By Lorenz Roth

On December 12, 2013 at 11 am EST, it was suddenly everywhere on the internet: The Washington Post, LA Times, Natural Geographic, and many more shared the story about water vapor plumes being discovered on Europa. How is it possible that an analysis of a bunch of pixels showing ultraviolet, or UV, light from a moon of Jupiter could be interesting to so many people? Well, in the end it is the idea that we might have come a little closer to discovering a habitable world outside Earth. For me, being a postdoc for not even a year, I was astonished, overwhelmed and a bit intimidated. I was also proud, in particular, that our method of using the ultraviolet images from the Hubble Space Telescope to search for water vapor had been successful.

This idea - to look for signs of plume activity in UV images of Europa's glowing aurora - actually existed before I even knew there was a moon of Jupiter out there called Europa. I cannot tell the story of a kid who spent every night on the rooftop staring into the sky with his self-earned telescope, hoping to fulfill his dream of becoming an astronomer. Instead, when I was looking for an interesting Master's thesis topic at the University of Cologne about seven years ago, I was interested in the methods and techniques in geophysics from a lecture given by Professor Joachim Saur. When I asked him about thesis topics, he showed me images of Europa's UV aurora taken by Hubble in 1999 that revealed a surprisingly irregular emission pattern. He explained that the observed irregularity might come from local enhancements in Europa's thin atmosphere, which could occur if active venting was present. However, there were many things to consider for the analysis and interpretation of the images.

After a while I realized that there seemed to be a rather strong hydrogen signal above the limb of Europa, which cannot be explained by reflected sunlight from the surface.

I liked the topic right away, although I had never been very optimistic that active plumes really exist on Europa and are observable in UV images, especially after looking at the unsuccessful searches for active venting with high-resolution imaging data provided by the Galileo and Voyager missions. However, the relatively small possibility that we might find hints of active plumes with these intriguing UV aurora images motivated me. We had to understand two things to be able to correctly interpret these UV images. First, the excitation mechanism of the aurora was unclear. When and where do electrons from the Jovian magnetosphere collide with Europa's neutral gas to trigger the aurora? And second, do we fully understand the complexities of the Hubble spatial-spectral images? Hubble images record not only the auroral emissions we want to study, but also light from the geocorona, which is the hydrogen-dominated outermost region of Earth's atmosphere; the interplanetary medium, surface-reflected sunlight; and other sources, each of which considerably complicates the analysis of the images.

The first question is better understood as a result of our recent work, but is not yet fully answered. In principle the auroras of Europa should be brightest where the most electron-neutral collisions take place. However, the interaction between the moon and the plasma around Jupiter complicates this correlation. The plasma - electrons and ions in Jupiter's magnetosphere - rotates with the planet's magnetic field with a short period of about 10 hours. Europa, in contrast, needs more than 85 hours to orbit Jupiter, so it is constantly overtaken by the plasma. The relative movement leads to various collisions between the Jovian plasma (which is electrically charged) and the neutral gas environment of Europa; electric currents are driven through the ionosphere of Europa and then northward and southward along the magnetic field all the way to the upper atmosphere of Jupiter, where they close the circuit. This magnetosphere-atmosphere interaction leads to a strong diversion of the Jovian plasma around Europa, somewhat like water flowing around a solid obstacle. Generally, more neutral gas means more collisions and a stronger diversion until the electric current is finally saturated. Thus, although a local gas enhancement (such as from a plume) might lead to a brighter aurora, the flow diversion issue makes a correlation between neutral gas and aurora brightness uncertain. In short, it needs further study.

When it comes to the second task, understanding the Hubble data, I made great progress when I worked for a couple of months at the Johns Hopkins University in Baltimore. I learned to process the raw image data from Hubble's Space Telescope Imaging Spectrograph, or STIS. I learned to extract images of the fingerprints of oxygen (at wavelengths of 130.4 nanometers and 135.6 nanometers) and of hydrogen (the so-called "Lyman-alpha" line at a wavelength of 121.6 nanometers), which are simultaneously recorded on the detector.

With our initial understanding of the excitation processes, we then critically re-analyzed STIS images of Europa from 1999 and found some possible, yet inconclusive, hints of anomalies in the atmosphere-related aurora. A similar intriguing surplus has also been detected in Hubble images from 2008. STIS is the best-suited camera on Hubble for these aurora observations, and after it was repaired during the final Space Shuttle mission to Hubble in 2008, we thought that we should try again (maybe one last time?) to monitor Europa's atmospheric glow with its unique capabilities.

The new observations were carried out right before I started my postdoctoral position with Kurt Retherford, at the Southwest Research Institute in San Antonio, in January 2013. Joachim Saur, who led the observation campaign, agreed to let me lead the analysis of the new images, while he remained involved in all steps. I was of course happy to work with the promising new Hubble data. At first glance there were not any obvious anomalies. But after a while I realized that there seemed to be a rather strong hydrogen signal above the limb of Europa, which cannot be explained by reflected sunlight from the surface. The coincidence with an oxygen emission surplus indicated that these signals must indeed originate from water above the surface of Europa! And water can, in principal, only exist if a local, active source is present, since it would immediately freeze on Europa's surface. So, the only reasonable explanation for our signal was active water vapor plumes!

We submitted the results to Science magazine and the paper was accepted just 10 days before the fall meeting of the American Geophysical Union (AGU) in San Francisco, where I was already scheduled to present a talk about Europa's UV aurora. The Science editors managed to make the paper public online on the same day as my talk, and a press release on the discovery also went out at the same time. And NASA even organized a press conference at AGU to announce the discovery.

So, six years after I walked into the office of Joachim Saur in Cologne to ask him for an interesting topic to work on, I was sitting next to him - along with Kurt Retherford and Jim Green from NASA Headquarters - to announce that our search for hints of water vapor in Europa's UV aurora had been successful. Half an hour later I opened my web browser and checked the big news sites to find overwhelming coverage of the story.

The excitement surrounding the whole thing is the idea that these plumes are connected to subsurface liquid water on Europa, which is speculated to provide a stable environment and all the ingredients for life: water, energy and the essential chemical elements. If the plumes are indeed connected to lakes or an ocean within Europa, we might soon be able to probe the composition of the liquid water and add another important piece of information to our understanding of this potentially habitable world.