Astrophysics with LOFAR

There are many different types of stars in the Universe. One particular type, called M dwarfs, are known to produce massive explosions on their surface. These explosions of energy, known as flares, can be detected across the electromagnetic spectrum from ultraviolet and x-ray to optical and radio waves. M dwarfs are much smaller and fainter than most stars, and LOFAR, as the largest radio telescope operating at low frequencies, is ideally suited to detect these faint radio sources in the sky. Radio observations allow us to understand the processes which drive these flares. Studying and understanding flares on other stars is very important for exoplanet research. Many exoplanets (planets orbiting other stars) have been detected around M dwarfs, and they are known to orbit much closer to their star than the Earth around the Sun. This means that any possible atmosphere of these exoplanets may be affected by large flares from the star. Observing these stars with LOFAR can provide information on the type of emission from stellar flares, and whether they can damage exoplanet atmospheres, potentially redefining the habitable zone around these stars. Better understanding these flares also allows us to compare other stars to our Sun, to better understand the potential for large solar flares. 

The Earth’s atmosphere frequently generates some of the most powerful electrical events in nature. This happens in the form of lightning during intense thunderstorms, when the atmosphere can produce megavolt electric fields and lightning can release up to a gigajoule of energy. This energy release causes the air around the lightning to get hotter than the surface of the Sun, exciting electrons which generate strong radio signals ranging from kHz – MHz, which we can detect with telescopes like LOFAR.

These extreme weather patterns are not unique to the Earth. Although the atmospheres may differ, powerful lightning strikes have been observed on other solar system bodies like Jupiter, Saturn, and Uranus. LOFAR has the frequency range and sensitivity required to detect these solar-system lightning storms from space. Regular monitoring of the most likely candidates to produce storms in our solar system at radio frequencies can help us understand the processes and formation of these space-storms and their powerful lightning events. This research could allow us to extend ground-based radio lightning observations to nearby exoplanets that are expected to have extended turbulent atmospheres. Observations like these could possibly place excellent constraints on the atmospheric variables for such a planet.

The solar atmosphere regularly releases huge amounts of magnetic energy, resulting in the acceleration of particles and the ejection of billions of tonnes of material into the solar system. These particles and eruptions can cause a threat to a variety of Earth-based technologies, such as damage to satellites and interruptions to electricity grids, as well as cause radiation hazards for astronauts or the crew of flights close to the Earth’s poles. Hence there is a need to understand the origin of this eruptive activity so that forecasts can be made of any resulting ‘space weather’. The energetic particles that accompany this activity are powerful sources of radio emission known as solar radio bursts (SRBs). The automatic detection of SRBs can allow us to study the statistics of solar eruptions, providing insight into the origin of such activity. However automatically detecting and classifying SRBs is a major challenge made more complex in recent years with new technology such as the Low Frequency Array (LOFAR). Each individual LOFAR station produces high-volume data streams (up to 3 Gigabits per second), hence processing and classifying SRBs in this data stream with accuracy is a significant computational challenge.

As the world of computing accelerates the digitalisation of everything, the world has a wonderful opportunity to solve ‘big global problems’. With that in mind, one specific aspect of a global problem is the effect the sun will have on our earth’s future. Artificial Intelligence (AI), can now help solve some of those problems. AI has made significant advancements in autonomous driving, robotics, voice recognition, and the recent landing of the “perseverance” probe on Mars. AI technology is now an everyday enabler in how we run our daily lives, the speed of how we can solve problems, and make decisions.

A recent example is how we managed to create an artificially intelligent algorithm known as YOLO (You only look once) that not only identifies SRBs but does it at high speed as well. What YOLO does is take hundreds of thousands of examples of SRBs and trains itself to identify different features of SRBs such as shape, colour, and intensity. YOLO is then presented with an observation made by LOFAR and scans the observation for SRBs. If YOLO identifies an SRB it then highlights it as an area of interest with what is called a bounding box as seen in the image.

Scientific and Technological Excellence by Leveraging LOFAR Advancements in Radio astronomy (STELLAR) is a collaboration between Institute of Astronomy and National Astronomical Observatory (IANAO) in Bulgaria, Netherlands Institute for Radio Astronomy (ASTRON), the Dublin Institute for Advanced Studies (DIAS), and the Technical University of Sofia (TUS), which is funded by EU’s Horizon 2020 Twinning program. STELLAR is a transformative project for training the next generation of Bulgarian radio astronomers, and will significantly increase the LOFAR technical and scientific expertise at TUS and IANAO.

STELLAR will achieve its objectives through carefully planned trainings for IANAO and TUS staff at ASTRON and DIAS, including lectures, workshops, summer schools, and research staff exchanges. The project will allow IANAO and TUS to develop and strengthen collaborations with ASTRON and DIAS.

The STELLAR project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 952439. Learn more at https://lofar.bg/stellar