I. Introduction

4. Cautionary Notes

The WISE survey data acquisition strategy, data processing and quality assurance measures were all designed to yield data products of high reliability and fidelity. However, the accelerated schedule of the Preliminary Release combined with challenge of validating such a large data set in a limited time, resulted in a number of limitations and problems that persist into the Preliminary Release products.

The following sections contain summaries of known features and limitations that pertain to WISE Preliminary Data Release data products. Section I.4.a presents general Cautionary Notes. Sections I.4.b and I.4.c address essential information concerning the Source Catalog and Image Atlas, respectively. Section I.4.d contains Cautionary Notes for users who make use of the Release Ancillary Data products, including the Single-Exposure Source Database and Image Archive, and the Known Solar System Object Possible Candidate Association Table.

Users are strongly encouraged to read and familiarize themselves with these Cautionary Notes before using the WISE Preliminary Release data products.

a. General

i. First-Pass Processing Limitations

The WISE Preliminary Release data products are based on the results of the first-pass processing of WISE data. First-pass processing used an early version of the WISE Science Data System (WSDS) that utilized calibration information and data reduction algorithms based on pre-launch ground and very early in-flight tests. It was not possible to incorporate into this early processing system version calibrations and algorithms that optimally deal with all on-orbit data characteristics, particularly for features that were not predicted by ground testing.

All WISE cryogenic mission data will be reprocessed during the planned second-pass processing, using an updated version of the WSDS that incorporates the best available calibrations, knowledge of the flight system performance, and improved reduction algorithms that are based on analysis of data and flight system performance from the full survey. The results of the second pass processing will be used to generate the products for the WISE Final Data Release in the Spring of 2012.

ii. Sky Coverage

1. The WISE Preliminary Data Release covers approximately 57% of the sky

The WISE Preliminary Release does not cover the entire sky. By design, the WISE Preliminary Release products were generated using only data acquired during the first 105 days of cryogenic survey operations (V.2). This resulted in products that cover approximately 57% of the sky (~23,600 deg²), in two large contiguous regions approximately bounded by ecliptic longitude limits 27.8°<λ<133.4° and 201.9°<λ<309.6° (Figures 1, 2 and 3).

The WISE Final Data Release scheduled for the Spring of 2012 will incorporate all cryogenic survey data covering the full sky.

Detailed statistics of the effective sky coverage in the WISE Preliminary Data Release are presented in VI.2.

Figure 1 - Equatorial projection	Figure 2 - Ecliptic projection
Figure 3 - Galactic projection
Sky maps showing the ~57% of the sky nominally covered by the Preliminary Data Release. The colors encode the average W1 frame depth-of-coverage in 15'x15' spatial bins. Atlas Image and Source Catalog data may be missing for regions with very low or no coverage within the Release Area.

2. The Preliminary Release outer boundaries are complex

The sky was tessellated into an equatorial grid of 1.564° x 1.564° Atlas Tiles for the purpose of combining Single-exposure image data and performing sensitive source extraction in Multiframe pipeline processing. The Preliminary Release area is comprised of 10,464 of those Tiles. Because Atlas Tile Right Ascension boundaries do not align between declination bands, the outer edges of the Preliminary Release area are irregular (Figure 4). Determining if a particular object or location on the sky that is near the boundaries can be difficult using a low resolution map.

The following online service uses actual Tile boundaries to determine if a user-specified object or location of interest is within the nominal boundaries of the Preliminary Release, and if so, in which Tile it is located.

WISE Preliminary Data Release Tile Lookup Service

The coordinates of Preliminary Release Tile corners in ecliptic and unit sphere coordinates are also provided in Table 1 of Section V.2.b.

Figure 4 - A color-composite mosaic of Preliminary Release Atlas Images (equatorial projection) for a region near the north ecliptic pole illustrating the complex outer boundaries of the Preliminary Release area. See Figure 7 in VI.2 for a full-sky Atlas Image mosaic.

As shown in Figures 1, 2 and 3, the effective depth-of-coverage for the WISE Preliminary Data Release varies considerably around the sky because of the WISE survey strategy and operations, the delivered data quality, and the Atlas and Catalog production algorithms. As discussed in VI.2, the effective depth-of-coverage for the Preliminary Release is also slightly different for the different bands. Atlas Images and Survey Catalog sensitivity will vary correspondingly with the depth-of-coverage in each band.

The systematic increase in coverage depth towards the ecliptic poles, visible in Figures 1-3, is a natural product of the WISE scanning strategy (III.3). Multiple, independent exposures on each point of the sky build up the net exposure depths to achieve the survey's sensitivity goals and enable effective suppression of single-exposure transient events such as cosmic rays during image coaddition. The most common Single-exposure depth near the ecliptic is 12, while the converging scan paths yield over 200 independent exposures for the highest ecliptic latitudes in the Preliminary Release.

The finer-scale coverage features, seen in Figures 1-3, result from a number of factors relating to survey operations, delivered data quality and data processing. As described in V.2.c, individual Single-exposure framesets and frames were rejected from the Multiframe processing that produced the Preliminary Release Atlas and Catalog if they did not satisfy minimum quality criteria. Furthermore, individual frames, and individual pixels within frames were eliminated dynamically during the image coaddition process to suppress contamination from cosmic rays, satellite streaks, hot pixels, and non-spatially repeating scattered light from bright sources such as the Moon. This filtering results in effective depth-of-coverage in the Preliminary Release products that varies on scales ranging from individual pixels up to large fractions of an Atlas Tile, and coverage that can differ between bands.

Use the Atlas Image Depth-of-Coverage maps to determine the pixel-level depth of coverage achieved in the Preliminary Release Atlas Images and resulting measurements. The number of independent exposures that were used in the measurement of each source in the Source Catalog are reported in the w?m and w?cov columns.

The following list summarizes many of the different factors that affect the final depth-of-coverage for the Release.

• Survey Factors


• The nominal WISE survey strategy leads to increasing depth-of-coverage towards ecliptic poles.


• Moon-avoidance maneuvering to recover regions of the sky near the Ecliptic obscured by the moon produces "spokes" of deeper coverage away from the Ecliptic.


• Survey data acquisition was interrupted typically four times per day to body-point the WISE spacecraft fixed high gain antenna towards TDRSS spacecraft to downlink data. Downlink contacts were limited to when WISE was within approximately 45° of the ecliptic poles where the nominal coverage was deepest.


• The Si:As W3/W4 detectors were annealed twice per day to purge accumulated long-term latent images. The detectors were briefly to approximately 15K from their nominal 7.8K operating temperatures during the anneals. W3 and W4 Single-exposure data were unusable for approximately 2000 seconds following the anneals because of elevated and rapidly changing detector dark current, read noise and responsivity. One anneal per day was performed during a TDRSS contact but one was conducted during normal survey scanning, resulting in W3 and W4 coverage loss.


• One special set of tests of the W4 bias setting were made on 16 April 2010 during the survey. The W4 frames taken during these tests (in scans 03752a through 03761b) were not used for Multiframe processing for the Preliminary Release, although the W1-W3 frames were used.
• Image Quality Factors

  Single-exposure framesets were rejected from Multiframe processing if they exhibited degraded image quality, as measured during Scan/frame processing quality assurance assessment. Framesets rejected because of poor image quality have noise quality factors qi_fact=0 in the Single-exposures Frame Metadata table. Degraded image quality occurred for a number of reasons, including the following:

• The WISE spacecraft tracking required a short time to settle following slew maneuvers, including preparation for new scan stars, or return from TDRSS pass. This often resulted in the first few exposures within scans to be smeared.


• Accumulated angular momentum was dumping using magnetic torque rods on the WISE spacecraft. Early in the mission, the torque rods were enabled when the spacecraft approached within 45° of the ecliptic poles. Images were often slightly smeared because of tracking jitter following the torque rods were enabled, so there was an accumulation of poor quality exposures near the same location of the sky. This caused the streaks of low coverage seen near the ecliptic poles in Figures 1-3. Torque rod enabled was staggered between 45°, 57.5° and 70° on alternate orbits starting on 2010 May 2.


• When the one of the two WISE star trackers viewed the Moon directly, tracking stability was sometimes temporarily affected, resulting in a few framesets with smeared images. The star tracker boresites are offset by 45° in elevation from that of the telescope, so this effect produced minor but systematic coverage loss around ±45° of ecliptic latitude.


• Severe source confusion contaminated the image quality estimates for exposures taken near the Galactic center in the first-pass processing pipeline. This resulted in a number of framesets being artificially excluded from Multiframe processing in these regions. The effect of this can be seen as the swath of lower coverage region in Figures 1, 2 and 3 that follow the Galactic plane near the Galactic center.
• Image Noise and Pixel Outlier Rejection

Single-exposure framesets were rejected from Multiframe processing if they exhibited elevated noise levels, as assessed during Scan/frame processing quality assurance. In addition, individual frames and individual pixels within frames were rejected during dynamic outlier rejection in the image coaddition process.

• Severe detector electronic features could result in elevated image noise metrics that would result in the rejection of a frameset from processing. Framesets rejected for elevated noise have noise quality factors qn_fact=0 in the Single-exposures Frame Metadata table.


• The presence of very bright satellite trails contaminating a exposure could result in elevated image noise.


• Frames affected by scattered moonlight could be rejected either because of elevated noise levels, or during dynamic frame rejection during the Multiframe pipeline coadd process.


• Dynamic outlier pixel rejection in the Multiframe coaddition process masked pixels in Single-exposure frames that were inconsistent with majority of pixels from other frames covering that position on the sky.
• Source Detection and Catalog Selection Factors


• Very bright stars effectively mask nearby fainter sources because of their extended scattered light halos, diffraction spikes, latent images, glints, etc. Detection of faint sources near brighter ones is suppressed by the elevated signal levels from the bright source. When fainter stars are detected near bright sources, their measurement is often contaminated by the image artifacts produced by the bright source.


• The elevated background levels from the Zodiacal emission near the Ecliptic plan, and the complex diffuse structure near the Galactic plane raises the effective noise level in images and therefore lowers the source detection flux thresholds.


• The Preliminary Release source selection criteria required an object to fall >50" from a Tile edge to be included in the Source Catalog. This results in a lower effective Catalog coverage near the outer boundaries of the Preliminary Release areas.

4. There are regions in the nominal Preliminary Release area for which there is very low or no effective coverage

Because of the many factors that impact the usable depth-of-coverage, there are region within the footprints of the 10,464 Atlas Tiles in Preliminary Release for which the effective depth-of-coverage is very low or even zero (e.g. Figures 16 and 17 in VI.2.c). If there is insufficient coverage in these regions, the corresponding pixels in the Atlas Images may have "NaN" values, and no sources can be detected in those locations.

Consequently, even if a position or object is within the nominal Release area, it may not have sufficient coverage to have corresponding Atlas Image or Source Catalog data.

Viewing the Atlas Depth-of-Coverage maps is the best way to determine if a region of interest to you may have reduced coverage, and therefore lower sensitivity, and/or missing detections in the Source Catalog.

Last update: 2011 April 18