THE WIDE-FIELD INFRARED SURVEY EXPLORER (WISE): MISSION DESCRIPTION AND INITIAL ON-ORBIT PERFORMANCE

It is easiest to observe sources when the Sun is not in the same part of the sky, since maximizing the elongation from the Sun makes it much easier to keep optics cool and to avoid scattered light from the Sun. But to scan the entire sky one must observe the ecliptic poles, which are always 90° from the Sun. Thus, WISE adopted a basic strategy of scanning great circles with a center located at the Sun, which keeps the solar elongation at 90°. These circles scan from the North Ecliptic Pole (NEP) to the South Ecliptic Pole, crossing the ecliptic at longitudes ±90° from the Sun. The entire sky is covered by this scan pattern in half a year as the Earth–Sun line turns by 180°.

Confusion noise would prevent detection of key WISE science targets unless the beam size were less than 50 arcsec², and since there are 5 × 10¹¹ arcsec² in the sky, WISE clearly needs to transmit large quantities of data to the ground. This led to choosing a low Earth orbit to minimize the transmission loss. The actual WISE orbit started with a mean radius of 6909 km in 2009 mid-December, and the radius had decreased by 0.8 km due to atmospheric drag by 2010 mid-June. In a polar orbit, the average geoid height is (R_eq + R_p)/2 = 6368 km, so the average altitude above the geoid is 540 km. This altitude was chosen to reduce the exposure to trapped radiation belts in the South Atlantic Anomaly (SAA). Since looking at the Earth would both blind the detector arrays and overwhelm the cryogenic system, a Sun synchronous orbit was chosen, with an inclination (975) chosen to give a precession rate of 360° in one year. The right ascension of the ascending node was chosen so that the local time at the equator crossings is either 6 AM or 6 PM. For the actual WISE launch at 14:09:33 UT on 2009 December 14 the ascending node is at 6 PM local time. Since the end of the orbit normal close to the Sun is at −75 declination for dawn launch from the Western Test Range (at Vandenberg AFB, CA), during June the angle between Sun and the orbit normal can be as large as 31°, and eclipses occur over the South pole. COBE was in a 99° inclination orbit at 900 km altitude with a June-centered eclipse season. The alternative of a 6 PM launch from the WTR gives an eclipse season centered on the December solstice, as was the case for IRAS.

By synchronizing the scans and the orbit so the telescope is pointed at the NEP when the satellite is closest to the North pole of the Earth, one achieves a scan pattern that covers the entire sky, with the LOS never closer to the Sun than 90° and never further than 31° from the zenith.

WISE has adopted the Two Micron All Sky Survey (2MASS) strategy of moving the telescope at a constant inertial rate and using a scan mirror to freeze the sky on the arrays for the integration time. The field of view (FOV) of WISE is 47', and by choosing to have a small (10%) overlap between frames, the cadence between frames is set at 11.002 s per frame. For WISE 8.8 s of this is spent integrating and it takes about 1.06 s to read out the arrays. The LOS then jumps forward 42' in 1.1 s. There are about 15 orbits per day, so the scan circle advances by about 4' per orbit, giving about 12 orbits where the FOV hits a given source. Figure 2 shows how the redundancy builds up, while Figure 3 shows scans on the celestial sphere. Figure 4 shows the relationship between the WISE orbit, the Earth–Sun line, and WISE line of sight (LOS) during the eclipse season.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The Moon crosses the scan circle twice a month, and this would leave gaps in the sky coverage because the Moon is bright enough to cause stray light problems when it is less than 15° from the LOS. Thus, WISE uses a modified scan pattern where the scan circle gets slightly ahead before the Moon interferes and then drops slightly behind to recover the region the Moon blotted out. The Moon moves 13° per day in ecliptic longitude, so with a 15° exclusion zone WISE needs to be 12 ahead just before the Moon crosses the scan circle, and then drop back to 12 behind. In addition, the SAA affects part of four orbits per day, and to mitigate reductions in the sky coverage due to the SAA, a small toggle of $\pm 0.2\deg$ is added to the longitude of the scan circle center, with the sign changing with a two orbit period. Finally, the scan circle central longitude is biased by +5° on the side of the scan where the Sun is approaching the scan, which is the 6 PM side of the orbit. This means that the scans are actually 90° and 95° from the Sun when crossing the ecliptic, allowing for a 5 day recovery period after a satellite anomaly without causing a gap in sky coverage.

The WISE mission design provides at least eight frames on over 99% of the sky in a 6 month survey interval after allowing for data lost to the Moon and the SAA. Heinrichsen & Wright (2006) describe the mission operations system in more detail. Figure 5 shows the survey progress versus date. In 2010 mid-July, WISE completed its first pass over the sky, but will continue to survey until its cryogen runs out.

Zoom In Zoom Out Reset image size

Duval et al. (2004) and Mainzer et al. (2005) describe the WISE hardware design.

The WISE telescope, imager, dichroic beamsplitter, detectors, and cryostat were built by the SDL of Utah State University with major subcontracts from Lockheed Martin Advanced Technology Center, L3 Communications SSG-Tinsley, and DRS Technologies. Larsen & Schick (2005) describe the WISE science payload in more detail. The instrument was tested, then passed a vibration test, and was tested again, before being delivered to Ball Aerospace in 2009 May for final integration with the spacecraft. The testing included a determination of the optimal focus, a measurement of the point-spread function (PSF), and a determination of both the relative spectral response and the absolute sensitivity of the system.

The WISE short-wavelength channels employ 4.2 and 5.4 μm cutoff HgCdTe arrays fabricated by Teledyne Imaging Sensors with 1024 × 1024 pixels each 18 μm square. For the long-wavelength channels, the detectors are Si:As BIB arrays from DRS Sensors & Targeting Systems with the same 1024 × 1024 pixel format and pitch. The first four and last four pixels in each row and column are used as non-illuminated reference pixels, so the effective size of the arrays is 1016 × 1016 pixels. The median pixel scale is 2757 pixel⁻¹ with a range of ±0.6% among the different axes and arrays. The band 4 data are binned 2 × 2 on board, giving 55 pixel⁻¹. All four arrays image the same FOV simultaneously using three dichroic beam splitters.

All of the arrays use sampling up the ramp. The 11 s cadence between frames is divided into 10 parts. A reset-read of the arrays occurs during the first 1.1 s, followed by eight read cycles each 1.1 s apart and a final reset cycle while the scan mirror flies back. For the W3 and W4 arrays, the nine reads are multiplied by weights of −4, −3, −2, −1, 0, 1, 2, 3, and 4 and summed to give the slope of the ramp. The W1 and W2 arrays showed excess noise in the first sample so the weights used are 0, −7, −5, −3, −1, 1, 3, 5, and 7.

The long-wavelength arrays show long lasting latent images in the flat field which are removed by annealing these arrays twice per day. The annealing heaters are on for 90 s and heat the arrays from their 7 K normal operating temperature to 15 K, which removes the latent images.

The spectral response of the system has been determined in three different ways: the whole system response was measured using a Fourier transform spectrometer (FTS), the system response was estimated using the product of component data (measured with an FTS), and the response was computed from component design calculations. Since the sharp edges in the response function cause ringing and negative values in the FTS spectrum, a "best" combined estimate of the relative spectral response was computed as follows: first, a power law was fit to the ratio of the total system measured response to either the component prediction or the design prediction. The predictions were then adjusted to better match the measured transmission using these fits. The fits were weighted by the square of the response function times a 1/σ² statistical weight, so the correction is best in the filter passbands. After the adjustment the three methods gave consistent responses. Finally, the best estimate for the spectral response was made using a weighted mean of the measured, predicted from components, and design values. In the weighted mean, zero or negative responses were given zero weight, the design and component predictions were given unit weight, and the measurements were given weights of signal-to-noise ratio (S/N)². This weighted mean relative spectral response is shown in Figures 6 and 7. The curves in the logarithmic plot have been normalized to a peak value of 1. Note that the small red leak in the W3 transmission is based on data with a low S/N and may not actually exist.

Zoom In Zoom Out Reset image size

WISE measures the signal S given by

$$\begin{equation} S \propto \int R(\lambda) F_\nu d\ln \nu \propto \int R(\lambda) \lambda F_\lambda d\lambda \end{equation} \tag{ 1 }$$

since the relative spectral response R is electrons per photon and dn_γ ∝ F_νdν/(hν) ∝ F_νdln ν.

Specifying a central wavelength for broad bands like the WISE filters, especially the very wide W3 band, is always ambiguous. We have chosen to specify the isophotal wavelengths for our filters. The isophotal wavelength and magnitude zeropoint are defined by requiring that a constant F°_λ gives the same signal as Vega in a WISE band, and that F°_λ = F^Vega_λ(λ_iso). Tokunaga & Vacca (2005) give a clear definition of the isophotal wavelength and values for Mauna Kea Observatory filter set. Cohen et al. (1992) give the IR spectrum of Vega, based on a model from Kurucz (1991; M. Cohen 2010, private communication). It is important to remember that WISE saturates on Vega, so the magnitude zeropoint is actually based on fainter stars calibrated to the Vega system. Therefore, the debris disk around Vega (Aumann et al. 1984) is relevant only if it has affected the calibration used to determine the magnitudes of the fainter standards. A convenient and accurate fit to the continuum of our Vega spectrum in the 2.5 < λ < 29 μm range we use is given by

$$\begin{eqnarray} F_\lambda & = & 1.0158 \times 10^{-16} (1-0.0083\ln (\lambda /8.891\,\mu \mbox{m})^2) \nonumber \\ & & \times B_\lambda (14454\;\mbox{K}), \end{eqnarray} \tag{ 2 }$$

where B_λ is the Planck function, which matches the continuum spectrum to an average absolute error of 0.045%. The temperature in this ad hoc fit is much higher than the effective temperature of Vega and the solid angle is much lower than the real solid angle due to the increase with wavelength of the free–free and bound–free opacities. Absorption lines reduce the in-band fluxes by only 0.73, 0.57, 0.28, and 0.14% in bands 1 through 4. To avoid possible multiple solution we use the continuum spectrum when solving for λ_iso. The isophotal wavelengths so defined are 3.3526, 4.6028, 11.5608, and 22.0883 μm for W1,..., W4. The measured system transmissions were smaller than the expected values at long wavelengths, leading to an effective wavelength for WISE band 4 of 22 μm instead of the expected 23 μm. Absolute measurements of Vega by MSX (Price et al. 2004) show that a 2.7% upward offset from the model spectrum is needed at 21.3 μm (Cohen 2009). We have applied this correction to the WISE 22 μm band, giving magnitude zeropoints on the Vega system in the WISE passbands of F°_λ = 8.180 × 10⁻¹⁵,  2.415 × 10⁻¹⁵,  6.515 × 10⁻¹⁷, and 5.090 × 10⁻¹⁸ W cm⁻² μm⁻¹, which convert to F°_ν = 306.681,  170.663,  29.0448, and 8.2839 Jy in W1,...,W4 using F°_ν = (λ²_iso/c)F°_λ. There is an overall systematic uncertainty of ±1.5% from the Vega spectrum in these flux zeropoints.

Zoom In Zoom Out Reset image size

We have also observed in-flight a discrepancy between red (typically sources with F_ν ∝ ν⁻²) and blue (stars with F_ν ∝ ν²) calibrators in W3 and W4, the 12 and 22 μm bands. This amounts to about −17% and 9% in the fluxes for W3 and W4, with the red sources appearing too bright in W4 and too faint in W3. The flux differences could be resolved by adjusting the effective wavelength of W3 and W4 3%–5% blueward and 2%–3% redward, respectively. This would change the zero magnitude F°_ν by about −8% in W3 and +4% in W4. But the zeropoints and isophotal wavelengths reported here are based on the relative spectral responses derived from ground calibration without any such adjustment. The instrumental zeropoints that define the conversion from counts to magnitudes have been based on standard stars, which are the blue calibrators. Given the discrepancy between red and blue calibrators we estimate that the conversion from magnitudes to Janskys are currently uncertain by ±10% in W3 and W4. Updated values will be provided in the Explanatory Supplement accompanying the preliminary data release.

This definition of the isophotal wavelength and flux zeropoint means that the color correction term for a source with a different spectrum than Vega vanishes by construction when F_λ is a constant (F_ν ∝ ν⁻²) and very nearly vanishes for Rayleigh–Jeans sources with F_λ ∝ λ⁻⁴ or F_ν ∝ ν². These spectral energy distributions bracket the vast majority of WISE sources, so the color corrections are generally small. But the extremely wide W3 filter does lead to color corrections as large as 0.1 mag for a constant F_ν. Table 1 gives the flux correction factors in the WISE bands for several input spectra and the WISE colors for these spectra. These factors multiply the signal S given by a spectrum F_ν = F°_ν(λ_iso/λ)^β. Thus, a spectrum with F_ν = const = 29.0 Jy gives a signal that is f_c(W3) = 0.9169 times the signal from Vega, so one would need a constant F_ν of 29.0/0.9169 = 31.7 Jy to give zero magnitude.

Outer solar system objects like Centaurs will have color temperatures close to 100 K and for a 100 K blackbody the flux correction factor f_c(W3) = 2.6588 is quite large. Thus, a 100 K blackbody would need 29.0/2.66 = 10.9 Jy at 11.56 μm to give zero magnitude in W3.

The most common stars in the WISE catalog at high galactic latitude should be G-K dwarfs, so the flux correction factors and colors for a K2V and a G2V have been included in Table 1 using Kurucz (1993) spectra. Since the W2 band includes the fundamental CO bandhead, the flux correction factors give the ratio between the signal from the star to the signal from a constant F_λ equal to the average over a log normal passband with a 9.1% FWHM centered on λ_iso.

2.2.1. Optics

2.2.2. Cryostat

The WISE spacecraft was built by Ball Aerospace, with a design based on the RS300 series of single string spacecraft. The NextSat component of the Orbital Express mission was the first RS300 spacecraft to be launched, and it worked flawlessly. SWRI provided the spacecraft avionics.

WISE is a three-axis controlled spacecraft that is commanded to follow long scans at a constant inertial rate. The spacecraft provides attitude control of better than 75'', jitter of less than 13, and drift rate variation of less than 02 s⁻¹ over 9 s. Angular momentum is stored on board in four reaction wheels, and any buildup of excess angular momentum is dumped using magnetic torquer rods. Primary attitude information is provided by two Ball CT-633 star trackers. A fiber optic inertial measurement unit, Sun sensors, and magnetometers are used during safe and emergency modes of the spacecraft.

A fixed solar panel provides over 500 W of power. WISE is oriented so the solar panel is always pointing very nearly at the Sun. During eclipse, a 20 A hr lithium-ion battery provides flight system power. Science data and flight system telemetry are stored on a 96GB flash memory card for later transmission to Earth. The science data volume is about 50 Gbyte per day of uncompressed data, but loss-less compression reduces this by roughly a factor of two. A fixed high gain antenna is used to transmit data to the Tracking and Data Relay Satellite System (TDRSS). The spacecraft body steers the antenna to TDRSS for downlink. Science operations have to be stopped during data transmission, so the data downlinks are scheduled while the satellite passes over the poles of the Earth, where the sky coverage is highly redundant.

Except for looking up instead of down, WISE is very similar to an Earth observing satellite. WISE is in a Sun-synchronous low Earth orbit like the Landsats and many other Earth observers, and it uses the TDRSS to download a large volume of data. Thus, WISE is being operated by the Earth Sciences Operations Center at JPL which also operates the JASON ocean topography mapping mission.

The preparation of a command load for WISE starts with a set of survey planning parameters, which then are used along with a predicted orbit to generate the survey plan for a week in the future. Times when a turn to TDRSS would be compatible with the WISE Sun and Earth pointing constraints are chosen, and a complete sequence of pointing commands covering half a week that combines observing with TDRSS downloads is created. Additional commands for momentum unloading and annealing the W3 and W4 detector arrays are inserted into the command sequence. This process is repeated twice a week to allow for the uncertainties in the predicted orbit caused by the fluctuating atmospheric drag due to solar activity. In the first half of 2010, the solar activity has been anomalously low, minimizing the errors in the predicted orbits.

WISE science data processing, archiving, and distribution are performed by the IPAC, California Institute of Technology. The processing software and operations system is based on algorithms, pipelines, and architecture used for 2MASS, the Spitzer Space Telescope, and GALEX projects.

Following each TDRSS downlink contact, science image data packets are sent to IPAC where they are uncompressed and assembled into raw FITS format images in the four WISE bands. The raw FITS image headers are populated with flight system engineering telemetry, navigation data, and survey events information that is sent asynchronously from Mission Operations at JPL. The images from each downlink transfer are then passed to the WISE scan/frame pipeline that performs instrumental calibration (droop correction, linearization, bias subtraction, flat fielding), source extraction, photometric and astrometric calibration, and artifact identification on the individual exposure frames. The WISE Multiframe pipeline operates on collections of calibrated single-frame images that cover a region on the sky. The single-frame images are co-added by registering them onto a common pixel grid, matching local background levels, and optimally combining resampled pixel values using outlier rejection to suppress transient pixel events such as cosmic rays, noise excursions, and fast moving objects (Masci & Fowler 2009). Sources are detected and characterized on both the resulting combined images and individual frames, and spurious detections of image artifacts are identified and flagged. Both scan/frame and multiframe processing are accompanied by detailed quality assessment to monitor data accuracy relative to WISEs science data requirements.

Source detection and characterization in the scan/frame and multiframe pipelines exploit WISEs simultaneous four-band measurements. Sources are detected by thresholding on a combined four-band, matched-filter image, similar to the "chi-square" detection method described by Szalay et al. (1999). Profile-fit photometry is performed on all bands simultaneously, using a maximum-likelihood fitting procedure that also operates on multiple, independent image frames in the multiframe pipeline.

Astrometric calibration of single-exposure WISE images is performed by solving for the frame center positions and rotations using the in situ measured position of 2MASS point source catalog astrometric reference stars. Band-to-band rotations and offsets, plate scales, and field distortion are determined by off-line analysis using the apparent positions of reference stars in thousands of image frames. Full WCS information derived from the astrometric solutions is included in the single-frame image headers. The single-frame astrometric solutions are propagated to the co-added images and extracted source lists in the multiframe pipeline.

Processed WISE images and extracted source data and metadata will be archived and served to the community via the Web-based and machine-friendly interfaces of the NASA/IPAC Infrared Science Archive (IRSA). WISE data products will be interoperable with other data centers and services through the IRSA infrastructure.

Brown dwarf stars are very faint, since they are not massive enough to fuse hydrogen into helium. As a result, they gradually fade and cool, and old brown dwarfs will be very cool and faint. Both Jupiter with L = 10⁻⁹ L_☉ and Gliese 229B with L = 10⁻⁵ L_☉ have very strong emission at 4.6 μm due to a lack of methane absorption at this wavelength (Kirkpatrick 2005). Thus, the 4.6 μm band of WISE is a powerful tool for finding cool brown dwarfs. Most of the brown dwarfs found prior to WISE have been discovered using the z' band of SDSS or the J band of 2MASS or UKIDSS, but fairly high temperatures are required before there is substantial emission in either of these short-wavelength bands. As a result the currently known sample of brown dwarfs is biased toward the hotter, and thus younger, objects. WISE is able to find 10 Gyr old brown dwarfs and thus should detect a high density of stars in the solar neighborhood. The expected number density of brown dwarfs is one to two times the density of ordinary stars (Reid et al. 1999; Chabrier 2002).

The expected number of brown dwarfs that WISE will see can be computed using models for the emitted spectra of brown dwarfs and the luminosity and effective temperature of brown dwarfs as a function of mass and age (Burrows et al. 2003). These model spectra are plotted as νF_ν/F_bol in Figure 13. The spectra can also be used to compute the dimensionless ratios of νF_ν/F_bol for each WISE band, which turn out to be well described by simple functions of the effective temperature T_e. The functional forms plotted in Figure 14 are given by

$$\begin{equation} \frac{\nu F_\nu }{F_{\rm bol}} = 0.16 \frac{15}{\pi ^4} \frac{x^4}{e^x-1} \end{equation} \tag{ 3 }$$

with x = hν/(1.34kT_e) for band 1 in blue,

$$\begin{equation} \frac{\nu F_\nu }{F_{\rm bol}} = 2.25 \; \mbox{min}(T_e/230,1)^2 \; \frac{15}{\pi ^4} \frac{x^4}{e^x-1} \end{equation} \tag{ 4 }$$

with x = hν/(2.29kT_e) for band 2 in green,

$$\begin{equation} \frac{\nu F_\nu }{F_{\rm bol}} = 0.55 \frac{15}{\pi ^4} \frac{x^4}{e^x-1} \end{equation} \tag{ 5 }$$

with x = hν/(1.05kT_e) for band 3 in orange, and

$$\begin{equation} \frac{\nu F_\nu }{F_{\rm bol}} = 0.9 \left\lbrace \begin{array}{@{}l@{\quad }l}\exp [-5.4\log (T_e/170)] & {\rm if}\ T_e > 317 \cr \exp [-20\log (T_e/170)^2] & {\rm otherwise}\end{array}\right. \end{equation} \tag{ 6 }$$

for band 4 in red. Thus, WISE band 2 sees a flux for T_eff>230 K that is similar to a graybody with a temperature 2.25 times higher than T_eff, and with an emissivity 2.25 times higher than a blackbody. The fluxes in WISE bands 1 and 3 also look like higher temperature graybodies but with emissivities much smaller than a blackbody. The flux in WISE band 4 does not behave like a blackbody.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

But when Eisenhardt et al. (2010) surveyed 10 deg² with Spitzer to a flux limit at 4.5 μm 20 times deeper than the WISE all sky sensitivity, they found fewer very red sources than predicted by combining the Burrows et al. (2003) models with either the Chabrier (2003) log normal IMF or the Reid et al. (1999) power laws. The models could be brought into agreement with the observed counts by assuming that about 50% of the IRAC 4.5 μm flux was suppressed by an absorber other than methane or water vapor. The amount of absorption needed is highly correlated with the assumed mass function, but the Eisenhardt et al. (2010) observation of about 10 brown dwarfs gives a model-independent prediction that WISE should find about 10(41253/10)/20^1.5 ≈ 400 late-T and Y class brown dwarfs. Table 2 shows the number of brown dwarfs in various temperature or distance bins that are detectable by WISE for different assumed mass functions. The age distribution is assumed to be uniform from 10⁸ to 10¹⁰ yr, and the mass function is integrated from 1 to 80 M_J. For each mass and age bin, the luminosity and effective temperature are found by interpolation. A fraction S of the 4.6 μm flux is then suppressed, following Golimowski et al. (2004), who found a suppression factor 1/(1 − S) = 1.5 to 2.5, implying S = 0.5 ± 0.1. The volume in which the object could be detected is computed, multiplied times the space density for the mass and age bin, and summed over all masses and ages to give the numbers in Table 2. There are two lines for each mass function: the first does not include the W2 flux suppression, and the second has S adjusted to give the expected 400 detectable objects with T_e < 750 K. Matching the Eisenhardt et al. (2010) source density with a flux suppression in the range recommended by Golimowski et al. (2004) requires a high number density of brown dwarfs, as given by Chabrier (2003) or Reid et al. (1999) with a fairly steep power-law index.

The WISE sample of brown dwarfs will be 3 mag brighter than the Spitzer sample of Eisenhardt et al. (2010), and thus much easier to follow up with the Hubble Space Telescope (HST) or the Keck telescopes. Astrometric follow-up will be needed to pin down the properties of old, cold brown dwarfs.

WISE will detect the most luminous galaxies in the universe. The Ultra-luminous Infrared Galaxies (ULIRGs) are due to mergers that lead to dust-enshrouded star formation and also to active galactic nucleus (AGN) activity as gas is disturbed out of stable circular orbits and falls into the central super-massive black holes in the merging galaxies. Figure 15 shows typical ULIRG and QSO spectra compared to the WISE flux limits.

Zoom In Zoom Out Reset image size

In the concordance ΛCDM cosmology structure formation proceeds in a bottom-up manner, so large objects are formed by the mergers of smaller objects. The peak of large mergers occurs at redshifts 1 to 2 and leads to the peak in the star formation history of the universe, and WISE has sufficient sensitivity at 22 μm to see the top end of the luminosity function at redshift 3. As a result, WISE will detect many galaxies with L>10¹³ L_☉. Quasar activity is also associated with mergers, and the evolving quasar luminosity function of Hopkins et al. (2007) predicts that WISE will see over 1.5 × 10⁵ quasars at 22 μm and over 3.6 million quasars at 4.6 μm, with redshift distributions shown in Figure 16.

Zoom In Zoom Out Reset image size

The abundance of ULIRGs as a function of luminosity and redshift will provide important information about the formation mechanisms that have led to the galaxies and clusters we see at low redshift today.

4.2.1. Red/Dusty AGNs and QSOs

Asteroids are typically fairly dark, but do have a large range of albedos. As a result, the visual magnitude is an imprecise indicator of the size of an asteroid. Stuart & Binzel (2004) show a range in albedos from 0.023 to 0.63 for near Earth asteroids, leading to a 5:1 range in estimated diameters and a 125:1 range in volumes. But the light that is not reflected is absorbed and reradiated in the thermal infrared. Figure 17 shows fluxes from an asteroid with low, median, and high albedos. The ratio of optical to infrared fluxes gives the albedo, and the infrared flux and color temperature give a good indication of the diameter of an asteroid.

Zoom In Zoom Out Reset image size

WISE is sensitive enough to see 1.2 km diameter asteroids in the main belt at 2.7 AU from the Sun, and 0.13 km diameter near Earth objects (NEOs) that are 0.5 AU from the Earth, and 1.1 AU from the Sun. These limits are for objects with known orbits so that multiple frames can be stacked and are for 5σ. This gives 10% accuracy in diameters, which is comparable to the errors caused by the unknown thermal inertia and rotation poles of most asteroids (Wright 2007). WISE will be able to provide radiometric diameters for about 3 × 10⁵ asteroids using image stacking analysis. For an asteroid with a bolometric albedo of A = 5%, the WISE stacked frame sensitivity of 800 μJy at 12 μm corresponds to an optical flux of roughly

$$\begin{equation} F_\nu ({\rm opt}) \approx \frac{\nu _{\rm IR}}{\nu _{{\rm opt}}} \frac{A}{1-A} F_\nu ({\rm IR}) = 2\;\mu \mbox{Jy,} \end{equation} \tag{ 7 }$$

which is an optical magnitude of 23. If good optical data are available, all of these asteroids will have well-determined albedos.

For new asteroids, a diameter of 2.4 km in the main belt or 0.25 km for a NEO at a distance of 0.5 AU gives enough S/N in a single frame to allow detections. The NEOWISE program (A. K. Mainzer et al. 2010, in preparation) is searching the WISE data for moving objects and sending positional information to the IAU Minor Planet Center. WISE should detect more than 10⁵ asteroids on single frames.

WISE provides about 10 measurements of the flux spaced over 30 hr for a typical asteroid which can be used to derive rotation periods for tens of thousands of asteroids.

The accurate diameters for previously known asteroids plus new asteroids derived by the WISE task will allow a significantly more precise knowledge of the size frequency distribution (SFD) of sub-kilometer NEOs. Since only a few dozen sub-kilometer objects currently have diameters that have been accurately measured by either radar or thermal IR data, WISE data will significantly improve our knowledge of the SFD of these smaller objects. Diameters derived from WISE measurements will reduce uncertainties in the SFD associated with the traditional reliance on the optical H magnitude. Also, since infrared discoveries of new asteroids are relatively insensitive to albedo, the set of new discoveries made by WISE are much closer to a diameter-limited survey than optical surveys, resulting in much better SFDs.

By observing asteroids at the time of their local afternoon and comparing this to their thermal fluxes in their local morning, WISE would enable measurement of the temperature differentials that give rise to the Yarkovsky effect, a phenomenon that is often the dominant uncertainty in the long-term orbit propagation for small asteroids. Measuring the Yarkovsky effect via the thermal infrared provides rather different information for a different population than the radar method (Chesley et al. 2003). The radar technique uses precision observations of an asteroid's trajectory over a decade to look for small changes to the ephemeris caused by the Yarkovsky force. In order to obtain sufficient astrometric precision, only NEOs which are very nearby can be observed. Radar data allow one to compute the tangential acceleration a produced by the Yarkovsky effect. However, using infrared to look for temperature differentials between an asteroid's local morning and evening requires only two epochs of observation separated by 3–8 months. The infrared technique is not limited to NEOs and will also work for Main Belt asteroids such as the Baptistina family. The infrared data allow one to compute the diameter which gives the mass m, and the morning versus evening difference gives the tangential force F. Thus, combining IR and radar data allows measurements of all the quantities in F = ma. For asteroids in a family, the change in semi-major axis divided by the age of the family gives an estimate for the acceleration a, so the IR data can be used to verify the time since the collision that created the family. The hypothesis (Bottke et al. 2007) that the K/T impactor was part of the Baptistina family depends on the estimated 160 Myr age of this family, so measuring the Yarkovsky force and diameter for a sample of Baptistina family members will provide a valuable test of this idea. The infrared Yarkovsky detection technique is improved by knowledge of an objects light curve, well measured by WISE.

An all-sky survey in the thermal infrared can be used to address many scientific questions, not all of which require surveying the whole sky. But by surveying the whole sky, WISE will enable these projects among many others.

4.4.1. Comet Trails and Zodiacal Bands

4.4.2. Young Stars and Debris Disks

WISE will provide a robust statistical database for studying star formation, the evolution of circumstellar disks, and the dissipation of zodiacal clouds around stars of all masses, extending even into the brown dwarf mass regime. Detecting the small IR excess from the later stages of disk dissipation requires the ability to accurately predict the long-wavelength photospheric emission and the ability to accurately measure the total long-wavelength emission. The simultaneous four color photometry provided by WISE will enable the required predictions, and the sensitivity of WISE will enable detection of optically thick disk emission for solar-type stars out to 1 kpc distance and down to the hydrogen-burning limit in the Taurus and Ophiucus star-forming regions. Because young stars are variable, the simultaneous four-color coverage of WISE provides a precise means of characterizing thermal excess from circumstellar disks or zodiacal material. (A possible second coverage of the object by WISE 6 months after the first will even provide clues as to the nature of the variability, e.g., extinction versus accretion.)

Because WISE is an all-sky survey it will reveal isolated young stars well outside obvious star-forming regions. For example, two of the nearest clusters (of any age) were discovered only within the last decade though their young stellar objects (YSOs) like properties: the TW Hydrae Association (d = 55 pc, by infrared emission) and the Eta Chamaeleonis Cluster (d = 96 pc, by X-ray emission). Both have about a dozen known members and are younger than 10 Myr. Surprisingly neither is associated with a molecular cloud. TW Hya covers more than 700 deg². A complete survey would be an impossible task for missions like Spitzer but is the natural focus of WISE. Such intermediate-age very-nearby clusters provide an ideal laboratory for studying the late stages of planetary accretion.

WISE will find young stars which are still accreting molecular cloud material over a substantial fraction of the Galaxy: WISE will measure stars with optically thick circumstellar disk emission down to the hydrogen-burning limit (0.08 M_☉) out to 500 parsecs (e.g., the Orion molecular cloud). Solar-type "Class I" and T Tauri stars (TTS) with optically thick disks will be found to substantial distances (Figure 18 shows the M0 TTS Sz 82 scaled to a distance of 2.5 kpc), thanks to their strong excesses at 12 and 22 μm. Expanding upon Spitzer's survey of selected regions of a few molecular clouds, WISE will provide a full inventory of low-luminosity YSO in many dozens of star-forming clouds extending over the entire sky. WISE will also discover hundreds of the very youngest stars: the elusive "Class 0" sources. These objects, which are in the early stages of gravitational collapse, are rare because they remain in this phase for less than 100,000 yr. They are deeply embedded in dust, so they are visible only at wavelengths longer than 20 μm, although they may be seen in scattered light in the mid-IR.

Zoom In Zoom Out Reset image size

WISE will survey PDDs around thousands of stars in the solar neighborhood. During the early stages of planet formation micron-sized grains accrete into successively larger particles. The resulting reduction in grain surface area makes the circumstellar disk optically thin, and the infrared excess substantially decreases. The disk does not clear uniformly due to both the variation in accretion/dynamical timescale with distance from the star, and the shepherding effects of forming planets (Kuchner & Holman 2003). This non-uniformity is evident in the distribution of infrared excess as a function of wavelength through the mid-infrared. WISE will exploit these signatures to reveal the systematics of and refine the timescales for disk clearing and solar system formation.

Although the planetesimal/planet formation phase is short-lived, dust excess signatures can persist into the main-sequence phase as demonstrated in stars like Vega by IRAS (Aumann et al. 1984). These excesses arise from small grains with Poynting–Robertson lifetimes of tens of thousands of years. Their existence argues that they must be actively replenished by the erosion of planetesimals and thus they are signatures of successful planetary formation. Indeed, high-resolution 20 μm imaging of infrared-excess objects like HR4796 reveal dust rings which may betray the existence of invisible planets. Figure 18 also shows the known debris disk system ζ Lep scaled to a distance of 300 pc. With the short-wavelength WISE bands pinning down the photospheric flux, the excess at 22 μm is easily detectable.

The sensitivity to see the photospheric flux and enough wavelength coverage to predict the photometric flux are, once again, the key to revealing excesses from eroding planetesimals. In addition to studying the statistical distribution of such small IR excesses, of great interest to the TPF mission, WISE will provide James Webb Space Telescope (JWST) with hundreds of targets for high-resolution imaging akin to the ground-based and HST images of HR4796.

4.4.3. Interstellar Dust

4.4.4. Galactic Structure

4.4.5. Nearby Galaxies

The angular resolution and sensitivity of WISE allow detailed study of the internal structure of galaxies in the local universe, whose properties provide our basic understanding of high-redshift galaxies. For the ∼5000 galaxies within 20 Mpc of the Sun, WISE spatial resolution is <500 pc (<1 kpc at 22 μm). Of this set, nearly 400 will be highly resolved, delineating globular clusters, giant molecular clouds, and other discrete sites of star formation. Figure 19 shows WISE images of four well-resolved nearby galaxies. The short mid-IR bands of WISE are an effective probe of the underlying stellar population and are nearly immune to extinction. The long bands are sensitive to hot dust associated with star formation, thus bridging the time span between evolving generations of stars. WISE data will complement the Spitzer Infrared Nearby Galaxies Survey Legacy project (SINGS; Kennicutt et al. 2004), expanding the SINGS sample of 75 galaxies to many thousands of galaxies, covering the widest range in morphological type. It will also supplement the IRAC 3.6/4.5 μm Spitzer Survey of Stellar Structure in Galaxies (Sheth et al. 2008): WISE will provide the crucial mid-IR tracers of star formation activity for this complete sample of galaxies within 30 Mpc.

Zoom In Zoom Out Reset image size

4.4.6. Galaxy Clusters