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

This paper describes the Wide-field Infrared Survey Explorer (WISE) mission and reports its initial on-orbit performance. The central “research question” is not a hypothesis test in the usual sense, but rather: can WISE successfully execute an all-sky mid-infrared survey with the designed sensitivity, angular resolution, and astrometric precision, and what are the early in-flight characteristics that determine the quality of the resulting astronomical catalogs? This matters because all-sky infrared surveys underpin long-lived reference datasets for essentially every subfield that needs uniform mid-IR photometry—solar system science (asteroids, comets), stellar astrophysics (brown dwarfs, young stellar objects, debris disks), galactic structure (stellar populations and dust), and extragalactic astronomy (dusty galaxies, obscured AGN, quasars).

The paper’s significance within the field is framed by comparison to prior all-sky infrared surveys, especially IRAS (1983) and 2MASS/DIRBE. WISE uses four mid-IR bands centered at 3.4, 4.6, 12, and 22 m (W1W4) and a 40 cm telescope feeding large-format detector arrays (4 million pixels total). The authors emphasize that the increased detector count and improved observing strategy yield dramatically better sensitivity than earlier missions. In the 12 m band, they state WISE is more than 100 times more sensitive than IRAS.

Methodologically, the paper combines (i) mission design choices (orbit, scanning law, cadence, redundancy), (ii) instrument characterization (detector types, readout scheme, spectral response calibration, photometric/astrometric calibration approach), and (iii) early on-orbit performance metrics derived from the first sky coverage. The study design is therefore an engineering-to-science validation: it uses in-flight measurements of noise, point-spread function, astrometric residuals, saturation behavior, and the realized point-source sensitivity as a function of coverage (number of frames). The “data sources” are WISE in-flight observations (single frames and coadded multiframe products), cross-matched astrometric reference catalogs (2MASS PSC and UCAC3), and ground-based instrument calibration measurements (including Fourier transform spectrometer measurements of system response).

Key quantitative results include the achieved point-source sensitivities and angular/astrometric performance. The abstract reports 5sigma point-source sensitivities better than 0.08, 0.11, 1, and 6 mJy in unconfused regions on the ecliptic in W1W4 (3.4, 4.6, 12, 22 m). Sensitivity improves toward the ecliptic poles due to denser coverage and lower zodiacal background. The angular resolution (FWHM) is reported as 6.1arcsec, 6.4arcsec, 6.5arcsec, and 12.0arcsec for W1W4. For astrometry, the paper states that for high signal-to-noise sources the astrometric precision is better than 0.15arcsec (1axis, 1sigma).

The on-orbit performance section provides additional detail on how sensitivity was derived. The authors note that the specified sensitivity is for 8 coverages (frames), but that even on the ecliptic the most likely number is 11. They analyze the scatter among individual-frame measurements to infer magnitude noise. They present a noise model of the form $\sigma (m)=0.01+[2.5/\ln (10)]n/10^{-0.4m}$ and identify the magnitudes where their curves cross a dashed reference line corresponding to a 5:1 signal-to-noise in 8 frames. Those crossing magnitudes are 17.11, 15.66, 11.40, and 7.97 mag, corresponding to raw 5sigma sensitivities of 44, 93, 800, and 5500 $\mu$Jy in W1W4 for 8 frames. They then account for confusion noise (not included in the statistical noise estimate) to obtain the final confusion-limited sensitivities of 0.08, 0.11, 1, and 6 mJy.

Astrometric accuracy is assessed by comparing WISE positions to UCAC3 for sources with SNR>20. At lower SNR, they add an additional centroiding error term approximately scaling as $\mathrm{FWHM}/(2\mathrm{SNR})$ in quadrature. Using the observed distribution width, they infer a scatter corresponding to $\sigma =0.17$ arcsec for SNR>20, and conclude that WISE positions for high SNR sources should be better than 0.15 arcsec for 1sigma and 1 axis.

The paper also reports instrument and operational constraints that affect performance. A major limitation is cryogen lifetime and resulting band degradation. The cryostat was predicted to have an on-orbit lifetime of $10.75_{-0.5}^{+1}$ months, but in-flight performance deviated: the secondary hydrogen tank ran out on 5 August after 7.7 months in orbit. The telescope warmed to 46 K, causing large backgrounds at 12 and 22 m. The 22 m channel stopped producing useful data after 8 August 2010. The 12 m integration time was progressively reduced (cut in half on 14 August, again on 20 August, and again on 23 August), and observations continued in W1W3 until the primary tank ran out on 29 September 2010. This is an explicit acknowledgment of a performance limitation relative to design.

Other limitations are implicit in the methodology: sensitivity estimates depend on the number of frames and on whether regions are “unconfused,” and the authors explicitly note that confusion noise is not included in the statistical noise curves. Additionally, photometric calibration uncertainty is discussed for W3 and W4 due to an observed discrepancy between red and blue calibrators in-flight. They estimate that the conversion from magnitudes to Janskys is currently uncertain by $\pm 10\%$ in W3 and W4, with updated values promised in an explanatory supplement.

Practical implications are broad. For astronomers, the paper provides the mission’s realized capabilities that determine survey completeness and selection functions: the four-band photometry with specified sensitivities, the angular resolution enabling source separation down to the confusion limit, and the astrometric precision enabling cross-matching with optical/near-IR catalogs. For solar system researchers, the sensitivity and cadence enable asteroid detection and radiometric diameter estimation via thermal emission; for example, the paper states WISE can provide radiometric diameters for about $3\times 10^5$ asteroids using image stacking, and detect new asteroids at the single-frame level (with NEOWISE expected to find more than $10^5$ asteroids on single frames). For extragalactic science, the 22 m band (until cryogen depletion) and the 12 m band enable detection of luminous dusty galaxies and obscured AGN across large redshift ranges.

Overall, the paper functions as a mission validation and performance characterization report. Its core contribution is demonstrating that WISE can deliver the promised all-sky mid-infrared survey quality—especially the achieved 5sigma sensitivities (0.08, 0.11, 1, 6 mJy in W1W4), the $\sim 6$ arcsec angular resolution in the short bands, and $<0.15$ arcsec astrometric precision for high SNR sources—while transparently documenting the cryogenic limitation that curtailed the longest-wavelength channel and reduced 12 m integration time later in the mission.