What we’re doing with PS1

The “Pan” in Pan-STARRS stands for panoramic, but roughly a quarter of the telescope’s time is spent staring at the same ten points in the sky over and over again. These may make up under a quarter of one percent of area of the map of the sky Pan-STARRS1 is making, but they have paid big scientific dividends and have the prospect to produce even more interesting research.

The sky is big, really, really big, it’s the size of almost 200,000 full moons. Pan-STARRS1 is spending about half its time mapping three quarters of this multiple times in different colours of light. However some science goals need more regular observations of patches of the sky. One such goal is looking for exploding stars known as supernovae. If you regularly image a patch of the sky you won’t see exactly the same image over and over again. Sometimes you’ll notice that stars are changing in brightness. These can be because they are pulsating stars which vary slightly in brightness. However there are some stars which are so high in mass that they burn out their fuel quickly and then explode as supernovae. By observing the same part of the sky over and over again, the rapid increase in brightness of supernovae in distant galaxies can be observed. This allows the detection of new supernovae with Pan-STARRS1 which can then be followed up with other telescopes. Have a look at the movie below which shows a supernova detected by Pan-STARRS1.

A short movie of a supernova found in one of the Medium Deep Fields of Pan-STARRS1. Notice there is nothing in the middle at the start but gradually the supernova appears and becomes one of the brightest stars in the field before beginning to fade. Credit: Michael Wood-Vasey/Alisa Rachubo/Alex Damewood/University of Pittsburgh/PS1SC

So far Pan-STARRS1 has detected thousands of probable supernovae, many of them a bit strange. Matt McCrum in Belfast has been looking at supernovae which do not have a clear host galaxy. Laura Chomiuk also wrote a post about some especially strange, extremely bright supernovae. Single exploding stars are not the only sources of supernovae. Binary stars can sometimes be so close that one star will rip matter off the other and suck it on to its own surface. If the star receiving the matter is a white dwarf, there is a limit to how much matter can be dumped on it. Once the white dwarf reaches about 1.4 times the mass of the Sun the physical mechanism that supports the white dwarf against collapse is overpowered. The white dwarf then suffers an cataclysmic collapse which also generates a massive explosion. Because this happens at a specific mass, the brightness of these type of supernovae is approximately the same. Hence these can be used as standard candles to measure distances in the universe. Find this type of supernova and you measure how fast it appears to be receding from us and you can measure how fast the universe is expanding at a particular distance. Build up a large sample and you can work out if the universe is accelerating or decelerating. This is was used in the 1990s to determine that the universe was in-fact accelerating. Pan-STARRS1 is currently building up a sample of this type of supernovae to constrain the expansion history of the universe even more.

An example of one of the deep, stacked image from one of the Medium Deep Fields. It shows NGC 7684 (lower centre) along with other galaxies. Credit Nigel Metcalfe/PS1SC

Staring at the same place over and over also means you get more and more photons. You can add these up by “stacking” images. While this isn’t literally stacking photographic plates on top of each other, it produces a similar effect, detecting fainter objects than you would from just a single image. Do this with years worth of images over the ten repeatedly surveyed fields (known as Medium Deep Fields) and you get wonderfully deep, detailed images. These fields aren’t chosen at random, they tend to point away from the galaxy so the number of foreground stars is low. Many other surveys have taken data in infrared light as well as the X-ray and radio. By picking areas also covered by these surveys for their deep, repeatedly surveyed fields, Pan-STARRS1 scientists can use these other datasets for studying galaxies in far-flung corners of the universe.

So yes, these fields may be small by Pan-STARRS standards, but the repeated sampling makes them fantastic for characterising parts of the universe that are far away.

Charting the heavens has always been one of the prime motivations of astronomy. However once larger telescopes began to be developed in the early twentieth century, the need for large continuous surveys of the sky to provide targets for these instruments became apparent. This need was filled by the telescopes invented by Bernhard Schmidt, an Estonian optician. Always an experimenter, Schmidt had lost his right-hand while playing with explosives as a child. His telescope combined lenses and mirrors to give sharp images over a wide field of view. Telescopes based on this design in California, Australia and Chile surveyed the whole sky over the course of fifty years. These surveys covered the sky multiple times in many different coloured filters. The frequent surveying of the sky led to the first classification system for exploding stars called supernovae and the discovery of thousands of nearby stars moving slowly across the sky.

The PS1 Observatory on Haleakala. Photo Credit: Rob Ratkowski, PS1SC

This frequent surveying of the sky is one of the primary aims of the Pan-STARRS1 3Pi survey which gets it’s name from the area it covers. A steradian is a measurement of angular area and full sky is four pi steradians, but almost a quarter of that is not visible from Haleakala where PS1 is based. Hence by mapping almost all the sky over Maui, Pan-STARRS1 will cover three quarters of the sky, three pi steradians. The survey will be done in five colour filters named g, r, i, z and y. Three of these g (roughly green light), r (red light) and i (almost into the infrared) cover wavelengths visible to the human eye. The z and y filters (which like many things in astronomy, have names which don’t make much sense) cover far-red light mostly invisible to humans. Each of these are planned to be observed on up to six separate nights over three years. This mapping will be done forty times faster than the photographic plate based Schmidt telescopes.

Each of the six nights of observation per filter will consist of two images, half an hour apart. This interval will allow asteroids (including some that pass close to the Earth) which move quickly across the sky to be identified. In between nights more distant icy bodies can be identified by their celestial motions. Finally over a timescale of months and years nearby faint stars can be found, again by their movements on the sky. It’s not just varying positions which can be found by the survey, objects can change brightness too. Young stars sucking matter from disks of material around them can have leaps in brightness. Exploding stars can also be detected by their sudden changes in luminosity

Finally all the observations from the PS1 survey will be combined to make one deep multi-colour image of three quarters of the sky. This can be used to probe deeper both within and outside our Galaxy. It can peer into the depths looking at how galaxies are distributed in space, probing the mysterious dark matter which makes up nearly a quarter of our universe. It is also planned to use these images to look for subtle warping in the images of galaxies which are imprinted with clues about the strange dark energy which appears to drive the expansion of the universe.

You would think all this would be enough for one telescope, but PS1 is also undertaking a series of smaller surveys. Watch this space for more details.

Stars, such as our Sun, may appear to be ever-present sources of heat and light, but they are highly dynamic objects which will eventually reach the end of their evolutionary paths and die. Although the death of the Sun will be comparatively quiet, more massive stars (with masses greater than 8 times that of the Sun) can be ripped apart when their iron core collapses to an extremely dense stellar remnant known as a neutron star and a shock wave driven by the production of a huge number of particles called neutrinos blasts away the outer layers of material in a supernova explosion. This ejected material does not go to waste however and the elements created, such as carbon and oxygen, provide the Universe with the building blocks it needs for the formation of rocky planets like Earth and all living things upon it.

The process outlined above will lead to only one of the different types of known supernovae, called core-collapse supernovae. These core-collapse type events can then be further divided into sub-classes (type Ib/c, type II) depending on the mass of the progenitor star which explodes. One of the most luminous type II explosions was recorded at 100 billion times the brightness of the Sun and could easily be identified despite being 4.7 billion lightyears away. A second type, a thermonuclear supernova (type Ia), is thought to occur when a dead remnant of a star like the Sun known as a white dwarf star which happens to be in a close binary system accretes matter from its companion, pushing it past its upper mass limit (the Chandrasekhar limit) of 1.4 solar masses. In the last two years incredibly bright supernovae (100 times brighter than an average SNe) have been discovered which may require a new physical explanation for their luminosity and observations from the Pan-STARRS 1 survey (PS1) will help further our understanding of these ultra-luminous events and better define the very interesting physics which leads to them.

Pre-supernova (top) and post-supernova (bottom) PS1 images showing the discoveries of a supernova with a clear host (left) and a hostless supernova (an orphan, right).

The basic discovery and classification of supernovae is a two-part process. Difference imaging, where nightly images are subtracted from a previously obtained reference image to highlight any objects with a variable flux, is used to detect these very characteristic transients. This is essentially an interstellar spot-the-difference and some examples of PS1 images containing supernovae can be seen in the images to the left. After the initial discovery of a transient that is thought to be a supernovae, follow up spectroscopy must be done in order to confirm the nature of the variable object (it might be simply a variable star or an active galactic nucleus). If it is a supernovae, the spectra can be used to determine elements involved in the explosion and thus deduce its type. By compiling large datasets of confirmed events, the rates of different species of supernovae can be estimated giving us invaluable information on stellar populations within various forms of galaxies and the intergalactic medium.

Up until now, both professional and amateur searches for supernovae have concentrated their efforts on large galaxies with high star formation rates to ensure a large number of successful discoveries. The PS1 survey is not a dedicated search for transients but hosts a multitude of varied science projects meaning that the Hawai’i based, 1.8m telescope, with its 7 square degree field of view focused onto the world’s largest CCD (1.4 gigapixels), surveys massive, unbiased portions of the sky every night. This results in supernovae discoveries in unexpected places as demonstrated in the right hand set of images above. This type of transient is known as an Orphan due to its apparent lack of association with any other object. The object may very well have a host galaxy, but one that is simply too faint to be imaged with the PS1 survey suggesting that it is either at a high redshift or has a low proportion of elements heavier than Helium. If the galaxy has a high redshift (and thus is at a great distance from us) the high luminosity of supernovae events allow us to probe deeper into the universe than the photometric limit of a telescope may at first suggest. If the event occurs within a faint host galaxy, the type of supernovae can give us evidence as to the population in these mysterious stellar neighbourhoods. A possible progenitor channel for the ultra-luminous supernovae mentioned earlier may be pair instability in stars which are quite deficient in the heavy chemical elements (like carbon and all heavier elements). Previous searches in relatively nearby, metal rich galaxies have therefore missed these exiting and currently obscure phenomena.

The lightcurve of a hostless SNIIn from PS1 photometry. The different coloured points represent brightness measurements in different filters (g, r, i and z).

Throughout the 3 year period of my PhD at Queen’s University Belfast, under the supervision of Professor Stephen Smartt and Dr. Rubina Kotak, I am hoping to classify as many of the orphans from the PS1 survey as possible, offering statistical analysis and comparison of their rates with current estimates. The PS1 telescope has an r-band photometric limit deeper than magnitude 21 (over one million times fainter than what the human eye can see on a dark night) which is easily capable of detecting these events and producing thorough and defined lightcurves, such as the one shown on the right. Of the almost 250 supernovae with faint or no visible hosts already compiled, at least 3 are thought to fall into the ’ultra-luminous’ category and with hundreds of new transients being detected by PS1 each week, more are sure to follow. With a little luck I will be able to obtain decisive experimental data and perform intense follow-up observations of some of these objects in an attempt to further our understanding of this latest addition to the study of supernovae. Another group, the Palomar Transient Factory, is also looking for these fascinating events (more information on their project can be found here). Possible collaborative work with this team, and some constructive rivalry, will ensure that the science goals presented here are carried out to optimal completeness.

The vicinity of the Solar system, the nearby few hundred light-years, has been a prime source of astronomical discoveries for many centuries. The average distance to stars that are similar to the Sun and visible to the naked eye is less than 120 light-years. This is the very neighborhood of the Sun when you think that the center of our Galaxy is located some 25,000 light-years away! Those are the stars that made up the “sphere of the fixed”, past the Moon, against which astrologers and early astronomers such as Kepler measured the movement of celestial bodies of the Solar system and eventually, let Galileo and Newton discover the law of gravity.

When the invention of the telescope in the 17th century allowed astronomers to observer fainter stars and accurately measure their position, it appeared that the “fixed stars” were not that fixed. Many showed small motions compared to other stars over the years, and some additionally show a circular, yearly motion, reflecting the revolution of the Earth around the Sun. These motions, which are greatest for the stars closest to the Earth, allowed scholars to discover the Sun’s neighbors and to measure their distance.

In the 20th century, the observations of the sky with photographic plates led to the discovery of most nearby stars more massive than a few tenths of the Sun’s mass. Today, Pan-STARRS 1 will use exactly the same methods, with greater accuracy and sensitivity, to detect all nearby stars in the Northern sky and measure their distance. The greater accuracy will allow us to expand the “Solar neighborhood” to about 250 light-years, and the greater sensitivity to detect the coolest stars, as well as compact objects not massive enough to burn hydrogen as stars do, which are called brown dwarfs.

One key project of the Pan-STARRS 1 Science Consortium is dedicated to the study of our stellar backyard. Astronomers from across the PS1SC including Eugene Magnier and Michael Liu from the University of Hawaii and I have been working to take a complete census of the nearby low-mass objects, both stars and brown dwarfs. The detection of all very-low mass stars and brown dwarfs in the Solar neighborhood may reveal the closest Solar neighbor, possibly closer than the current holder of that record, Proxima Centauri at 4 light-years. Because they are close, those objects are the brightest among their peers, so that it is possible to observe and characterize them with greater details.

2MASS J09201223+3517429 imaged by Pan-STARRS 1, indicated by an arrow. While it appears as a single red dot, it is actually a binary pair of brown dwarfs. Pan-STARRS 1 will discover hundreds of new brown dwarfs. Credit: PS1SC

Like stars, brown dwarfs form from the collapse of a huge cloud of gas. However unlike stars they cool and dim with time, because they lack an internal energy source. This makes it difficult to determine important parameters like age or mass, because a young, low-mass brown dwarf will look like an older, more massive one. Sometimes we can learn more from a stellar companion, or from the stellar cluster if the brown dwarf belongs to one.

The atmospheres of very low-mass stars and brown dwarfs have temperatures of a few hundred degrees to a few thousands Fahrenheit. Because of such a large range of temperatures, the atmospheric chemistry is diverse and complex. We know that dust clouds are present, but sometimes hidden below a layer of water and methane. The coolest known dwarfs have ammonia; even cooler dwarfs may have water ices. It seems that different dust properties change the colors of brown dwarfs, as well as the proportion of heavier elements they have, or ages. Only when we discover and characterize more objects will we understand better their properties. This will also help us to understand the atmospheres of gas giant planets around other stars. But these cold brown dwarfs won’t be the only near neighbors of the Sun we will search for with PS1.

Most of the stars in the Solar neighborhood belong to a flattened structure of our Galaxy called the Thin Disk, which we see as the Milky Way on the sky. That disk sits in a dimmer and older structure, almost spheroidal, called the Halo, which extends to great distances. A small fraction of the nearby objects are actually members of that Halo. Detecting and studying them, for instance old white dwarfs, which are the dead remnants of aged stars which were originally similar to the Sun, provides information on the Halo, such as its age and origin.

On the contrary, another small fraction of the Solar neighborhood is made of young objects which just escaped their stellar nursery, or maybe were born in isolation. Again studying them with great precision allows us to understand the formation of very low-mass stars and brown dwarfs, which may differ from both higher-mass stars, and lower-mass extrasolar planets. We can also search for young planets and gas and dust disks orbiting the young dwarfs. With the present instrument those are visible only close to the Earth.

Finally, thanks to our repeated observations of the whole Northern sky, we will discover most variable stars having variations as small as 1% in their light output. These stars may be young stars, still erratically accreting material from disks of material left over from their formation; or eclipsing stars, whose light is occasionally blocked by a companion or a planet; or magnetically active stars which show variations due to massive flares in their atmosphere.

The Solar neighborhood is a diverse mixture of stellar populations reflecting the past and present of the Galaxy. As we come to describe it more completely, we learn about the life of stars from birth to death, or their companions and planets, and of the physics that control their atmospheres and their evolutions. Pan-STARRS 1 will be a key tool in these investigations over the coming years.