Dear friends,
I have been eager to write and convey the excellent news that our observing session at the Green Bank Telescope (GBT) on Friday went extremely well. The students and I are thrilled beyond measure. Before I describe the observations, let me summarize the events of the past week.
In order to minimize the time spent repositioning the telescope, I asked the students to restrict their proposed observations to a region of the sky that includes about 30% of the celestial sphere. This area includes the Kepler field, which is a 10° × 10° region of the sky that the Kepler Space Telescope observed continuously between 2009 and 2013. Thousands of exoplanets have been discovered in this region, including planets that are potentially habitable. I encouraged each student to nominate one planet and to become primarily responsible for analyzing the corresponding observations. The students submitted 13 suggestions, including many of the most promising habitable worlds in the Kepler field. Because our observational strategy works best with an even number of sources, I picked a planet as well. One of the students (Akshay) also suggested an appropriate pulsar for that part of the sky (B2021+51).
It was important to arrange the 14 planets in a logical sequence for our GBT observations. During our class on Wednesday, I described the travelling salesman problem, a classic computer science problem that is often stated as follows: "Given a list of cities and the distances between each pair of cities, what is the shortest possible route that visits each city exactly once and returns to the origin city?'' The students wrote computer programs to solve the traveling salesman problem in the context of our observing sequence, seeking to minimize the time spent repositioning the telescope. They also learned about the important business-world implications of this problem: courier companies save millions of dollars when optimizing their routes according to such algorithms. This course module works very well because there are a variety of algorithmic approaches to solving the problem that suit every level of computer programming experience. There are simple nearest-neighbor algorithms (always select the closest city), brute force methods (try every possibility), and relatively complicated heuristic approaches (efficiently find an answer that is not perfect but good enough). Students tried at least 4 different approaches, and we settled on the optimal sequence.
|
|
|
|
The travelling salesman problem applied to U.S. capital cities (courtesy of SAS).
|
|
|
|
The ordered sequence of 14 exoplanets that we observed with the Green Bank Telescope (courtesy of Akshay).
|
|
|
|
|
|
Our final observing sequence was designed as follows: First we would spend about 30 seconds recording a test tone signal, a signal of known frequency that can be used as an end-to-end test of the receivers, frequency downconversion chain, data-taking systems, and data-processing software. Second, we would record pulsar B2021+51 for 2 minutes. Third, we would observe each one of the 14 planets twice. Finally, if there was any time remaining at the end of the 2-hour session, we would observe a GPS satellite. The hardware setup accommodated recording of the observations with two separate data-taking systems: one with a wide (800 MHz) bandwidth and coarse (2-bit) sampling, and the other with a narrower (200 MHz) bandwidth and finer (8-bit) sampling. This redundancy was meant to minimize the risk of a backend not working as expected. My main concern with the data-recording – that the voltage levels might not be properly adjusted for the 2-bit sampling system – turned out to be unfounded. One of the students (Adam) wrote a program that allowed us to check the data in real time. It turns out that the system is designed very well and automatically samples the signal appropriately. Adam was also in charge of verifying the test tone signal, which he did successfully in the first few minutes of our observing run.
Our observing session took place on Friday, April 15 from 9 a.m. to 11 a.m. PDT. We were able to project the telescope control and monitoring windows on two screens, and all the students were able to see the commands that we issued to the telescope and which area of the sky was being observed. We were assisted by Ryan Lynch and Frank Ghigo at the GBT, both of whom did an excellent job. The observations went very smoothly. However, the overhead time required to change from one source to the next was a little longer than I had estimated. About 90 minutes into the session, it became clear that we would have to sacrifice either observations of the last planet in our sequence or the GPS satellite in order to release the telescope to the next observer on time. Observing a GPS is challenging because the signal is so strong that it saturates the receivers and makes adjustment of the voltage levels throughout the receiver chain difficult. I was therefore relieved that the students preferred sacrificing observations of the GPS. In the end, we were able to observe all 14 exoplanets during an extremely successful and memorable observing session. We transferred all 7 terabytes of data to our dedicated storage server at UCLA and will start analyzing the data soon.
|
|
|
|
The UCLA classroom during our observations with the GBT on April 15, 2016. We operated the telescope remotely and projected all control and diagnostic windows on 2 large screens.
|
|
|
|
The GBT status window during our observations. It shows the source name, coordinates, configuration, and weather information.
|
|
What about the GPS data? We had originally intended to use the GPS observations to validate our software. Instead, we will use Voyager 1 observations recently released by the Breakthrough Listen project. I verified yesterday that the signal from Voyager 1, observed when it was 134 astronomical units from Earth, is indeed detectable in these data with the expected frequency drift rate. These data provide a much better test of the software than GPS data because the Voyager 1 signal is relatively faint. In addition, the students may be quite excited to search for and identify a signal emitted by a spacecraft at the edge of the Solar System.
The successful GBT observations are etched in my memory as one of the most meaningful events of my career. Thanks to our benefactors and GBT staff, UCLA students are looking for signs of extraterrestrial intelligence in data that they obtained at the telescope! I hosted a celebratory party at my house on Saturday night, with a sighting of the International Space Station. At the predicted time, all of us went outside and marveled at space activities around planet Earth, wondering what other civilizations might be constructing in space. If you have never seen the space station, I recommend that you download the heavens-above.com app or check out their web site. It's fun!
Warm regards,
Jean-Luc Margot
|
|
|
|
|
