WiseView

This week, I’m excited to present this guest post from two of our users, Dan Caselden and Paul Westin.  They wrote their own tool for viewing the WISE data, called “WiseView”.  It provides some useful options you won’t find at the backyardworlds.org site.  Enjoy!

Marc Kuchner


As you may know, the images at ByW: P9 all ultimately come from a database called unWISE, which is a project that reprocesses WISE single exposures to generate coadded images with improved clarity. Since we citizen scientists with ByW: P9 are always eager to know more about our subjects, we found ourselves often visiting the unWISE site to obtain different views of our favorite patches of sky.

However, we felt that unWISE’s packaging could stand to be a little more user friendly. So, after a while, we decided to add a friendly wrapper, to make this data easier to examine, and share it with you. We’re a far cry from User Interface/User Experience professionals, but, hey, it’s a start!

image2

Our tool, wiseview (http://byw.tools/wiseview), displays two sets of cutouts (i.e., portions of larger images of the sky). At the top, wiseview flashes coadded imagery from the WISE satellite. These cutouts come from unWISE.  unWISE currently contains coadded images for three data sets: AllWISE, NeoWISE-R1, and NeoWISE-R2. unWISE coadds are full-depth. That is, unWISE NEO1 also incorporates the single exposures used by unWISE AllWISE, and unWISE NEO2 also incorporates the single exposures used by unWISE NEO1 and unWISE AllWISE. Consequently, particularly high proper motion options will appear to stretch, or even fade in one position and appear in another.

Since ByW: P9 participants are on the hunt for things that move in WISE data, unWISE images are a natural resource for further investigation. After identifying coordinates of a pattern possibly indicative of proper motion, participants can zoom in with wiseview to see a closer representation of the underlying data from the flipbooks. The “field of view” parameter selects what size cutouts to display, in arcseconds, and the zoom slider blows up the unWISE cutouts. WISE W1 and W2 bands can be isolated with the WISE band field (W1 for W1, W2 for W2, and W1+W2 for both), and the “Speed” slider changes how quickly the cutouts flash.

The second cutout is a composite image from PanSTARRS-1, created in the same way as the default PanSTARRS-1 cutouts: band y colors red, band i colors green, and band g colors blue. PanSTARRS-1 cutouts are great for comparison versus unWISE because many unwanted sources and some of the brighter and/or earlier brown dwarfs show distinguishably.

unWISE Post-Processing

unWISE cutouts are normalized with astropy.visualization.AsinhStretch, and mapped to a colormap with matplotlib. The following images show AsinhStretch applied to a greyscale gradient with differing values of ‘a’. The ‘Linear’ parameter in wiseview is directly passed through to this parameter in AsinhStretch. ‘Linear=1.0’ applies a purely linear normalization to the image, which has no effect.

Images3-5

Lower values highlight lower intensity pixels, which is useful for observing faint sources, or those obscured by other, brighter, sources. For example, The images below show Ross 458C  with varying values of ‘Linear’.

Images6-8

However, purely AsinhStretch normalization can make modest proper motions difficult to discern. Observationally, the normalization appears to lose dynamic range at the edges of sources, which is where the eye seems to most perceive motion in these images.

The three modes, ‘fixed’, ‘percent’, and ‘adapt’, attempt to compensate for this by capping intensity ranges before AsinhStretch normalization. ‘fixed’ caps the maximum intensity to an absolute number supplied by the slider ‘Trim Bright’. ‘percent’ caps the maximum intensity to a percentile within the image, again using the slider ‘Trim Bright’. ‘adapt’ is very much a work in progress that (poorly!) attempts to find a good intensity range automatically.

Why wiseview?

We wrote wiseview to improve our accuracy (and sate our curiosity!) when classifying candidates in the ByW: P9 flipbooks. With wiseview, curious participants can investigate their subjects to show whether their candidates demonstrate proper motion. For particularly challenging candidates that are not easily distinguishable in other available imagery like 2MASS, comparing unWISE coadds can be our only option to demonstrate proper motion.

Although we originally wrote wiseview for use with ByW: P9, its applications are more general; anyone searching near-infrared for objects in the solar neighborhood may find it helpful. In fact, multiple ByW: P9 participants discovered candidates in side projects using wiseview.

By the way, if you’re interested in unWISE, another great resource is legacysurvey.org’s Sky Viewer. For getting a quick big picture of what’s going on in a portion of the sky and disambiguating sources, their tool is invaluable. It also provides other image sets and catalog overlays, like DECaLS and SDSS, if your coordinates are lucky enough to fall within those surveys. Very nice!

Thank you!

We are Dan Caselden and Paul Westin, two computer security researchers from California with absolutely zero background in Astronomy. Thanks for reading!

 

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s