Hunting Voorwerpjes from Arizona

We have a team working at Kitt Peak again, this time using a spectrograph to chase down Voorwerpjes. As the Dutch diminutive indicates, these are like Hanny’s Voorwerp, only smaller. They are clouds of gas within galaxies (or out to their edges) which are ionized by a luminous active galactic nucleus. In most of these, unlike Hanny’s Voorwerp, we can see other signs of the active nucleus, but the same considerations of hidden versus faded are important. Zooites have given us a rich new list of potential objects, many from the special object hunt set up by Waveney incorporating database queries done by laihro, and more from reports on the Forum. They often show up as oddly-shaped blue zones on the SDSS images, when strong [O III] emission lies in the SDSS g filter. At some redshifts, they look purple, when Hα enters the i filter.

I’m also working with four summer students from the SARA consortium at our 0.9m telescope, normally operated remotely but this time hands-on. (Last night was the first time I’ve ever operated its camera while in the same building as the telescope). One of these students , Drew, is spending the summer working on Voorwerpjes, and is also working on the spectra. Our first night here was devoted just to training at the SARA telescope.

Kitt Peak vista

Last night we started at the Kitt Peak 2.1m telescope with a long-slit spectrograph known as GoldCam (for its color). For each galaxy, we’ve used whatever previous data were available – SARA images, processed SDSS data, a few observations by other people – to work out the most informative direction to align the spectrograph slit, which then delivers data all along that line on the sky. To set the orientation, we physically rotate the spectrograph on the back of the telescope, taking care not to snag any of the cables. This made it interesting when there was a failure of the hydraulic platform we usually use to get to the spectrograph – it’s been ladders and flashlights to do this tonight. We have a luxurious span of 7 nights (although they are practically the shortest of the year), so we can plan a pretty extensive study. We needed to concentrate on one spectral region for best sensitivity and spectral resolution, so we are using the blue range (3400-5400 Angstroms). For the redshifts of these galaxies, that lets us measure the strong [O III] emission lines and look for the highly-ionized species He II and [Ne V]. These two species are signposts that the gas is irradiated by the UV- and X-ray-rich spectrum of a quasar or Seyfert nucleus, not a star-forming region. Our first task is to conform that this is the case for many of our candidates. Beyond that, the ratios of the emission lines tell us how dense the gas is in each region, and how strong the ionizing ultraviolet is. That, in turn, suggests whether the nucleus has remained at about the same luminosity over the timespan that its light took to reach these clouds, and whether it is hidden from our view by dense absorbing material. The most exciting cases may the the galaxies that seem to have [O III] clouds but no optical AGN; they could be additional examples of the kind of rapid fading from an active nucleus that we believe went on with IC 2497 and Hanny’s Voorwerp.

One of the galaxies I most wanted to see spectra from is UGC 7342, among the greatest hits of the forum and Voorwerpje hunt. It has roughly symmetric regions of highly ionized gas reaching 45,000 light-years from the nucleus on each side, which the images suggested were probably ionization cones. These are the result of radiation escaping the nucleus only in two conical regions on opposite sides (around a thick obscuring disk). This phenomenon is seen in some other type 2 Seyfert galaxies, and if the cones are pointed in another direction, we don’t expect to see deep into the nucleus directly. These pictures were done with the SARA telescope (as I sat in my den with the cats). From left to right, they show [O III] emission, Hα emission, and the starlight alone as seen in a red filter (which has been removed from the emission-line images).

UGC 7342 emission-line clouds: O++, H-alpha, and starlight alone.

We aligned the slit with the long axis of the emitting gas (and just about across the bright star at lower left). This paid off spectacularly. Here’s a first look at a 45-minute exposure from earlier tonight. Wavelength increases from left to right, and the bright streak across the bottom is the foreground star’s spectrum. At the wavelets of such lines as Hβ and [O III], the gas glows across a huge region around the galaxy (extending vertically in this display), lit up by an active nucleus which is partially hidden from our viewpoint.

Two-dimensional spectrum of UGC 7342 showing very extensive ionized gas.

This is a chance to mention how we (truth in advertising, mostly Drew over the last couple of weeks) have been using the SDSS images to narrow down the most likely candidates for [O III] clouds and get their exact locations for the spectrum. For objects with very strong emission lines in only one or two SDSS filters, we can use one of the other filters as a guess for what the starlight of the galaxy would look like in one of the emission-line filters. We subtract various amounts of this estimate from the filters with [O III] and Hα, and select the one that isolates the clouds best. It’s not perfect, since stars in different parts of galaxies may have different average colors, but does a pretty good job as a screening tool for these active galaxies.

We want to look not only at the best candidates, but a representative set of all kinds that have turned up. This includes “purple haze” a fairly shapeless glow combining the colors of [O III] and Hα, which we see almost solely around the brilliant nuclei of type 1 Seyfert galaxies. This may be what an ionization cone looks like when we look down its axis.

We’re coordinating what we do with three nights coming up in July using a double spectrograph at the 3-meter Shane telescope of Lick Observatory, being carried out by Vardha Bennert. The telescope is larger and the instrument can get good resolution in blue and red simultaneously, so that it makes sense for us to treat some of what we do now as a screening study which can be followed up next month. (Vardha checked a couple of our candidates during a slow part of the night last December, and confirmed a purple-haze object as genuinely large emission-line clouds. This allayed my concern that these might be artifacts of incomplete registration of the three SDSS filters going into color images).

We’re still going, with six more nights and an encouraging weather forecast…