“Everyone agrees your method is not going to work”: William Borucki on the power of resolve in NASA’s Kepler mission

The primary investigator of NASA’s historic Kepler mission talks data, algorithms and determination

Kepler PI William Borucki with Kepler scientist Joergen Christensen-Dalsgaard of Aarhus University, Denmark. (Photo: NASA/Ames Dom Hart)

When NASA’s Kepler space telescope began gathering its first bytes of data 10 years ago, on May 13 2009, it marked a major development in our understanding of the galaxy, revealing that the night sky was filled with billions of hidden planets. Over its nine-and-a-half-year mission to establish the probability of planets beyond the solar system, in particular Earth-size planets in the habitable zone, Kepler observed a total of 530,506 stars and detected well over 2,000 planets. The data it gathered led to the publication of more than 2,800 peer-reviewed articles.

But behind the numbers, the mission was testament to the perseverance and commitment of the women and men at NASA who worked to turn an idea into a reality.

The mission generated a huge amount of interest in the scientific community and popular media. As Editor Dr. Jack J. Lissauer of NASA’s Space Science and Astrobiology Division notes in the introduction to a new virtual special issue about the Kepler mission published in Elsevier’s journal New Astronomy Reviews:

Kepler appears to have burst upon the scene ready for battle, like Athena springing forth, fully formed, from the head of Zeus.

New Astronomy Reviews virtual special issueRead our virtual special issue

Our virtual special issue on the NASA Kepler Mission in New Astronomy Reviews celebrates 10 years since Kepler first started gathering data. Articles are freely available through May 2020.

William BoruckiFrom the perspective of the mission’s principal investigator, William Borucki, Kepler embodied something exceptionally thrilling about the nature of scientific progress:

Here was a mission that would help us understand the possibility of life in our galaxy, whether there are Earth-sized planets in the habitable zones of other stars, and do most stars have planets. The answer was ‘yes,’ and that meant that life could be plentiful in our galaxy. It also confirmed that the theory that when a star forms, planets are also formed from the leftover gas and  dust, is correct.

Going from speculation to knowledge is what science is all about. The whole world shared in the excitement of the new discoveries of planets, planetary systems, and of the properties of the stars they orbit.

 This illustration shows a hypothetical planet covered in water around the binary star system of Kepler-35A and B. With two suns in its sky, Luke Skywalker's home planet Tatooine in "Star Wars" looks like a parched, sandy desert world compared to the possible water worlds Kepler found. The data show that 2-star systems support planets and planetary systems. (Image: NASA/JPL-Caltech)

Overcoming doubt

While the Kepler mission may have appeared “fully formed” to the world at large, the reality was that it was the result of decades of resolve from Bill and his colleagues:

For 14 years, the feeling in some quarters was ‘your method will never work, do something useful, Bill.’ But I recognized that if we’re going to explore the galaxy, it was critical to determine whether Earth-like planets were common or rare.

The method Bill proposed was to use a photometer on board a space telescope that would monitor the brightness of hundreds of thousands of stars, producing data that would then be analysed to detect exoplanets that caused stars to dim by crossing in front of them.

Early in its development, the project faced a critical test:

We had a new division chief – Dave Morrison – who came into the organization, and as he got to know the goal of each person. He called me into his office and said, ‘Look, Bill everyone agrees that what you’re doing is not going to work. However, I will set up a non-advocate committee – led by Jill Tarter, the former director of the Center for SETI Research – who will look at the project without bias and listen to your team’s reasons for continuing. If the committee decides in the negative, I will terminate the project.’

After the day-long presentation, the committee agreed that it might be successful. In fact, several of the committee members asked to join the team, and Dave Morrison said, ‘OK Bill – since they believe it could be successful, I’ll get you the funds you need to continue the development of the mission concept.’

Carl Sagan (Image: Nasa/JPL)Among the team members was the late Carl Sagan, the American astrophysicist known for – among many other things – assembling the physical messages sent into space on the Voyager Golden Record. Bill recalled his unwavering support for the endeavor:

Every time we proposed the project – in 1992, 1994, 1996, 1998 – we were rejected, and the team got to be known for rejected proposals. So, every time I went to propose again, I would ask my team, ‘Are you sure you want to be a part of this? Do you still believe in it?’ And Carl said, ‘Yes, I want to be a part of it. I know you failed – keep trying.’

I still have a letter on my wall from Carl saying that he wished he was in better health so he could help more.

When Kepler started to generate results, Bill saw it as testament to the hard work of the women and men involved:

I felt very happy for all the people who joined me to make the mission successful. I advocated and led the project development for almost two decades, but when it came to executing it, there were thousands of people involved. They believed in it, they worked hard for it, and I was delighted that we together succeeded in accomplishing something important for humankind.

AI helps NASA researchers draw new insights from data

The special issue in New Astronomy Reviews focuses on Kepler’s first discovered exoplanets, with each contribution highlighting a different planet. As Dr. Jack Lissauer explained: “Special issues like this are very useful as they bring together relevant articles around a certain theme and produce an overall volume in which both authors and readers benefit from the synergy between individual contributions. A journal like New Astronomy Reviews is particularly suitable for themed issues as this journal has such a broad and large audience.”

Borucki’s contribution to the issue concerns the planet Kepler-62f, and in the paper he suggests that based on its size, stellar host and orbital distance, it is the known exoplanet most likely to be habitable. Kepler 62f was identified as a planet candidate in early Kepler catalogs but was labelled a false positive by the “Robovetter,” an algorithm designed to ensure uniformity of criteria of credibility of planet candidates.

In his submission to New Astronomy Reviews, Bill and his co-authors argue that comprehensive analysis of all the observations indicates that Kepler-62f is a true planet. For Bill, that represents AI working as it should in science – supporting the work of researchers and prompting further investigation, but not acting as a mindless replacement:

AI is a wonderful development in respect to finding signals in noise. You have all this data, and an algorithm can allow you to dig deeper into the noise to find a weak signal that inspires further investigation. There are probably 100 or more signals from small planets buried in the Kepler data that have not been found because not all the structures in the noise have been removed. One day, people who use more powerful analysis methods might be able to find them. I really like the idea that in the future, people will dig through the data and find those small planets.

Researchers should get credit for their data

Indeed, the Kepler telescope transmitted huge amounts of data, which were then shared with the broader science community and the public. Elsevier has indicated before that for data sharing to become established practice, researchers need to be able to first take ownership of their data – a point that Bill strongly agreed with:

When people build a mission like this, it’s not something they do for an hour a week or an hour a day. It’s 12 hours a day, six days a week, year after year. That means you’re not publishing papers, and if you’re not publishing papers, your career is severely impacted. Note how few experimentalists sit on high-level committees.

The people who spend a decade or more to develop the mission depend on having sufficient proprietory time to be the first to publish papers describing the mission discoveries. That often takes awhile because the data reduction programs take time to handle the usually unexpected complexity of real data. Only then can analysis and the confirmation observations be made to validate mission results.

Releasing the data to the public and the science community starts a period when everyone explores the data to produce an explosion of even more and varied discoveries.

Continuing the mission

Through this ongoing work, insights into the data captured by Kepler will continue and further develop our understanding of the way planets are formed and the probability that there is intelligent life elsewhere in our galaxy.

“The excitement that Kepler generated in the public and scientific communities, and the influx of many young scientists, are vital aspects of the scientific process,” Bill said. “The public deserves to know what we’re finding so that they can develop their own understanding of our place in the Universe.”

Quick question for you

Which terms do you most associate with Elsevier? (check all that apply)

Data and analytics
Research platforms
Technology
Decision support tools
Publishing
Books and journals
Scientific articles
Healthcare content

Tags


Contributors


Related stories


Comments


comments powered by Disqus