Machine learning will help us identify habitable exoplanets

Machine learning will help us identify habitable exoplanets

To date, 4,940 exoplanets have been confirmed in 3,711 planetary systems, with an additional 8,709 candidates awaiting confirmation. With so many planets available for study and improvements in telescope sensitivity and data analysis, the focus is shifting from discovery to characterization. Instead of simply searching for multiple planets, astrobiologists will scan "potentially habitable" worlds for potential "bio-signatures."

They are looking for chemical signatures associated with life and biological processes, one of the most important of which is the 'water. As the only known solvent of which life (as we know it) cannot exist without, water is considered the divinatory rod to find life. In a recent study, astrophysicists Dang Pham and Lisa Kaltenegger explain how future investigations (when combined with machine learning) could discern the presence of water, snow and clouds on distant exoplanets. This importance is reflected in NASA's slogan - "just follow the water" - which also inspired the title of their article.

Dang Pham is a graduate student at the David A. Dunlap Department of Astronomy & Astrophysics at the University of Toronto, where he specializes in planetary dynamics research. Lisa Kaltenegger is an associate professor of astronomy at Cornell University, director of the Carl Sagan Institute and a world-renowned expert in modeling potentially habitable worlds and characterizing their atmospheres.

Currently, to identify the presence of water, astronomers have limited themselves to looking for the absorption of the Lyman-alpha line, which indicates the presence of gaseous hydrogen in the atmosphere of an exoplanet. This is a byproduct of atmospheric water vapor that has been exposed to solar ultraviolet radiation, causing it to chemical dissociation into hydrogen and molecular oxygen (O2), the former of which is lost to space while the latter is maintained.


Thanks to their large primary mirrors and advanced spectrograph suite , chronographs, adaptive optics, these instruments will be able to conduct direct imaging studies of exoplanets. This consists in studying the light reflected directly from the atmosphere or from the surface of an exoplanet to obtain spectra, allowing astronomers to identify which chemical elements are present. But as Lisa Kaltenegger and Dang Pham indicate in the article, this is a time-consuming process. Astronomers begin by observing thousands of stars for periodic dips in brightness, then analyzing the light curves to locate the signs of chemical signatures. Currently, exoplanet researchers and astrobiologists rely on amateur astronomers and mechanical algorithms to sort the volumes of data obtained from their telescopes. Looking ahead, Pham and Kaltenegger show how more advanced machine learning will be crucial.




"Next-generation telescopes will search for water vapor in planets' atmospheres and water on the surface of the planets, ”Kaltenegger said. "Machine learning allows us to quickly identify the optimal filters, as well as the trade-off in terms of accuracy in various signal-to-noise ratios," added Pham. "In the first task, using [the open source algorithm] XGBoost, we get a ranking of which filters are most useful for the algorithm in its water, snow or cloud detection tasks. In the second task, we can observe how much better the algorithm performs with less noise. "

To make sure their algorithm was up to the task, Pham and Kaltenegger did a remarkable calibration. This involved creating 53,130 spectral profiles of a cold Earth with various surface components, including snow, water and water clouds. They then simulated the spectra for this water in terms of atmosphere and surface reflectivity and assigned color profiles. As Pham explained: “We train XGBoost on these colors to perform three separate objectives: to detect the existence of water, the existence of clouds and the existence of snow”.


The water is more difficult to detect, but identifying water, snow and clouds through photometry is doable, ”said Pham. Likewise, they were surprised to see how well the trained XGBoost could identify water on the surface of rocky planets based on color alone. The proposed method does not identify water in the atmospheres of exoplanets, but on the surface of an exoplanet by photometry.

Furthermore, it will not work with the Transit Photometry method, which is currently the most used and effective medium for detection of exoplanets. This method consists of observing periodic dips in brightness from distant stars, attributed to exoplanets passing in front of the star relative to the observer.

Sometimes, astronomers can obtain spectra from the atmosphere of an exoplanet while making a transit, a process known as 'transit spectroscopy'. As sunlight passes through the exoplanet's atmosphere from the observer, astronomers analyze it with spectrometers to determine what chemicals are there. For example, using its sensitive optics and suite of spectrometers, the JWST will rely on this method to characterize the atmospheres of exoplanets.






Powered by Blogger.