basic interprocess communication schemes which led to slowing down
or crashing APECS. It also required using an old version of the Linux operating system (RedHat 7.2). The latter
prevented us from upgrading the server hardware which was beginning to age and which was not able to handle
the higher data rates of current and future instruments.
It thus became increasingly urgent to upgrade to a more recent version of ACS which had been much improved
within the ALMA project. During 2007 and 2008 TICS and APECS were ported to the latest ACS. Technical
problems in the deployment of the CORBA libraries on the VxWorks real-time computer delayed the upgrade.
In early 2009 the new software was deployed at APEX in Chile. Parallel to the software upgrade the network
infrastructure was changed from 100 Mbps to gigabit speeds and physical network separation for the control and
data acquisition parts of APECS.
3%)4. )A 037B V);. DDED DDED0GM5
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
The new APECS version 2.0 was based on Scientific Linux 5.2, VxWorks 6.6 and ACS 8.0. It was able to
handle higher data rates than the old setup and it was much more stable. After this rather larger upgrade we
now try to keep APECS aligned with new ACS versions. In early 2010 version 2.2 of APECS under Scientific
Linux 5.3 and ACS 8.1 was installed in Chile. This year we plan to move to ACS 9.0 and VxWorks 6.8. The
upgrades are always performed during the shutdown and maintenance times in the Chilean summer time.
Keeping ACS current will allow to continuously upgrade hard- and software and to benefit from software
improvements. In addition to these upgrades we also keep adding new features to APECS for new observing
modes and instruments.
3. FACILITY INSTRUMENTATION
The instrumentation that is available to the communities of the APEX partners and Chile are described in the
following. For their performances, individual calibration strategies we refer to M. Dumke et al. in this issue.13
3.1 Swedish Heterodyne Facility Instrument (SHeFI)
The facility receivers which are open to the public for spectral line observations are the ones available in the
Swedish Heterodyne Facility Instrument (SHeFI). This instrument was built by the GARD group at the Chalmers
University, Sweden with the Onsala Space Observatory. The SHeFI consists of four wideband heterodyne receiver
channels for 230 GHz (APEX–1), 345 GHz (APEX–2), 460 GHz (APEX–3) and 1300 GHz (APEX–T2). All
the bands are located inside a single closed-cycle cryostat which cools the coldest part at a temperature close
to 4 K. The first three bands for 230, 345 and 460 GHz employ superconductor-insulator-superconductor (SIS)
mixers. APEX–1 and –2 behave as single sideband receivers (SSB) and APEX–3 as double sideband receiver
(DSB). The 1.3 THz band is built using hot electron bolometer (HEB) mixer technology and is a DSB receiver.
For a complete description, please refer to Vassilev et al. (2008)14 and Lapkin et al. (2008).15
SHeFI is located in the Nasmyth-A cabin of the telescope. The optical path is fairly complex and includes
not less than 11 mirrors with the main dish and the subreflector. The SHeFI receiver support structure is firmly
attached to the Nasmyth elevation tube flange and therefore temperature deformations of the A cabin have little
to no effect on the receiver alignment.
The cryostat was built at the Chalmers University and uses a closed-cycled Sumitomo cooling machine providing temperatures at the coldest stage below 4 K, which is needed for the best operations of the superconducting
mixers.
One receiver band is in operation at a time. A rotating mirror selects the desired band and the receiver
output is routed to the backends (spectrometers and total power). These heterodyne instruments all use direct
multiplication chains for the local oscillator (LO) injection. The reference signal comes from a signal generator
providing a strong clean signal from 10 to 20 GHz.
In detail, the available bands are:
• APEX–1: 210–280 GHz, with an average noise temperature of Trec = 140 K (SSB).
• APEX–2: 285-380 GHz, with an average noise temperature of Trec = 160 K (SSB).
• APEX–3: 385-500 GHz, with an average noise temperature of Trec = 100 K (DSB).
• APEX–T2: 1220-1385 GHz, with an average noise temperature of Trec = 1250 K (DSB).
The output of these receivers is spectrally resolved in a pair of Fast Fourier Transform Spectrometers16
(FFTS), performing an optimized FFT on the data in a 1 GHz bandwidth.
3%)4. )A 037B V);. DDED DDED0GMN
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
3.2 LABOCA
The Large APEX BOlometer CAmera (LABOCA) is a 295-pixel, 345 GHz (870 µm) bolometer array. The field
of view (FoV) is 11x11 arcmin2 . It was successfully commissioned in May 2007. It was built by the bolometer
development group of the MPIfR. A complete description is given in Siringo et al. (2009).17
The instrument was first installed at APEX in September 2006. It was then continuously improved until the
commissioning and science verification in May/June 2007. Since then, it is available for regular observations.
The sensitivity of LABOCA is 60 mJy s1/2 . The LABOCA observations is usually organized in dedicated timeblocks because the operation depends on the availability of liquid helium on site. These time-blocks will usually
be announced in the corresponding calls for proposals.
The software system offers various observing modes for LABOCA: Spiral mode. In this mode the telescope
moves - relative to the tracking position - from a start radius to an end radius along a spiral pattern. This allows
to obtain a fully sampled map of the total FoV of LABOCA. The spiralling is done with a constant angular
speed. This mode can be combined with a raster mode that allows for larger map sizes than the instrument’s
FoV. On-the-fly mode. This is a mode used for mapping sources larger than 30" . The telescope scans along
map rows or columns in either horizontal or equatorial coordinates with a continuos speed. Photometry, or
On-Off mode. During source tracking, a combination of nodding and chopping allows to obtain highly accurate
photometry measurements of compact sources (i.e. smaller than a pixel’s beam).
3.3 SABOCA
The Submillimetre APEX BOlometer CAmera (SABOCA) is a bolometric continuum receiver operating at 810
GHz (in the 350 µm atmospheric window). The FoV is 1.5x1.5 arcmin2 and its detector array consists of 39
superconducting transition edge sensor (TES) bolometers with SQUID (Superconducting Quantum Interference
Device) amplification and time-domain multiplexing. The receiver has been designed and integrated by the
bolometer group at the MPIfR in collaboration with the Institute of Photonic Technology (IPHT, Jena, Germany). The MPIfR group has a long track record in the development of bolometers and bolometric cameras
for astronomical applications. IPHT is known for building state-of-the-art superconducting devices for over 15
years. The collaboration to build SABOCA merges the technology expertise provided by the two groups.
The active development process took several years as it involved a large number of theoretical studies, cycles
of manufacture and tests in the laboratory. A prototype system was successfully tested on APEX during May
2008. Some technical problems were identified and fixed. Thus, commissioning began in September with an
improved version of the receiver. The final version of SABOCA was installed in 2009 and commissioning was
completed in 2010, see Siringo et al. (2010)18 for details. The observing modes available for LABOCA can also
be used by SABOCA.
3.4 Principal Investigator instrumentation
A number of PI instruments are or have been installed at the different instrument slots. CONDOR: a 1.5
THz heterodyne receiver from the University of Köln;19 PI1100: a 1.1THz single pixel heterodyne SIS receiver
(4 GHz IF bandwidth) from MPIfR; CHAMP+ : a dual-color 2×7 element heterodyne array from MPIfR for
operation in the 810 GHz and 690 GHz atmospheric windows;20, 21 FLASH: a dual-color (initially 810/460, now
460/345 GHz) single pixel heterodyne receiver from MPIfR;22 p-ArTeMiS: a prototype bolometer (450 µm)
from CEA Saclay;23, 24 APEX-SZ: a 150 GHz bolometer designed from University of California at Berkeley
for studying the Sunyaev-Zel’Dovich Effect;25 PolKa: A polarization add-on for LABOCA from MPIfR , see
Siringo et al. (2010),26 this SPIE. Detailed descriptions of these instruments are beyond the scope of this paper.
4. OPERATIONS
The APEX science operations are carried out by the science operations group and can be considered either facility
observations or principal investigator (PI) observations.
The idea during facility observations is to work through a queue of observing projects that have been approved
by time allocation committees (TAC) of the partner’s communities. There we operate in service mode, observing
3%)4. )A 037B V);. DDED DDED0GMD
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
the projects that are scheduled around the constraints of instrument requirements, instrument availability, and
weather conditions (mostly determined by the content of precipitable water vapor along the line-of-sight).
PI observations usually include more staff of the investigator group, operating instrumentation which requires
staff presence at the antenna, and their observing projects are as well approved and granted by a TAC.
During daytime the antenna and the instruments are operated from the Chajnantor control room by a team
of at least two persons, of which at least one is an APEX staff member and site certified (i.e. trained to handle
the antenna, knowledgeable to take safety measures). They also, shared with members of the engineering branch,
take care of instrument maintenance (e.g. refilling cryogenics, recycling of instrument fridges etc.). During night
time, the antenna and instruments are normally operated from the Sequitor control room.
4.1 16 h to 24 h – operations experiences
Since the inauguration in 2005, the baseline facility observations were carried out in a 16 h observation mode,
covering the night shift from sunset to sunrise and a morning shift until local noon. The afternoon times were
reserved for engineering access.
In June 2009, the APEX team adopted 24 h Operations Plan for its implementation on tentative basis. The
reason for this decision was to optimize the efficiency of the observatory by switching from 16 h operations per
day to 24 h, which resembles an increase of 50% with respect to the conventional shift system. This plan foresees
to invite at least three external support visitors (SV) sent by the partners that complement operations teams in
order to cover three daily shifts. Finally, there are always two persons in either a night shift, morning shift or
afternoon shift. Since the day shifts are to be executed at the Chajnantor site for antenna safety reasons, one
of the SVs per team requires to be site certified. The 24 h Operations plan was formally accepted by the APEX
board in February 2010.
4.2 Efficiency improvement
The efficiency measurement requires some detailed explanations: At APEX, we currently have two facility
instruments that require a manual refilling with cryogenic refrigerants (liquid helium and nitrogen) and recycling
of the focal plane fridge to reach mK temperatures. For LABOCA, this takes 2.3 hours (once or twice per
day, depending on demand), and for SABOCA about one hour every 48 h. These recyclings cannot be done in
parallel due to geometric constraints in the telescope. These can sum up to 4.5 h of ”dead time” daily, if both
instruments are requested, but the loss is mitigated due to the fact that observations with another instrument
may continued in a limited elevation range while recycling LABOCA.
In the 16 h operations scheme, 2.5 hours of these recycling times can be scheduled to one of the day shifts
(typically in the afternoon shift with least favorable weather conditions), so that they were not counted as hours
lost to science operations.
In the 24 h operations scheme, these 4.5 hours on average, must be deducted from the total time available, so
that, on average, 19.5 hours are available for observations per day. This renders an effective efficiency increase
of 22% (effectively 3.5 hours per day), when shifting from the 16 h operations to the 24 h operations scheme.
In the summer months, the total time available is even further reduced due to the ”sun avoidance” time in
which the telescope shall not be operated due to the angular proximity of the sun to the zenith position (sun
avoidance time). Of course, instrument recycling is usually scheduled during sun avoidance time, hence these
dead times are used for instrument maintenances.
5. STATISTICS
5.1 Observing statistics
As the problems got fewer and operations got more streamlined, the on-sky integration time has increased at
APEX. The total on-sky time rose steadily from 2291 h in 2007 via 3093 h in 2008 to 3443 h in 2009. With on-sky
time we mean the accumulated times of each scan time, where each scan time is counted from the timestamp
the scan was submitted to the observing engine to the timestamp where it was successfully finished.
3%)4. )A 037B V);. DDED DDED0GMG
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
Total on-sky time (h)
450
400
350
300
250
200
150
100
50
0
1
*
-0
10
20
-0
1
-0
09
20
*
1
-0
09
20
08
20
-0
*
-0
08
20
1
-0
0*
20
*
-0
0*
20
06
20
<ear-month
Figure 3. The on-sky time as function of year (and month) between June 2006 end of May 2010. The on-sky time is
defined as the time between an observing command has been sent until the scan has finished. No science observations
are scheduled during the summer months (Jan-Mar). During these months, and when the weather and staffing situation
allows, the antenna is used for technical observations, such as pointing, calibration, technical verification, etc.
Figure 3 shows the on-sky time as function of month and year since mid-2006, and it seems as if we have
reached a plateau of ∼350 h on-sky time per month. As from June 2009 we are most of the time in 24 h operations
mode (see Sect.4.1), and this has a discernable, positive effect on the usage of the antenna.
5.1.1 Statistics per instrument
Taking 2009 as an example, and summing the on-sky time per instrument, see Table 1, we see that the facility
receivers APEX-1, APEX-2 and LABOCA account for about 2/3 of the total on-sky time. SABOCA and the
THz instruments (APEX T2 and PI1100) were still under commissioning and received relatively little on-sky
time. The PI instruments APEX-SZ, p-ArTeMiS, FLASH, and CHAMP+ were only operated during limited
time periods.
The total time on sky is 3136h, which amounts to about 348 h per month (calculated over 9 months), or about
12 h per day. The remaining time is on average spent as 5 h on non-science operational time (tuning, refilling
and recycling, optical pointing, holography, etc.), and 7 h downtime (mostly weather, hardware and software
problems).
Table 1. The number of hours spent on-sky per instrument during 2009. Instruments in roman font are facility instruments,
and instruments in italic font are PI instruments. The instruments are ordered in frequency and divided in heterodyne
and bolometer. Non-scientific frontends, like the holography receiver, are not listed here. See section 3 for description of
the instruments.
Heterodyne
APEX T2
3
PI1100
11
CHAMP+ 287
FLASH
102
APEX-2
685
APEX-1
640
Total
1728
Bolometer
SABOCA
79
p-ArTeMiS 116
LABOCA
811
APEX-SZ
402
1408
5.2 Weather statistics
Apart from science data, APEX has also been collecting climatological data for the past 5 years. A key parameter
for submillimeter astronomy observations is of course the precipitable water vapor (PWV). Figure 4 shows the
3%)4. )A 037B V);. DDED DDED0GMS
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
Figure 4. The precipitable water vapor measured by the APEX radiometer as function of the time of the year. Data have
been collected over the years 2006-2010. The blue, green, and red bars show the 1st, 2nd, and 3rd quartile of the data.
PWV measured by the APEX radiometer as function of the time of the year. This image shows that during more
that 50% of the time during the period when we are conducting science operations (April 1st to January 1st),
the PWV is below 1mm (green quartile), and during the 5 best months (May-September), the PWV is below
0.5 for more than 25% of the time (blue quartile). We have seen PWV<0.5mm during all weeks of the year,
and PWVs as low as 0.07mm (confirmed by heterodyne calibration observations) has been seen a few times, or
zenith opacity τ < 0.1 at 870 µm (LABOCA).
Furthermore, Figure 5 shows temperature, humidity, windspeed, and barometric pressure as function of time
of the year. The southern winter (June-August) is characterized by low temperatures, low humidity, higher wind
speeds, and slightly lower pressure. The figures is based on median values per hour, which tends to smooth the
appearance, and sorted per week. The reason for taking the median value and not the average value is that
the median value is not affected by non-representative data (e.g. wind gusts or the occasional bad data). The
smoothing of the data affects in particular the plot of the wind; There are in the raw data base measurements
of wind gusts up to 46 m/s, and sustained wind for more than one hour up to 32 m/s.
The diurnal variation of temperature and wind in winter (June) and summer (December) is shown in Figure
6. Typical night-time temperatures are -6 to -10 ◦ C during winter and 0 to -4 ◦ C during summer, and the
temperatures rising slowly after sunrise, reaching normally around -4 ◦ C and 4 ◦ C in winter and summer,
respectively. Extreme night and day temperatures are -20 ◦ C and 12 ◦ C, respectively.
The wind pattern is during winter characterized by steady wind 5-7 m/s during night time, increasing steadily
after sunrise, reaching 6-11 m/s at 12-14 (CLT), and then starting to drop. The wind pattern during summer
look slightly different, with a minimum at 22-02 (CLT) and 2-4 m/s, gradually increasing up to 19 (CLT) and
8-10 m/s, after which the wind speed drops relatively abrupt. This winter pattern normally starts in October
and continues until March.
6. SELECTED SCIENCE RESULTS
Over its first five years of operations, APEX has significantly contributed to a wide variety of science areas. A
total of 26 articles appeared in a special (August 2006) issue of Astronomy and Astrophysics with the telescope
receiving title page prominence. Since then, about 50 further refereed papers have appeared in journals (out of
which 29 during 2009) and many more are in the process of being published. To name but a few highlights:
• Red-shifted 0.89 absorption of the ground-state lines of ortho-water and -ammonia has been detected
toward a gravitational lens27
3%)4. )A 037B V);. DDED DDED0GMJ0
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
Chajnantor Temperature
ChaKnantor Bumidity
10
100
90
*0
)0
0
Bumidity (G)
Temperature (C)
5
-5
-10
60
50
&0
30
20
-15
10
-20
'an
Feb
Mar
Apr
May
'un
'ul
Aug
Sep
Oct
Nov
0
,an
Dec
Feb
Mar
Apr
May
,un
Month
,ul
Aug
Sep
Oct
Nov
Dec
Sep
Oct
Nov
Dec
Month
ED'F('(79- <=(> 4/**>
Chajnantor Pressure
560
20
558
Pressure (hPa)
<=(> ?/**> @AB?C
556
15
10
554
552
550
548
5
546
0
&'(
)*+
,'-
./-
,'0
&1(
&12
.13
4*/
567
89:
;*6
544
Jan
Feb
Mar
,9(7D
Apr
May
Jun
Jul
Aug
Month
Figure 5. Weather statistics (temperature, humidity, wind speed, and pressure) measured at the APEX weather station.
Data have been collected over the years 2006-2010. The blue, green, and red bars show the 1st, 2nd, and 3rd quartile of
the data. The raw data have a dump rate of 1 minute. From the raw data, the median value per hour was calculated and
these median values were sorted per weeks. The thin cyan line shows the tenth largest value in the data base, and the
thin purple line shows the tenth lowest value among the median values, per week.
• The first imaging results with the 14 element CHAMP+ heterodyne array have published28–30
• as have the first articles reporting results from the APEX Sunyaev Zeldovich survey project25, 31
• LABOCA mapping of nearby active galaxies, including the detection of, both, dust emission from the plane
of the Sombrero galaxy32 and synchrotron emission from the jets and lobes of the archetypical radio galaxy
Centaurus A33
• Galactic and extragalactic molecular line studies in the 650 and 350 µm windows, very few of which were
made in the past, e.g. the 1mm line survey in the Orion bar34
• The first detection of several molecular ions, including CF+, which is a ubiquitous component of the
interstellar medium,35 and the first interstellar detection of OH+36
• ATLASGAL, a submillimeter dust continuum survey of the inner Galactic plane with LABOCA has been
completed, and first results have been published37
• The first LABOCA 870 µm cosmological deep field imaging of a Ly α proto-cluster at z = 2.3838
• Discovery of strong [CII] emission in a lensed galaxy at high redshift (z=4.43)39
• The first results from the LABOCA survey of the Extended Chandra Deep Field South (ECDFS)40–42
• The finding of a sub-millimetre galaxy at redshift z=2.3259, which has been gravitationally magnified by
a factor of 32 by a massive foreground galaxy cluster lens43
3%)4. )A 037B V);. DDED DDED0GMJJ
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
37/89/906. *+,-+./01.+ 2:+;+,<+.4
$#
%
%
0
#
*+,-+./01.+ 234
Temperat1re 2C4
Cha8nantor Tempert1re 2:1ne4
10
-%
-10
-1%
!$#
!$%
0"
0'
0(
08
10
1"
1'
1(
18
"0
!"#
##
""
#'
#(
#)
$#
$"
$'
$(
561.
89:;+:+<57 )*+, =.//, 0>6+/3
89:;+:+<57 )*+, =.//, 0>/?/1@/73
%"
%"
$#
$#
$"
#
"
""
#"
5o1r
)*+, -.//, 012-3
)*+, -.//, 012-3
-"0
00
!%
$)
"#
""
$(
%"
%%
$"
#
"%
"&
"'
"(
$"
$%
$&
$'
$(
%"
%%
"
""
"%
"&
4567
"'
"(
$"
$%
$&
$'
4567
Figure 6. Diurnal variation of temperature (upper row) and wind speed (lower row) in June (left column) and December
(right column) as function of time (CLT). Data have been collected over the years 2006-2010. The blue, green, and red
bars show the 1st, 2nd, and 3rd quartile of the data. The raw data have a dump rate of 1 minute. From the raw data,
the median value per hour was calculated and these median values were sorted per weeks. The thin cyan line shows the
tenth largest value in the data base, and the thin purple line shows the tenth lowest value among the median values, per
week.
7. OUTLOOK
7.1 Project lifetime
The future is bright for APEX. The excellence of the site calls for more observations, both focussed as well as
surveys, and the oversubscription rate in the various partner’s time slots underlines this. Also, the experimental
character of APEX makes it suitable as a platform for testing new and/or different detection technologies.
Therefore, an extension of the lifetime of APEX beyond 2012 is under discussion among the project partners.
7.2 Future instrumentation
A second generation instrumentation suite is already being contemplated. State–of–the–art technologies for both
heterodyne and bolometer instruments are available, and it is a matter of funding and extension of the project’s
lifetime to get them to APEX.
On the spectroscopy side, heterodyne receiver arrays with ever wider bandwidth will allow to investigate
much more in–depth extragalactic features with more scrutiny.
The rather large optical field of view of the telescope calls for larger format bolometer cameras. Currently,
transition edge sensors (TES) or kinetic inductance detectors (KID) are under evaluation for larger formats.
The outcome of first tests as visiting instruments at APEX will allow to set course for developing more mature
incarnations of these technologies.
Also, a large dual-color heterodyne array is under development at MPIfR: 19 (7) pixels in the 460 (345) GHz
window respectively. Installation planned during 2011.
See Siringo et al. (2010),44 this SPIE conference, for more details about future receivers for APEX.
3%)4. )A 037B V);. DDED DDED0GMJ2
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
ACKNOWLEDGMENTS
Thanks to Ignacio Vera (ESO/DBCM), who collected and analyzed the APEX pwv data and produced Figure
4.
REFERENCES
[1] Güsten, R., Nyman, L. Å., Schilke, P., et al., “The Atacama Pathfinder EXperiment (APEX) - a new
submillimeter facility for southern skies -,” A&A 454, L13–L16 (2006).
[2] Hills, R. E. and Beasley, A. J., “The Atacama Large Millimeter/submillimeter Array,” in [SPIE Conference
Series], SPIE 7012 (2008).
[3] Güsten, R., Booth, R. S., Cesarsky, C., Menten, K. M., et al., “APEX: the Atacama Pathfinder EXperiment,” in [SPIE Conference Series], SPIE 6267 (2006).
[4] Wallace, P., [Tpoint software manual], Abingdon, Oxfordshire, UK (2003).
[5] Muders, D., Hafok, H., Wyrowski, F., et al., “APECS - the Atacama pathfinder experiment control system,”
A&A 454, L25–L28 (2006).
[6] Raffi, G., Chiozzi, G., and Glendenning, B., “The ALMA Common Software as a Basis for a Distributed
Software Development,” in [Astronomical Data Analysis Software and Systems XI], D. A. Bohlender, D. Durand, & T. H. Handley, ed., Astronomical Society of the Pacific Conference Series 281, 103–+ (2002).
[7] Glendenning, B., Brooks, M., Chiozzi, G., et al., “Test Interferometer Control Software Design Concept,”
tech. rep., ALMA Report (2001).
[8] The Common Object Request Broker: Architecture and Specifications. Framingham, MA, USA (1999).
[9] Muders, D., Hatchell, J., Lemke, R., Olberg, M., and Hafok, H., “Software Interfaces for Submillimeter
Telescope Instrumentation,” tech. rep., APEX (2002).
[10] Muders, D. and Hafok, H., “APEX Instruments Generic CORBA IDL Interfaces,” tech. rep., APEX (2007).
[11] Hafok, H., Muders, D., and Olberg, M., “APEX SCPI Socket Command Syntax and Backend Data Stream
Format,” tech. rep., APEX (2006).
[12] Muders, D., Polehampton, E., and Hatchell, J., “Multi-beam FITS Raw Data Format,” tech. rep., APEX
(2002).
[13] Dumke, M. and Mac-Auliffe, F., “The APEX calibration plan: goals, implementation, and achievements,”
SPIE this issue (2010).
[14] Vassilev, V., Meledin, D., Lapkin, I., et al., “A Swedish heterodyne facility instrument for the APEX
telescope,” A&A 490, 1157–1163 (2008).
[15] Lapkin, I., Nyström, O., Desmaris, V., et al., “Optics Design and Verification for the APEX Swedish
Heterodyne Facility Instrument (SHeFI),” in [Proceedings of the Ninteenth International Symposium on
Space Terahertz Technology, held April 28-30, 2008, in Groningen], W. Wild, ed., 351–+ (2008).
[16] Klein, B., Philipp, S. D., Güsten, R., et al., “A new generation of spectrometers for radio astronomy: fast
fourier transform spectrometer,” Millimeter and Submillimeter Detectors and Instrumentation for Astronomy III. Edited by Zmuidzinas 6275, 33 (2006).
[17] Siringo, G., Kreysa, E., Kovács, A., et al., “The Large APEX BOlometer CAmera LABOCA,” A&A 497,
945–962 (2009).
[18] Siringo, G., Kreysa, E., De Breuck, C., et al., “A New Facility Receiver on APEX: The Submillimetre
APEX Bolometer Camera, SABOCA,” The Messenger 139, 20–23 (2010).
[19] Wiedner, M. C., Wieching, G., Bielau, F., et al., “First observations with CONDOR, a 1.5 THz heterodyne
receiver,” A&A 454, L33–L36 (2006).
[20] Güsten, R., Baryshev, A., Bell, et al., “Submillimeter heterodyne arrays for APEX,” in [SPIE Conference
Series], SPIE 7020 (2008).
[21] Kasemann, C., Güsten, R., Heyminck, S., et al., “CHAMP+ : a powerful array receiver for APEX,” in [SPIE
Conference Series], SPIE 6275 (2006).
[22] Heyminck, S., Kasemann, C., Güsten, R., et al., “The first-light APEX submillimeter heterodyne instrument
FLASH,” A&A 454, L21–L24 (2006).
3%)4. )A 037B V);. DDED DDED0GMJE
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms
[23] Talvard, M., André, P., Rodriguez, L., et al., “Recent results obtained on the APEX 12 m antenna with
the ArTeMiS prototype camera,” in [SPIE Conference Series], SPIE 7020 (2008).
[24] Talvard, M., André, P., , Le-Pennec, Y., et al., “Status of the ArTeMiS camera to be installed on APEX,”
in [SPIE Conference Series], SPIE (2010).
[25] Halverson, N. W., Lanting, T., Ade, P. A. R., et al., “Sunyaev-Zel’Dovich Effect Observations of the Bullet
Cluster (1E 0657-56) with APEX-SZ,” ApJ 701, 42–51 (2009).
[26] Siringo, G. et al., “Beginning of operation on APEX of the polarimeter for the Large APEX Bolometer
Camera (LABOCA),” in [SPIE Conference Series], SPIE this issue (2010).
[27] Menten, K. M., Güsten, R., Leurini, S., et al., “Submillimeter water and ammonia absorption by the
peculiar z=0.89 interstellar medium in the gravitational lens of the PKS 1830-211 system,” A&A 492,
725–730 (2008).
[28] Pérez-Beaupuits, J. P., Spaans, M., Hogerheijde, M. R., et al., “CHAMP+ observations of warm gas in M
17 SW,” A&A 510, A87+ (2010).
[29] van Kempen, T. A., van Dishoeck, E. F., Güsten, R., et al., “APEX-CHAMP+ high-J CO observations of
low-mass young stellar objects. I. The HH 46 envelope and outflow,” A&A 501, 633–646 (2009).
[30] van Kempen, T. A., van Dishoeck, E. F., Güsten, R., et al., “APEX-CHAMP+ high-J CO observations of
low-mass young stellar objects. II. Distribution and origin of warm molecular gas,” A&A 507, 1425–1442
(2009).
[31] Nord, M., Basu, K., Pacaud, F., et al., “Multi-frequency imaging of the galaxy cluster Abell 2163 using the
Sunyaev-Zel’dovich effect,” A&A 506, 623–636 (2009).
[32] Vlahakis, C., Baes, M., Bendo, G., and Lundgren, A., “LABOCA and MAMBO-2 imaging of the dust ring
of the Sombrero galaxy (NGC 4594),” A&A 485, L25–L28 (2008).
[33] Weiß, A., Kovács, A., Güsten, R., et al., “LABOCA observations of nearby, active galaxies,” A&A 490,
77–86 (2008).
[34] Leurini, S., Rolffs, R., Thorwirth, S., et al., “APEX 1 mm line survey of the Orion Bar,” A&A 454, L47–L50
(2006).
[35] Neufeld, D. A., Schilke, P., Menten, K. M., et al., “Discovery of interstellar CF+,” A&A 454, L37–L40
(2006).
[36] Wyrowski, F., Menten, K. M., Guesten, R., and Belloche, A., “First interstellar detection of OH+,”
A&A (accepted) (2010).
[37] Schuller, F., Menten, K. M., Contreras, Y., et al., “ATLASGAL - The APEX telescope large area survey of
the galaxy at 870 µm,” A&A 504, 415–427 (2009).
[38] Beelen, A., Omont, A., Bavouzet, N., et al., “Submillimeter observations of the J2142-4423 Lyα protocluster
at z = 2.38,” A&A 485, 645–655 (2008).
[39] Maiolino, R., Caselli, P., Nagao, T., et al., “Strong [CII] emission at high redshift,” A&A 500, L1–L4 (2009).
[40] Weiß, A., Kovács, A., Coppin, K., et al., “The Large Apex Bolometer Camera Survey of the Extended
Chandra Deep Field South,” ApJ 707, 1201–1216 (2009).
[41] Coppin, K. E. K., Smail, I., Alexander, D. M., et al., “A submillimetre galaxy at z = 4.76 in the LABOCA
survey of the Extended Chandra Deep Field-South,” MNRAS 395, 1905–1914 (2009).
[42] Lutz, D., Mainieri, V., Rafferty, D., et al., “The LABOCA Survey of the Extended Chandra Deep Field
South: Two Modes of Star Formation in Active Galactic Nucleus Hosts?,” ApJ 712, 1287–1301 (2010).
[43] Swinbank, A. M., Smail, I., Longmore, S., et al., “Intense star formation within resolved compact regions
in a galaxy at z = 2.3,” Nature 464, 733–736 (2010).
[44] Siringo, G. et al., “A new generation of receivers on the Atacama Pathfinder EXperiment (APEX),” in
[SPIE Conference Series], SPIE this issue (2010).
3%)4. )A 037B V);. DDED DDED0GMJ4
Downloaded from SPIE Digital Library on 16 Aug 2010 to 134.171.162.132. Terms of Use: http://spiedl.org/terms