Professional astronomers spend a lot of time on reading other astronomers’ research to learn what’s going on in the field and to incorporate new, best results into their work.
Traditionally, that research is officially presented to the astronomical community in the form of a peer-reviewed article printed (or at least hosted) in a professional astronomical journal, such as “The Astrophysical Journal“, “The Astronomical Journal“, or the venerable “Monthly Notices of the Royal Astronomical Society“.
However, scientific research articles are very unlike newspaper or magazine articles — they don’t usually employ a narrative structure, and they include confusing words and references. Consequently, it can be hard for people new to the field to read and understand them.
So in response to insightful requests from my students, here’s a short primer about how to read astronomical research articles. A lot of this information probably applies to all scientific articles, but there are also some aspects unique to astronomical articles. If you have suggestions to improve this post, don’t hesitate to contact me.
How to Access Scientific Articles
There are lots of services to find astronomical articles, but the vast majority of astronomers use NASA’s Astrophysics Data System service to find articles. That service provides links to articles hosted on official journals’ websites (which may require a subscription to access) and (if they are available) to free versions on the open-access pre-print server astro-ph (which is part of the arxiv.org service).
Most journals allow the article authors to send out free versions of the published articles to anyone who requests them. So if you can work up the nerve (and most astronomers are very nice people or at least eager to have others read their work), e-mail the authors directly to politely request a copy. (Here’s a good article about how to e-mail scientists.)
Journals are the official repositories for the final versions of scientific articles. If an article appears in such a journal, it’s (probably) been through a review process (see below) and meets some basic standards of quality.
That doesn’t mean the results of an article are correct, and it’s not unusual for results in one article to be contradicted by subsequent articles (even subsequent articles from the same scientists). But with a published article, you can have some confidence in the results and process.
Many journals nowadays are managed by private companies, which must turn a profit. Consequently, subscriptions to some journals are very expensive, which severely limits public access to scientific research that has been supported by public tax dollars. (See stories like this one.)
This is why the astro-ph archive is a big deal – you can usually get free access to an article. One caution, though: ANYONE can post articles to astro-ph, and astro-ph articles have NOT necessarily been reviewed for accuracy by anyone.
How Articles are Written
The process of publication is, in many ways, arcane, confusing, and backwards, and scientists in many (but maybe not all) fields are working to improve it.
In a nutshell, a professional astronomer will spend several months, sometimes years, on a research project – running computer code, collecting observations, conducting experiments, etc.
At some point in the course of the project, the scientists will decide they have a self-contained, compelling story (knowing when to cut off a project and publish is almost more art than science). Then they will (if they haven’t already started) draft a scientific manuscript.
That manuscript usually includes
- Context and motivation for the project – What does recent, past work say about the problem? What questions remained unanswered?
- Technical aspects of their approach – What physical approximations were used in the code, and where might they fail? How long was each astronomical observation?
- Results from the project – What did the observations tell us about the planetary system?
- Summary of the conclusions and plans or suggestions for future work – How do the new results relate to the previous work? What observations should we collect next?
Eventually, the authors are satisfied with (or at least resigned to) the draft manuscript, at which point they submit to a journal.
The journal sends the article out to the other scientists (called “referees”) with relevant expertise for a (hopefully but not always) objective assessment of the work. There is usually some back-and-forth between the referees and authors (who are often kept anonymous to one another), with suggestions for improvements. Eventually, a final manuscript is “accepted for publication” and printed and/or posted online.
Reading Research Articles
Scientific articles can seem a little bit like a Gordian knot – convoluted and indecipherable. But the best way to read a scientific paper is to chop it into pieces, not to read the article from beginning to end like a short story. I’ll use a recent article from my own group as an example.
The image above shows the article’s first page. Different journals have different formats, but most will have the same information on the first page:
- Title of the article – (hopefully) tells you what the article is about
- Author list and affiliations – Who wrote the article, how can you get in touch with them. Usually, the person listed first (the “first author”) was in charge of the project and is the person you should contact if you have questions.
- Abstract – a short summary of article and main conclusions
- The introduction – Background and context for the project
Throughout the article, you will see references to previous work – for this article, those references look like “(Knutson et al. 2007)”. That means an article written by a group (“et al.“) led by someone named Knutson in 2007. In “The Astrophysical Journal”, the complete reference information is given at the end of the paper.
When I read a paper, I usually read the abstract carefully to get a clear sense for the paper’s about and what the authors conclude. Then I read the introduction if it’s pretty short (about a page). Longer than that, and I usually skim the introduction.
Usually the next section you see will be the technical description, which will include lots of figures and equations. I skip this section on my first read-through. It’s easy to get lost in the details, and if you’re not familiar with the techniques, this section will be undecipherable.
Then comes the results section. I will usually read this section if it’s short, with “short” meaning again one or two pages.
Finally, comes the conclusion and discussion. Though this is the last section of the paper, it’s usually the second section I read (after the abstract and/or introduction).
The last page of the article will usually include an Acknowledgments sections in which the authors will thank anyone who contributed to the article but is not listed as an author (including the anonymous referees who reviewed the paper).
After that comes the references section, which provides the citation information for all the previous work referenced. Journals nowadays often use abbreviations that can be a little cryptic, so let’s look at the earlier example:
The figure above shows the reference information. We see the first three authors listed: Knutson, H.A, and then Charbonneau, D., and then Allen, L.E. The “et al.” means there were more contributing authors, but they are not listed to save space.
The “2007” means that was the year the article was published (but not necessarily the year the work was done), followed by “Natur”. In “The Astrophysical Journal”, “Natur” is short-hand for the journal “Nature“. Some journals don’t use such abbreviations, and others will have different abbreviations.
Finally, we see “447, 183”. These numbers usually refer to the volume and page number(s) in the journal where the article appears. Not all journals have volume or page numbers like this, so reference styles may vary.
All of this information is helpful if you want to find the referenced articles, which you can usually do through NASA ADS. In fact, usually all you need is the first author’s last name and the year the paper was published to find it, and ADS has a good guide about how to use the service to find articles.
In spite of the ubiquity of dust on Mars, though, the physics of dust-lifting and transport remain mysterious. For instance, the seasonal appearance of dark streaks on slopes across the surface of Mars, called recurring slope lineae, were thought to result from flow of brine. Recently, though, we’ve found they are more likely granular flow, but what exactly drives their seasonality is unknown.
Experiments conducted by the aeolian physics group at University of Duisburg-Essen has brought a little clarity to the mystery of martian dust.
One of the biggest challenges for experiments exploring martian dust transport is replicating Mars’ low gravity (40% of Earth’s) and air pressure (10% of Earth’s). Since gravity and pressure help determine how winds move dust, accurate experiments must somehow create winds in a low-gravity environment under near vacuum.
To make a little pocket of Mars on Earth, Maximilian Kruss and colleagues took a small vacuum chamber centrifuge onto a “vomit comet” and conducted parabolic flights to create short periods of microgravity.
They filled the chamber with martian-like dust grains and turned up the fan to figure out when the winds were strong enough to start blowing the dust. By imaging the grain bed and tracking the grains, they estimated this threshold wind velocity, which is key to understanding when and where Mars can blow dust around.
Kruss and colleagues found, reassuringly, that theoretical models about dust transport were accurate. These results help us understand aeolian processes on a wide range of bodies, not only on Mars but any body with a low-pressure atmosphere.
Indeed, even comets play host to aeolian processes. When the Rosetta Mission flew past comet 67P, it saw features on the comet’s surface that looked for all the world like wind streaks and dune fields.
Kruss and colleagues suggest dust transport may be important on some exoplanets, where gravities and atmospheric pressures span an even wider range than in our solar system. And so, these results, taken from a tiny vacuum chamber, may bear on processes on worlds across the whole galaxy.
Even though the New Horizons Mission completed its flyby of the Pluto system more than four years ago, analyses of the treasure-trove of data returned continue to reveal how complex and colorful this system really is.
In our research group meeting today, we discussed a recently published analysis of landslides observed on Pluto’s moon Charon from Chloe Beddingfield of the SETI Institute. (Before continuing, let’s just pause for one second to reflect on the fact that we can talk about catastrophic geological events on a world almost 8 billion kilometers from Earth …)
As New Horizons zipped past the Pluto system back in 2015, it snapped lots of relatively high resolution photos of Charon’s surface, revealing an icy surface covered in craters, ridges, faults, and other weird geological features we don’t understand.
Among these striking features, landslides stood out prominently. Landslides on Earth often happen when an unstable hill or ridge becomes saturated after a rainstorm (as happened a few years ago in Boise).
On Charon, though, there is no rain (or atmosphere for that matter), so the landslides there probably occurred as the result of seismic activity, that is, a moonquake.
An enduring mystery of terrestrial and extraterrestrial landslides alike is why they are able to run out for such long distances even though they don’t fall very far.
These very long landslides are unimaginatively called long runout landslides. The Blackhawk Landslide in the San Bernardino mountains, for example, fell about 1 km but ran out for more than 8 km. Scientists have speculated the landslides may be cushioned by a layer of trapped air, which reduces friction as they rumble along.
Since it has no atmosphere, Charon’s landslides can’t be cushioned by air, and so Beddingfield and colleagues set out to determine whether these landslides are also anomalously long. By carefully mapping the landslides’ topographic profiles using New Horizon’s imagery and altimetry data, they found the landslides ran out between two and four times farther than they should have, based on the available gravitational energy.
These results confirm that long run-out landslides have made the long run-out all the way to the Pluto system, and something besides atmospheric cushions must be responsible. Other explanations include powerful sound waves carried along by the tumbling debris itself, a hypothesis called acoustic fluidization — which would be a great geology-based cover band name.
This study shows us we still have a lot of learn about the most distant objects in our solar system. Even though Charon may be the moon of a dwarf planet, it seems to have the same geological potency of a full-sized planet.
As you may have heard, exoplaneteers Michel Mayor and Didier Queloz shared this year’s Nobel Prize in Physics for the first discovery of an extrasolar planet around a Sun-like star.
These discoveries are now so commonplace, with thousands of exoplanets now known, it’s hard to remember back when each individual discovery was groundbreaking. So to reflect on how far we’ve come, we went back to Mayor and Queloz’s original 51 Peg b discovery paper at our research group meeting on Friday.
Even after decades of exoplanet discoveries, their paper is a gem, with bold scientific claims buttressed by meticulous observational data. As a gas giant circling its host star every 4 days, 51 Peg b presented a clear challenge to our notions of planet formation which said gas planets like Jupiter can only form very far from their host stars. Even so, Mayor and Queloz built a nearly bullet-proof argument for their discovery, and their results were confirmed within a week of their announcement.
But re-reading the paper this week, I was especially struck by how much our understanding of exoplanetary systems has changed and how many of their arguments, perfectly plausible at the dawn of exoplanet science, have been turned on their head — literally.
A Wobbly Rainbow
To find 51 Peg b, Mayor and Queloz used what has now become a standard exoplanet discovery technique, radial velocity measurements. The animation above shows how this works: as a planet circles its host star, the star also revolves around the planet. If the planet’s orbit is not too far from edge-on as seen from Earth, the Doppler effect will raise or lower the pitch (i.e., color) of the star’s spectral features as the star pirouettes toward and away from Earth.
With this technique, Mayor and Queloz detected the teeny gravitational tug of 51 Peg b on its host star to find the planet and estimate its mass (about half Jupiter’s).
As powerful as this technique is, though, if the planet’s orbit is not exactly edge-on as seen from Earth, the mass inferred is smaller than the actual mass. And so when Mayor and Queloz detected 51 Peg’s gravitational gumboot, they couldn’t be sure whether they had detected a gas giant in an orbit nearly edge-on or a small star in an orbit nearly face-on.
To address this uncertainty, Mayor and Queloz measured the star’s rotation and found the equator was nearly edge-on to Earth. Since the orbits for solar system planets are all nearly aligned with the Sun’s equator, it seemed obvious that 51 Peg b’s was as well.
So the inferred radial velocity mass must be close to the actual mass, and spectral oscillations were caused by a planet.
Later observations of 51 Peg b confirmed this alignment assumption. But we now know that many exoplanet orbits are severely misaligned compared to their stars’ equators. In some cases, the planets actually orbit at a right angle or even in the opposite direction to their stars’ rotation.
The reasons for these misalignments are not clear — in some cases, the stars might undergo an early phase of chaotic rotational evolution. In other cases, the planets might start out in well-aligned orbits, but gravitational interactions among planets or with a distant star can produce misalignment.
It Could (and Did) Happen Here
Even though 51 Peg b seems not to have experienced this misalignment, its discovery forced astronomers to reconsider the canonical wisdom of planet formation and think outside of the box about where we might find planets. Once it became clear that Jupiter-sized planets could occupy very short-period orbits, radial velocity observers sifted their data again and found dozens of planetary signals hiding where no one had thought to look before.
And the dramatic orbital evolution later invoked to explain 51 Peg b’s very small orbit prompted astronomers to re-visit previously puzzling aspects of our own solar system. Now we think the same kinds of prepubescent shake-ups that occur regularly in extrasolar systems probably also happened here, perhaps explaining why Mars is so much smaller than Earth and unlikely arrangements of orbits in the Kuiper Belt.
51 Peg b’s Legacy
51 Peg b is sometimes mistakenly called the first exoplanet discovered, but, in fact, the first confirmed exoplanet system was discovered in 1992 orbiting the pulsar PSR B1257+12. However, as the first planet orbiting a Sun-like star, 51 Peg b definitively demonstrated the existence of planetary systems resembling our own.
And here, 25 years after its discovery, we know planetary systems are common, with on average at least one planet for every star in our galaxy. The awarding of the Nobel Prize to Mayor and Queloz (as flawed as the Nobel awards are) is a rightful recognition of the profound importance of their work. Indeed, the discovery of 51 Peg b was not just a stunning testament to human achievement — it’s a response to the age-old question, “Are we alone in the Universe?”. Each exoplanet discovery since then whispers the answer, “No, we are not.“