Hunting Outbursting Young Stars with the Centre of Astrophysics and Planetary Sciences (HOYS-CAPS)

Dr Dirk Froebrich -- df@star.kent.ac.uk

HOYS-CAPS is a new citizen science project run by the University of Kent. It aims to engage amateur astronomers in the search for and characterisation of highly variable young stellar objects. This article briefly summarises the science goals of the project and the requirements for participation. More detailed information can be obtained from Dirk Froebrich or on the project webpage, which also showcases first results and images:

http://astro.kent.ac.uk/~df/hoyscaps/index.html

Science Goals

Understanding the mass accretion process onto the young stellar objects is one of the fundamental problems in star formation. In particular, the variation of accretion rates on timescales of several years is not well understood. Observed mass accretion rates of young stars tend to be much lower than predicted by theory. One solution for this problem is that young stars undergo rare and short, but intense bursts of increased accretion activity during which a large fraction of their total mass is accumulated. These burst can be observed as increases in luminosity by up to (or even in excess of) a factor of 100. Recent works have shown the most intense of these, the FU-Ori type bursts, are very rare and only happen every 5-50 thousand years for a particular young star. It is still not clear how long these busts last -- some known bursts have lasted for more than 100 years.

Our goal is to measure the occurrence rate of these bursts more accurately than the current order of magnitude estimate. This is important as it will allow us to accurately determine what fraction of the total mass young stars accrete during these bursts. Due to their intensity and thus vastly increased output of energetic radiation, these bursts have a severe impact on the chemistry and distribution of the material in the stellar accretion disks. Thus they are ultimately a great influence on the formation of planetary systems in these disks.

To identify these rare events we plan to photometrically monitor a number of nearby young clusters and to search for new, highly variable objects. Using clusters of young stars has the advantage that we know the distance and ages of all objects, as well as the total number of the monitored stars -- all vital for the scientific analysis. The envisaged long term monitoring will not just identify outbursting objects, but also other, rare events such as occultations of objects by their accretion discs (such as in KH 15D or V* V582 Mon) which can be used to study these discs in great detail.

Participating in the Observations

Our main aim is to find highly variable or outbursting objects, hence there are no particularly high requirements for the image quality. Nevertheless, to increase the usefulness and accuracy of the brightness measurements, all images should be subject to a bias and flatfield correction. This removes most of the systematic noise in the images and greatly improves the detectability of brightness variations, in particular for fainter objects. The project website has some detailed notes of the required data reduction procedures, and we are also able to run a workshop for small groups of interested participants should this be of interest.

There are no other stringent requirements for the images. Any image is useful and much appreciated. The website lists a number of target clusters/regions -- which will be extended should more participants join the project. Please centre your images on the provided coordinates to maximise the number of monitored stars. The field of view, filter used and integration times do not matter. Based on our experience in the first observing season, and the subsequent data analysis, the best 'strategy' seems to be if every observer concentrates his/her efforts on a small number of targets and spends a larger integration time on target to monitor fainter cluster members. Images can be taken at any time, ideally one every few weeks, but even a single image is helpful.  When you have done the data reduction of your images (dark current and flatfield corrections) you can upload your final images to our Dropbox account (details on request). We will then perform photometry of your images as soon as possible and check if there are new, highly variable stars in the image. We have started to develop software which automatically analyses the lightcurves for all stars in all images, hence the processing will eventually be done very quickly. Should any new potentially erupting objects be detected, we aim to perform follow up observations with a number of telescopes to verify their nature. Currently we will be able to quickly get access to the St. Andrews Observatory, the Liverpool Telescope, the Thuringian State Observatory in Germany and the newly constructed Beacon Observatory at the University of Kent.


Data Analysis and first Results

During the first observing season we have gathered 42 images of seven individual targets. We have analysed the six different RGB images of NGC2264 (Christmas Tree Cluster) in some detail. We attached an accurate coordinate system to every image using the astrometry.net software and then co-added all images with the Montage software package. This deep image was used to detect stars using the SourceExtractor software and photometry of all stars has subsequently been performed in each of the individual images. Using the assumption that on average all the stars in the image are not variable, the fluxes of all images are set equal and stars who are much more variable than the typical stars of the same brightness are selected. This procedure easily recovers many of the known and obvious variable stars in the images. Finally, the magnitudes of the stars are calibrated using the photometry from the Sloan Digitized Sky Survey.

In Fig.1 we show the typical variability of the stars in NGC2264 as a function of their R-band magnitude. Bright stars vary by about 0.1mag (or 10%), which is due to saturation and non-linearity in the detector which varies with the observing conditions such as the seeing and transparency of the atmosphere. For a range of magnitudes (R=13.5--15.5) the typical variability (or accuracy of the photometry) is better than 0.05mag (or 5%). The faintest objects detected in the images are about R=18mag.

In Fig.2 we show the R-band magnitudes as a function of time of six of the selected and well known highly variable objects. All of them vary by at least one magnitude and two of them completely disappear below the detection limit in at least one of the images. Hence some of the stars change their flux by almost a factor of 100.

Figure1: rms_ngc2264_r.ps
Figure2: variables.ps