Skip to main content

Finding faint fireballs in the dark desert sky

Research team: Dr. Paul Hancock, A/Prof. Randall Wayth, Xiang Zhang, Prof. Steven Tingay, Sean Mattingley, Hadrien Devillepoix, Prof. Phil Bland
CIC specialists: Shiv Meka, Dr. Kevin Chai

Researchers at the Curtin University node of the International Centre for Radio Astronomy Research are interested in radio emissions coming from fireballs (meteors).  Any transient radio emissions from these sources could be detected by radio telescopes such as the Murchison Widefield Array (MWA) and the future Square Kilometre Array (SKA), and have the potential to provide new insights into how meteors interact with the ionosphere as they fall to earth.  Being able to study meteors via their radio emissions would also give access to elusive daytime meteor showers that can’t be detected with optical telescopes.

The Desert Fireball Network (DFN) already has a network of 50 optical cameras across the Australian Desert, and software that can detect the brightest fireballs as they descend – the ones most likely to leave a meteorite.  But the research team needed to also detect the faintest meteor streaks in the sky, and correlate those sightings against radio telescope data to identify any characteristic radio emissions.

Working with Mr Hadrien Devillepoix from the DFN, the radio astronomy team installed a modified DFN camera with a narrower field of view and optimised for higher sensitivity, the ‘astrocam’, out in the Murchison to look at the same part of the sky as the MWA.  Using this they collected photos during nights of the Geminid meteor shower to correlate against the MWA data.  The difficult part was to then find all of these rare meteor trails, faint as well as bright, in thousands of photos of the night sky to compare with any radio signals originating from the same time and place.

Dr Kevin Chai and Mr Shiv Meka from the Curtin Institute for Computation joined the project to build a machine learning algorithm to automatically detect even the faintest meteor trails in the mass of digital photos.  To train it to identify meteor trails, Ms. Xiang Zhang sorted through 6,000 images and manually identified and annotated about 70 meteors, but this wasn’t a large enough data set to train the convolutional neural network effectively.

To bypass the lack of training data and avoid weeks of manual searching for more of these rare events, Chai and Meka created artificial training images, taking pictures of the night sky without meteor trails from the data set and adding random lines in a variety of positions, orientations, lengths, thicknesses and light intensities to mimic meteor trails.  They rapidly created a dataset of 30,000 simulated meteor trails which was then used to train the neural network.

The resulting algorithm was surprisingly successful at detecting meteor trails in the 6,000 real digital photos that had been manually examined and annotated by Zhang.  It found Zhang’s 70 meteors, but identified quite a few others.  What were initially considered ‘false positive’ results were in every case shown on manual re-examination of the images to be genuine examples of faint meteor trails, or other transient phenomena such as satellites.

The algorithm is now being trained on a larger artificial data set that incorporates different weather events such as high cloud, different phases of the moon, and different amounts of Galactic background (the Milky Way) in the sky at different times of year, to enable it to better identify meteor trails under varying conditions.

Once the astrocam data is calibrated against the DFN network, the algorithm will be able to sift through years of accumulated observations, finding rare events like meteor trails to compare against the radio emissions recorded at the same time and location by the MWA.  Automating feature detection in the optical and hence the radio data will allow the team to detect and study all sorts of transient events, from meteors to space junk re-entering the atmosphere.  The dataset is already there, but it will take a machine learning algorithm to find the matching needles in the haystack.