Spitzer: the first 30 months


The Prime Directive for the design of the Spitzer mission and the Spitzer observatory is to minimize the unnecessary heat reaching the helium tank. This dictates both the choice of Spitzer's unusual orbit and the unusual architecture of Spitzer's cryogenic/thermal system. The telescope and its three instruments are cooled to their ultimate operating temperatures by liquid helium cryogen, which itself achieves a temperature of about 1.2 K under the pumping action of the vacuum of space. The Spitzer cryogen tank (figure 2) held nearly 350 l of helium before launch; when the helium is gone, the primary mission ends. The helium is evaporated and boiled away by minute heat inputs – measured in handfuls of milliwatts – from the Spitzer instruments and structures around the telescope. The tip of your finger radiates ∼25 mW at body temperature; this much heat applied in the wrong way to Spitzer would drastically shorten the mission.

Figure 2 : .

The outer shell is the boundary of the Cryogenic Telescope Assembly (CTA), incorporating the telescope, cryostat, helium tank and instruments. Principal Investigator-led teams provided the instruments; Ball Aerospace provided the rest of the CTA; Lockheed-Martin provided the solar panel and the spacecraft bus. The observatory is about 4.5 m tall and 2.1 m in diameter; the mass at launch was 861 kg. The dust cover atop the CTA was jettisoned five days after launch. (Ball Aerospace)

A kinder, gentler orbit

All previous space observatories have used a low Earth orbit (IRAS and the Hubble Space Telescope) or a high Earth orbit with a period of between one and two days (ISO and the Chandra X-ray Observatory). Spitzer breaks this pattern by using an Earth-trailing heliocentric (solar) orbit. As seen from Earth, Spitzer recedes at about 0.1 AU per year and will reach a distance of 0.62 AU in five years. This orbit satisfies the Prime Directive in several ways. First, there is great benefit in getting away from the heat of the Earth. Secondly, the orbit allows radiative cooling into the refrigerator of deep space to play a larger part in cooling Spitzer and keeping it cool. Thirdly, because there are no eclipses and the Earth and Moon are far away, the orbit permits excellent sky viewing and observing efficiency – thus efficient utilization of the cryogen – even though, again for thermal reasons, Spitzer's pointing direction is constrained to an annulus that extends no closer than 80° towards the Sun and no further than 120° away from the Sun. As well as satisfying the Prime Directive, the solar orbit itself has enabled Spitzer to carry out several unique science investigations. These include conducting in situ studies of the Earth-trailing density enhancement in the zodiacal cloud, which Spitzer passes directly through, and obtaining constraints on the nature of objects responsible for background star microlensing in the Large Magellanic Clouds (LMC) and the Small Magellanic Clouds (SMC), using a parallactic view in conjunction with Earth-based measurements.

The orbit has one major disadvantage. As Spitzer moves away from the Earth, its radio signals become gradually weaker. Data is stored on board and radioed back to Earth through the Deep Space Network (DSN). The current downlink strategy of one or two ∼45 minute passes per day and a downlink data rate of 2.2 Mb s−1 works for the projected cryogenic lifetime of at least five years, although the largest DSN dishes (70 m) will be needed for the final three years.

The technical advantages of this solar orbit are shared by the L2 orbit in use for the Wilkinson Microwave Anisotropy Probe (WMAP) and planned for the Herschel Space Observatory, the Planck spacecraft, and the James Webb Space Telescope (JWST). The solar orbit has an additional advantage in that no station-keeping or propulsion is required following launch, while the L2 missions have a more favourable communications geometry.

The following sections review the key features of the Spitzer design. Further details can be found in a series of papers in the Proceedings of the SPIE (Roellig et al. 2004) and in a forthcoming paper to be published in the Reviews of Scientific Instruments (Gehrz et al. 2007).


Spitzer's novel cryogenic architecture and the overall configuration of the flight system is shown in figure 2, which rewards close study. The spacecraft and solar panel were provided by Lockheed Martin. The telescope, instruments, cryostat and associated shields and shells make up the Cryogenic Telescope Assembly (CTA), built by Ball Aerospace. Figure 3 shows the solar panel and Sun shield, which absorbs or reflects solar energy, contrasting with the black back side of the outer shell, which radiates heat to space.

Figure 3 : .

The Spitzer Observatory in five different orientations. (NASA/JPL-Caltech)

The novel features of the Spitzer cryo-thermal design and its benefits are illustrated in figure 4, which compares the architecture of Spitzer as it now flies with an earlier design that was closer to the predecessor ISO and IRAS missions. ISO and IRAS packaged the telescope inside the cryostat so that it was in contact with the liquid helium and thus cold at launch. Spitzer employs a novel design in which the greater part of the CTA was launched at room temperature; only the science instrument cold assemblies and the superfluid helium vessel were cold within the vacuum cryostat shell. As shown in figure 4, this allowed a much smaller vacuum pressure vessel and a smaller observatory mass than achieved by the cold launch architecture used in older missions.

Figure 4 : .

The warm-launch architecture compared with an earlier cold-launch concept similar to the IRAS and ISO missions. The warm launch achieves the same mission lifetime and primary mirror size, but at a fraction of the cost. Some of the reduction in cost and mass was achieved by streamlining Spitzer's measurement capabilities, but the warm-launch concept retains the core functionality required for Spitzer's most important programmes.

Of course, Spitzer operates properly only with the telescope at ∼5 K, and we use the deep space refrigerator to get us there after launch. The principles of this on-orbit cooling are deceptively simple. In the kinder, gentler solar orbit, we can keep the spacecraft oriented so that the Sun always falls on the solar panel, which shades the telescope outer shell. A carefully designed and fabricated system of reflective shells and shields ensures that little of the solar heat reaches the outer shell to diffuse inward to heat the telescope. A similar system deflects heat from the spacecraft bus, which operates near room temperature. With no heat input, the telescope should cool very rapidly by radiating its heat to space – the anti-solar side of the outer shell is painted black to facilitate this. Prior to launch we expected that the telescope would achieve a temperature of about ∼50 K through this radiative cooling, and that the evaporating helium gas, which is at a temperature of only ∼1.2 K, would cool the telescope to its operating temperature and keep it there.

When things appear this simple in the space business, “the devil is in the details” becomes a mantra. In this case the devil was on our side, and the system worked just as predicted (if not better). The initial (radiative plus gas-assisted) cooldown to ∼5 K took about 40 days. The outermost CTA shell has equilibrated at an operating temperature of between 34 and 34.5 K solely by radiative cooling. This establishes a very cold boundary around the telescope and the cryostat. Very little heat leaks inwards so a small amount of cold helium vapour can keep the telescope at operating temperature. In fact, it takes only about one ounce of helium per day to keep Spitzer cold. Not coincidentally, this is about the same rate at which the instruments boil away the liquid helium by dissipating power into the helium bath. So the system is both in equilibrium and in balance thermally. This is a tribute to excellent design, fabrication and test by the Ball Aerospace team responsible for the CTA. Additional information on the measured on-orbit performance of the Spitzer CTA can be found in Finley, Hopkins and Schweickart (2004).

Let us now fast-forward to that unhappy day, expected to come in the spring of 2009, when the last drop of helium boils away. The Spitzer telescope will start to warm up, but the passively cooled outer shell should stay close to 35 K. The telescope can be cooler because it can radiate (here we go again!) through the aperture at the top of the outer shell; the dust cover shown in figure 2 was successfully ejected a few days after launch – otherwise I would not be writing nor you reading this article. We predict that the telescope will reach an equilibrium temperature below ∼30 K in this scenario. The two shortest wavelength channels (bands 1 and 2) of the IRAC instrument (see below) should operate with full efficiency and sensitivity at this temperature, although Spitzer's other channels will not work. However, we plan to continue the mission for another two-to-five years in this “lukewarm” condition. Several of the most important investigations that Spitzer has carried out (see below) rely heavily or even exclusively on IRAC bands 1 and 2, and these and other science questions could be addressed in this extended mission.

The best of the rest

Additional innovations as well as careful design, fabrication and test characterize the rest of the Spitzer spacecraft. Key components, described in detail in Gehrz et al. (2007), include:

  • An all-beryllium telescope, with an 85 cm primary mirror. Beryllium was selected because of its high strength-to-weight ratio and reproducible cryogenic behaviour. The secondary mirror of the telescope can be moved in an axial direction to focus the telescope on-orbit. Pre-launch test and analysis determined the proper setting of the secondary. Only a small adjustment (less than 0.01 mm) in the position of the secondary was made after on-orbit cooldown to bring the telescope into proper focus. The telescope achieves diffraction-limited performance at all wavelengths longward of 5.5 μm across its entire focal plane. Spitzer achieves an image size of ∼2 arcsec at its shortest operating wavelengths.
  • An excellent pointing and control system. This system is built around a high-performance autonomous star-tracker, which points the telescope to the desired starfield by using a pattern-recognition algorithm based on an internal star catalogue. In addition, the system includes gyroscopes, reaction wheels to turn the spacecraft, and a visible-light sensor in the telescope focal plane to establish and track the line of sight of the cold telescope relative to that of the warm star-tracker. The overall performance is excellent; Spitzer can be pointed to an accuracy of less than a half arcsecond and achieves short-term stability much less than 0.1 arcsec.
  • A robust spacecraft. The spacecraft bus shown in figure 2 provides the Spitzer telescope and instruments with a robust and reliable support system. The spacecraft performs various functions including: pointing Spitzer; managing execution of the science programme by clocking through a series of pre-loaded commands; storing, compressing and telemetering the science data to Earth; and monitoring its own health as well as that of the telescope and instruments. The Spitzer spacecraft resembles other unsung heroes in that when it operates smoothly (which it does essentially 100% of the time) we are unaware of its existence. The spacecraft incorporates full redundancy, meaning that there are functional replacements for every one of its critical components. The Spitzer spacecraft includes one expendable substance, high-pressure nitrogen gas, used to spin down the reaction wheel assemblies when they start to accumulate too much angular momentum. Although only a limited amount of gaseous nitrogen is onboard, it should last far beyond Spitzer's cryogenic lifetime. The redundancy and the absence of expendables bode well for the proposal to operate Spitzer – using only the two shortest wavelength IRAC arrays – for up to five years after cryogen depletion.


Spitzer's three instruments are situated in the multiple instrument chamber behind the primary mirror. They share a common focal plane, with their fields of view defined by pickoff mirrors. The instruments achieve great scientific power with an uncomplicated design through the use of state-of-the-art infrared detector arrays in formats as large as 256 × 256 pixels. For broadband imaging and low spectral resolution spectroscopy, Spitzer has achieved sensitivities close to or at the levels established by the natural astrophysical backgrounds – principally the zodiacal light – encountered in Earth orbit. The only moving part in use in the entire science instrument payload is a scan mirror in the Multiband Imaging Photometer for Spitzer (MIPS).

Together, the three instruments provide imaging and photometry in eight spectral bands between 3.6 and 160 μm and spectroscopy and spectrophotometry between 5 and 95 μm. Table 1 summarizes instrument characteristics and performance. The Spitzer Science Center (SSC) at Caltech is responsible for Spitzer science operations, providing instrument handbooks and performance and operability information on its website at http://ssc.Spitzer.caltech.edu. The Jet Propulsion Laboratory, California Institute of Technology, managed Spitzer development for NASA and continues as the overall managing centre while carrying out the mission operations with the help of Lockheed Martin, Denver.

Table 1.  Spitzer instrumentation summary
λ (μm)array typeλ/ΔλF.O.V.pixel size (arcsec)sensitivity*
  1. *Sensitivity numbers are indicative of Spitzer performance in low-background sky and are expressed as the brightness of a source detectable by Spitzer with signal/noise of 5 in 500 sec. In stellar magnitudes, this sensitivity corresponds to a magnitude of 20.7 at 3.6 μm. The limiting magnitude becomes brighter with increasing wavelength. Detailed performance estimates should be based on tools available at http://ssc.Spitzer.caltech.edu/obs. 1 mJy = 10−29 w/m+2/Hz

  2. **Peakup imaging

IRAC: InfraRed Array Camera (PI: G Fazio, Smithsonian Astrophysical Observatory)
3.6InSb4.75.2′× 5.2′1.221.3 μJy
4.5InSb4.45.2′× 5.2′1.212.7 μJy
5.8Si:As (IBC)4.05.2′× 5.2′1.2218 μJy
8.0Si:As (IBC)2.75.2′× 5.2′1.2222 μJy
MIPS: Multiband Imaging Photometer for Spitzer (PI: George Rieke, University of Arizona)
24Si:As (IBC)45.4′× 5.4′2.5110 μJy
70 wideGe:Ga3.55.25′× 2.6′9.87.2 mJy
70 fineGe:Ga3.52.6′× 1.3′5.014.4 mJy
55–95Ge:Ga14–240.32′× 3.8′9.8200 mJy
160Ge:Ga (stressed)40.53′× 5.3′1624 mJy
IRS: Infrared Spectrograph (PI: Jim Houck, Cornell University)
5.2–14.5Si:As (IBC)60–1273.6″× 57″1.8400 μJy
13–18.5**Si:As (IBC)31′× 1.2′1.875 μJy
9.9–19.6Si:As (IBC)6004.7″× 11.3″2.41.5 × 10−18 W m−2
14–38Si:Sb (IBC)57–12610.6″× 168″5.11.7 mJy
18.7–37.2Si:Sb (IBC)60011.1″× 22.3″4.53 × 10−18 W m−2


The SSC maintains the user operational interface for Spitzer. It is responsible for science programme selection and scheduling, data calibration and pipeline processing, and data distribution and archiving. As is done for the other Great Observatories, the SSC annually issues calls for General Observer proposals, and the peer-reviewed competition for observing time is open to scientists from all countries. Spitzer's Legacy Science programme has innovatively engaged the scientific community. Six Legacy Science teams were selected prior to launch and awarded a total of 3000 hours of observing time. These teams have carried out large-scale projects with the joint objectives of creating a coherent scientific legacy and seeding the Spitzer archive with data that will stimulate follow-on proposals from the entire scientific community. All Legacy data are placed in the public archive when they reach the Legacy teams; in addition, the Legacy teams deliver higher order data products such as catalogues and spatial and spectral atlases.

Spitzer differs from HST in that only one instrument operates at a time; there are no parallel observations or internal calibrations. This greatly simplifies the onboard architecture and software with no penalty to data return. Because instrument cold power dissipation dominates helium use, operating two instruments at once would halve Spitzer's lifetime even as the instantaneous data rate doubled. The current scheduling approach is based on instrument campaigns of one-to-two weeks, cycling through the three instruments so that each instrument has at least one “on” period during the 40 days when a particular source at low ecliptic latitude is visible. This campaign scheduling supports a helium strategy in which the telescope temperature is allowed to vary between ∼5 and ∼15 K, as only the MIPS requires the telescope temperature to be as low as 5.5 K. This strategy (Lawrence and Finley 2004) has added about six months to Spitzer's expected cryogenic lifetime.

The Astronomical Observational Template (AOT) is the underlying principle of Spitzer's science operations. The AOT is a web-based form that the user completes to define a particular observation. A completed AOT becomes an Astronomical Observational Request (AOR), which expands directly into a series of spacecraft commands. The executed command sequence consists of a series of AORs, each individually lasting between ten minutes and six hours, interspersed with spacecraft activities such as pointing-system calibrations and data downlink. These command sequences are prepared in one-week blocks starting several weeks before they are uploaded and executed. For a high-priority target of opportunity, a new sequence can be generated and executed on board within about 48 hours of the decision to carry it out. This flexibility was used to observe both a gamma-ray burst afterglow and a microlensing event in the SMC during the first two years of the mission.

The campaign scheduling and the excellent sky visibility from solar orbit allow Spitzer to operate with very high efficiency. In a typical week, only 10% of the wall-clock time goes to downlinking, target-to-target slews and spacecraft engineering activities. Spitzer spends 90% of the time carrying out either observations or necessary campaign-level calibration activities.

Scientific results from Spitzer

There are many more results in hand from Spitzer than can be discussed in a short paper such as this. The initial results from the mission appear in a dedicated special issue of the Astrophysical Journal Supplement 1 September 2004 (Werner et al. 2004, and following papers). Werner et al. (2006) have prepared a review of the first two years of Spitzer galactic and solar system science; a similar review of Spitzer extragalactic results is planned for 2008. Additional early-mission Spitzer science results appear in the proceedings of the first two Spitzer science conferences “New Views of the Cosmos” and “Infrared Diagnostics of Galaxy Evolution”, to appear in the Astronomical Society of the Pacific Conference Series. The SSC maintains an archive of particularly newsworthy Spitzer images, data and scientific results.

We present below a few striking examples of the type of data that Spitzer is returning to the international scientific community.

Star formation

Spitzer's large-area surveys of regions of star formation facilitate the rapid identification and classification of young stellar objects (YSOs). Allen et al. (2004) showed how the IRAC colours of YSOs can discriminate between objects classified as Classes 1, 2 and 3 in the well-recognized evolutionary scheme first put forward by Lada and Wilking (1984). Here Class 1 objects are still embedded in their collapsing protostellar envelopes and have massive circumstellar discs and Class 2 objects have lost their envelopes and have more modest discs. Class 3 objects may show visible-wavelength evidence of accretion, but show little infrared emission above the stellar photosphere. Figure 5 (Allen et al. 2005) shows how the spatial distribution of YSOs in GL4029 aligns with this classification: the youngest objects are most closely clustered together.

Figure 5 : .

Spitzer/IRAC images of the GL4029 region. A bright star off to the west sculpted this complex structure out of a molecular cloud. Circles (top) mark Class I objects identified by their IRAC colours; squares (bottom) show more widely distributed Class II objects. Colour coding by IRAC band: blue = 3.6 μm; green = 4.5 μm; orange = 5.8 μm; red = 8.0 μm. The extended emission is dominantly red from vibration of large organic molecules in the interstellar medium. (NASA/JPL-Caltech/L Allen, Harvard-Smithsonian CfA)

Planetary formation

Spitzer studies the process of planetary system formation by tracing the dissolution of the circumstellar discs around Class 2 objects. The condensation of dust and gas within these discs gives rise to planets and also contributes to the disappearance of the discs. Spitzer observations of stars and clusters of differing ages suggest that the warm dust that arises from the terrestrial planet zone around solar-type stars disappears within a few million years, so that terrestrial planet formation occurs quickly. Some discs, such as that around CoKu Tau 4, shown in figure 6, show dramatic evidence for central clearing possibly attributable to planet formation.

Figure 6 (Left) : .

Figure 6 (Left) : .

IRS spectra of three young stellar objects in Taurus. CoKu Tau 4 shows a flux deficit in the 5–15 μm region, indicative of a central clearing in its circumstellar disc in comparison to that of FM Tau. V928 Tau no longer shows evidence for any circumstellar material, indicating complete dissipation of the protoplanetary disc. (Forrest et al. 2004, revised and reproduced by permission of the AAS). (Right): This series of cartoons, provided by the Spitzer Science Center, shows how a central clearing leads to a drop in the spectrum of a star + disc system similar to that seen in CoKu Tau 4.

Figure 6 (Left) : .

Figure 6 (Left) : .

IRS spectra of three young stellar objects in Taurus. CoKu Tau 4 shows a flux deficit in the 5–15 μm region, indicative of a central clearing in its circumstellar disc in comparison to that of FM Tau. V928 Tau no longer shows evidence for any circumstellar material, indicating complete dissipation of the protoplanetary disc. (Forrest et al. 2004, revised and reproduced by permission of the AAS). (Right): This series of cartoons, provided by the Spitzer Science Center, shows how a central clearing leads to a drop in the spectrum of a star + disc system similar to that seen in CoKu Tau 4.

A significant result from Spitzer's studies of discs and planetary-system formation concerns brown dwarfs, also called young substellar objects. These are bodies with mass less than ∼0.08 solar masses, that do not attain sufficient central pressure and temperature to ignite nuclear burning. They can be seen in the infrared as the heat of formation generated as they collapsed diffuses away. Spitzer has shown that they – even some with masses of 10 Jupiter masses or less (Luhman et al. 2005a,b) – show circumstellar discs as frequently as do more massive objects. This implies that isolated brown dwarfs form by collapse processes similar to those that form stars. Additionally, Apai et al. (2005) find evidence for grain and structure evolution in discs around young brown dwarfs similar to that thought to signal the first steps toward planet formation around young stars. Because brown dwarfs are at least as common in the solar neighbourhood as are stars of all other types, this raises the interesting possibility that the nearest extrasolar planets may be orbiting a brown dwarf, which itself may be thought of as a giant planet.

Nearby normal galaxies

Spitzer's wide wavelength coverage, 3.6–160 μm, samples radiation from a range of different constituents of a galaxy, from stellar atmospheres to cold dust in quiescent interstellar clouds. This scope, together with Spitzer's fundamental imaging field of view of 5 × 5 arcmin, allows Spitzer to provide striking images and new insights into the distribution of stars, interstellar matter and star formation throughout a nearby galaxy. A particularly striking example of this is shown in figure 7, which presents Spitzer 3.6–24 μm images of the nearby spiral galaxy M81. Images and complementary spectra are being obtained for some 50 nearby galaxies by the SINGS legacy team (Kennicutt et al. 2003). This will provide both a marvellous database for studies of nearby galaxies and invaluable ground truth for the interpretation of Spitzer's observations of more distant, spatially unresolved galaxies.

Figure 7 : .

Mosaic of M81 (18 × 23 arcmin across), from Spitzer's multiband imaging photometer and infrared array camera. IRAC 3.6 μm radiation, in blue, traces the stellar bulge, while IRAC 8 μm radiation, in green, picks up vibrational radiation from interstellar hydrocarbons in the spiral arms. MIPS 24 μm radiation, coloured red, highlights warm interstellar dust from recent star formation. (NASA/JPL-Caltech/K Gordon [Univ. Arizona] and S Willner [Harvard-Smithsonian Center for Astrophysics])

The most distant galaxies

Infrared radiation provides a natural probe of the distant and early universe, as the expansion of the universe shifts optical and even ultraviolet starlight from distant galaxies into the infrared. The sensitivity and image quality of Spitzer has permitted detection of galaxies as distant as redshift 6, which means that the light we observe now left the galaxy when the universe was only one-seventh of its present size, and about 10% of its present age. This is illustrated in figure 8, which shows HST, ground-based, and (longward of 3 μm) Spitzer observations of a galaxy at redshift z= 5.8. The break at an observed wavelength of ∼3 μm in the model spectral energy distribution fit to the data lies at the position of the Balmer jump, a persistent feature seen in the spectra of nearby galaxies at a wavelength ∼0.4 μm. The Spitzer observations establish the magnitude of the Balmer jump, which is critically important in determining the mass and age of the stellar population of the galaxy.

Figure 8 : .

HST and Spitzer observations of a galaxy at redshift z= 5.8. The best model fit to the data is a galaxy with mass ∼20% that of the Milky Way which formed at z∼ 10. (Adapted from Eyles et al. [2005])

Scientific utilization of Spitzer

As with NASA's other Great Observatories, Hubble and Chandra, observing time on Spitzer is available competitively to the international scientific community. General Observer Proposals for cycle 4, which will also accept proposals for archival and theoretical studies, are due around February 2007. If the pattern followed in the previous three cycles persists, cycle 4 will award around 6000 hours of Spitzer observing time, and individual programmes requiring as much as 500 hours will be accepted. Proposals for joint programmes with the other Great Observatories, and with ground-based optical and radio facilities, will also be welcomed. Interested scientists should monitor the Spitzer Science Center at http://www.ssc.Spitzer.caltech.edu for the schedule for the proposal process.


The technical and scientific triumphs of Spitzer set the stage for follow-on infrared missions planned by NASA, ESA and the Japanese Space Agency. The most ambitious of these, Japan's SPICA and the NASA/ESA/CSA James Webb Space Telescope, will launch in the next decade. Together, these missions promise the astronomical community continuous access to the infrared spectral band to extend the remarkable exploration begun by Spitzer, just as Spitzer itself builds on the IRAS, ISO and IRTS missions.


In his George Darwin Lecture, Michael Werner celebrates the achievements of the Spitzer Space Telescope, an observatory that owes much of its success to the fact that it is cool by design.