2018 June

Project EDEN’s goal is to find potentially habitable planets around stars in our celestial neighborhood. We will share our findings with scientists around the world, so they can further study, explore, and possibly visit them one day. My job is to find nearby stars that work well with our detection level and that are most likely to host habitable planets. To create a manageable list of stars, I am using many different criteria to filter out some of these targets and narrow the scope of the survey. With fewer stars to review, we can observe each target longer and have a greater chance of discovering an exoplanet.

Venus passing in front of, or transiting, the Sun. Image from NASA.

First, I will explain Project EDEN’s primary method of detecting exoplanets. Our solar system lies approximately in one plane, so if we were to stand on Mars and look directly at the Sun, we would eventually see Earth pass in front of, or transit, the Sun, because it is crossing through the area directly between our eyes and the Sun. Instead of standing on Mars, however, Project EDEN is using telescopes located on Earth, and instead of staring at the Sun, we’re pointing the telescopes at other stars in the Milky Way. If the angle of the plane of planets in that solar system sits within the area formed between the telescopes and the star, we’ll be able to detect planets passing in front of the star. This method of exoplanet detection is called the transit method.

Using the transit method to detect planets around all the stars in our solar neighborhood would take decades because planets can take anywhere from a few hours to a few years to cross in front of a star. You have to point a telescope at a star for slightly longer than the entire transit to detect that planet in order to create a graph of the light of the star before, during, and after the transit. This is called a light curve.

Actual data of planet WASP-10b transiting in front of star WASP-10. Visual from Institute for Astronomy, University of Hawaii.

If we see a little dip in the data, we know that a transit occurred for the duration of the dip since the planet blocks that star’s light as it passes in front of it. If the transit is long, we have to spend a long time looking at the star without any breaks. However, the smaller a star is and the closer a planet is to that star, the shorter the transit. The more time you spend observing a star, the higher probability you have of witnessing a transit. However, it’s very expensive to purchase a large amount of time on the types of telescopes required for our research. That’s one of a few reasons why Project EDEN is beginning observations with a short list of 40 cool, small stars. I wrote a program to extract star data from the astronomical database Simbad, and dump it into a localized database.

Initially, we extracted the data of approximately 100 stars. Then, I removed known double or multiple-star systems, which make transit observations difficult. Imagine that the double stars are like two bright light bulbs less than an inch apart. If we pass a tiny object in front of them (our exoplanet), it would be difficult to distinguish how much light one of the bulbs emits because you’re also receiving light from the other bulb. Finally, I reduced the database to 40 stars by choosing the 20 closest stars in each hemisphere since EDEN is focused on nearby exoplanets.

Gaia’s most recent all-sky map, colorized.

Going forward, I plan to collect more information on each of the 40 targets, such as mass and radius. Additionally, the second Gaia Data Release will be a great resource for us. Gaia is attempting to create a 3D map of the Milky Way, so it may provide us with information like distance that many stars in Simbad do not already have. 

Working on Project EDEN as a first-year undergraduate student has been a steep, but extremely interesting and rewarding learning curve. I had to greatly expand my knowledge of the programming language Python, as well as observational astronomy concepts. Even though I’ve been part of EDEN for more than six months now, I still have a hard time believing that I’m actually playing such an important role in the search to find life on other planets. I am excited to begin observing the stars selected in the first project phase. I look forward to continuing my work as EDEN grows larger and more developed as we move into the future.

As our hemisphere of this big rock slowly moves into summer, everything focuses on change. The local flora and fauna prepare themselves to weather the harsh summer rays, much better than a certain undergraduate researcher who has lived his whole life underneath the swaddling blanket of Oregon’s clouds. There is an air about the EDEN group that something is about to happen. Which would make sense. Over the past 6 months the group has gone on a handful of observing runs for over a month of collected observable nights. During these observing runs every single undergraduate researcher has undergone at least their first overnight observing run. We have developed advanced systems to process photometric data from our observations. We have gained experience navigating through scientific papers and other astronomical publications to create a detailed database of many potential targets perfectly ripe for observing. We have also grown our international family with the addition of the Max Plank Institute for Astronomy as an institutional partner, joining Steward Observatory and the Vatican Observatory The entire EDEN team has grown in experience on how a survey and more importantly a team works in harmony.

Some personal development is beginning work on EOPAST, or the EDEN Observation Planning and Analysis Strategy Tool. The purpose of this tool will be to query the databases that Jose and Allie (of the Reduction and Targets teams respectively) are working on to provide a myriad of easily accessible information on potential targets for an observing run. To develop EOPAST, I have begun taking online python coding classes to better understand how to obtain and implement code. My plan is to hopefully also include many functionalities in this tool. These could include providing guide stars for targets, available times during a night a target is observable, and the ability to write observing scripts for the telescopes that EDEN frequently uses.

Here is some background information on what some of these terms and things actually mean. EDEN is focused on observing some of the dimmest and faintest stars in our galaxy. These are much dimmer than what humans can see, and provide even a challenge to telescopes to directly focus on. Even small variations from the intended target can be quite detrimental. These telescopes however have a comparatively wide view of a particular portion of the night sky. So, to go around this problem, astronomers instead focus their telescopes on a much brighter star that is in the celestial neighborhood. By focusing the telescope on that guide star, combined with the larger field of view, whenever the telescope takes an exposure the target star is somewhere in the image. Focusing on the brighter target also leaves less room for the tracking system to vary as it moves through the night sky.

Target visibility is a slightly more difficult concept to grasp, but ultimately it’s not that bad. One major component is a concept in astronomy called ‘air mass’ which is basically how much air light from a star must travel through. Objects with a high air mass suffer from more distortions due to the atmosphere, with a perfect air mass level being 1.00. Air mass depends on a number of factors, which include the local ‘sidereal’ time (or just how far away the local time is from the Universal ‘UTC’ timezone), the elevation of the observatory, and their ‘declination’ (which is similar to celestial latitude). Stars that are higher in the sky have a higher air mass generally, with a star directly above you having an air mass of 1.00. You can picture this in the image to the right. The stars that make the small circles have lower air mass, while the stars that make large circles have a higher air mass. Every star that is observable during any particular season has an ‘air mass curve’ as the star moves in the sky over the night.  At the peak of this curve is how close the star gets to the minimum of 1.00.Using the air mass curve, we can then plan when we observe a target and for how long we can observe it. For reference, we can get really good data whenever we observe a target when it is below 1.50 air mass. So, we look at the air mass curve of the target, and only observe it when it is high enough in the sky. Every night of the year this curve shifts for every star, some more than others. Whenever the air mass curve of a star shifts so far that it peaks during the day, we say that the star is ‘out of season.’ We can then plan observations months in advance, picking targets that are in season and have a low enough air mass.

The last concept I want to explain is the process of writing observing scripts. Many modern telescopes are moving to automate many of the processes that are associated with observing. One of the telescopes that our survey uses, the Schulman telescope, has the ability to read observing instructions and follow a plan over the course of a night. This script includes when to open the telescope up, what the coordinates of the guide star are, what angle the telescope must rotate itself, and how many exposures to take and for how long. The cherry on top is that the telesccope will shut down and close the dome all by itself at 4 in the morning, which means we don’t have to stay up until then.

I am also assisting one of our new team members, Luigi Mancini of the Max Planck Institute of Astronomy, to plan observations for late June/ July to utilize the Calar Alto 1.23m telescope of Spain and the Cassini 1.5m telescope of Italy. Using all of the concepts and techniques I just described, there is an interesting phenomenon that we are using to make these observations as powerful as they can be. Due to the large time difference between Southern Europe and Arizona, the telescopes in Europe can begin observing a target in a night, and by the time it reaches a high air mass or the sun begins to come up, our telescopes in Arizona can begin their night and pick up where they left off. All told, this would mean that it would be possible to observe a target continuously for almost 15 hours! This dramatically increases our chances of detecting an exoplanet.

I hope that you now have a better understanding of some of the aspects that go into planning observations. I think this stuff is incredibly cool, and I’m excited for what the future holds for this survey. We are finally poised to move into the next stage of our observations, which is to observe stars looking for previously-unknown exoplanets. This is an incredible time to be alive. These stars are out there, and we are gazing upon an ocean of limitless mystery and wonder.

Image credits: Featured Image, Shah of onebiguniverse-blog. Image 1: davesastrotools.com. Image 2: Videohive.net, Image 3: nikonusa.com

Earlier this week University of Arizona student Allie Mousseau and graduate student Ben Rackham, members of the EDEN team, welcomed a group of 12 students at the UA SkyCenter atop Mt. Lemmon and introduced them to the methods we are using to find exoplanets. The activity was a collaboration between our EDEN team and Project POEM, which aims to foster interest in STEM careers among visually impaired middle and high school students. Since the Project POEM students have different levels of visual impairment, the traditional data visualization tools were not suitable for demonstrating the concepts to the students. We wanted to find a way that allowed them to get an intuitive sense of the exoplanet transit data – to bring the abstract concept closer to the students’ everyday experiences. Steward Observatory graduate students Alex Bixel and Ben Rackham and University of Arizona professor Daniel Apai came up with an innovative method to demonstrate exoplanet transits: by translating stellar lightcurves (brightness variations) into piano tunes, all students could easily pick up the events when planets pass in front of their host stars – the sort of transits Project EDEN is searching for.

The students heard examples of models of exoplanet transits to learn about how we can tell the size and orbital period of an exoplanet from its light curve.  Here, for example, is a visual comparison of the transit depths for three planet-sun combinations:  a Jupiter-like planet orbiting a Sun-like star, an Earth-like planet orbiting a Sun-like star, and an Earth-like planet orbiting an M-dwarf star:

And here are how those same model lightcurves sound when converted to pitches:

Jupiter-Sun:

Earth-Sun:

Earth-M dwarf:

As you can hear, it’s much easier to detect an Earth-sized planet around a small M-dwarf star than around a Sun-like star–the transit depth is almost as deep as a Jupiter-sized planet transiting the Sun!

Of course, real measurements always have noise as well, and the students got to experience this first-hand by analyzing a transit light/soundcurve of the sub-Neptune GJ 1214b that we recently observed with the Kuiper 61″ telescope on Mt. Bigelow:

Finally, the students listened to segments of a 20-day, nearly continuous lightcurve of TRAPPIST-1 taken with the Spitzer Space Telescope.  This amazing dataset, which was published in Nature last year, contains transits of seven Earth-sized transiting exoplanets, three of which have surface temperatures that could allow for liquid water!

This artist’s concept allows us to imagine what it would be like to stand on the surface of the exoplanet TRAPPIST-1f, located in the TRAPPIST-1 system in the constellation Aquarius.

The students listened to segments of this lightcurve and worked together to identify transits, breaks for data downlinking from the telescope, stellar flares, and even a simultaneous double-transit of two planets.  They used the open-source Aria Maestosa program to change the tempo and instrument used to represent the data in order to best analyze the data. As you can hear, the similarities of the transit depths between the planets and the practice of compressing many days of data into a few tens of seconds make this no easy task!

Samples of TRAPPIST-1 Spitzer Light Curve:

It was great to see the excitement of the students as they have explored worlds beyond the solar system through “interstellar songs”!

Listening to the melodies of exoplanet transits.