An international team of researchers has found that Artificial Intelligence (AI) can help identify hidden patterns within geographical data that could indicate life on Mars.
As there are only a few opportunities to collect samples from Mars in the search for life beyond Earth, it is crucial that these missions target locations that have the best chance of harbouring extra-terrestrial life. The new study, led by an international team of over 50 researchers, ensures that this can be supported by using Artificial Intelligence and Machine Learning methods. This technology can be used to identify hidden patterns within geographical data that could indicate the presence of life on Mars.
The work, ‘Orbit-to-Ground Framework to Decode and Predict Biosignature Patterns in Terrestrial Analogues,’ has been published in Nature Astronomy.
The resulting model was capable of locating biosignatures that have the potential to indicate life on Mars
The first part of the study, led by Dr Kimberley Warren-Rhodes at the SETI Institute, was an ecological survey of a 3 km² area in the Salar de Pajonales basin, at the boundary of the Chilean Atacama Desert and Altiplano in South America. This was used to map the distribution of photosynthetic microorganisms. Gene sequencing and infrared spectroscopy were also used to reveal distinct markers of life, called ‘biosignatures.’ Aerial images were then combined with this data to train a Machine Learning model to predict which macro- and microhabitat types would be associated with biosignatures that could indicate life on Mars and other areas.
The resulting model could locate and detect biosignatures up to 87.5% of the time on data on which it was not trained. This decreased the search area required to find a positive result by up to 97%. In the future, life on Mars could be detected through the identification of the areas most likely to contain signs of life. These can then be extensively searched by rovers.
Dr Freddie Kalaitzis from the University of Oxford’s Department of Computer Science led the application of Machine Learning methods to microhabitat data. He said: “This work demonstrates an AI-guided protocol for searching for life on a Mars-like terrestrial analogue on Earth. This protocol is the first of its kind trained on actual field data, and its application can, in principle, generalise to other extreme life-harbouring environments. Our next steps will be to test this method further on Earth with the aim that it will eventually aid our exploration for biosignatures elsewhere in the solar system, such as Mars, Titan, and Europa.”
On Earth, one of the most similar analogues to Mars is the Pajonales, a four-million-year-old lakebed. This area is considered to be inhospitable to most forms of life. Comparable to the evaporitic basins of Mars, the high altitude (3,541 m) basin experiences exceptionally strong levels of ultraviolet radiation, hypersalinity, and low temperatures.
Water availability is likely to be the key factor determining the position of biological hotspots
The researchers collected over 7,700 images and 1,150 samples and tested for the presence of photosynthetic microbes living within the salt domes, rocks, and alabaster crystals that make up the basin’s surface. Here, biosignature markers, such as carotenoid and chlorophyll pigments, could be seen as orange-pink and green layers respectively.
Ground sampling data and 3D topographical mapping were combined with the drone images to classify regions into four macrohabitats (metre to kilometre scales) and six microhabitats (centimetre scale). The team found that the microbial organisms across the study site were clustered in distinct regions, despite the Pajonales having a near-uniform mineral composition.
Follow-up experiments showed that rather than environmental variables, like nutrient or light availability, determining the position of, biological hotspots water availability is the most likely factor.
The combined dataset was used to train convolutional neural networks to predict which macro- and microhabitats were most strongly associated with biosignatures.
“For both the aerial images and ground-based centimetre-scale data, the model demonstrated high predictive capability for the presence of geological materials strongly likely to contain biosignatures,” said Dr Kalaitzis.
“The results aligned well with ground-truth data, with the distribution of biosignatures being strongly associated with hydrological features.”
The model will be used to map other harsh ecosystems
Now, the researchers aim to test the model’s ability to predict the location of similar yet different natural systems in the Pajonales basin, such as ancient stromatolite fossils. The model will also be used to map other harsh ecosystems, including hot springs and permafrost soils. The data from these studies will inform and test hypotheses on the mechanisms that living organisms use to survive in extreme environments.
“Our study has once again demonstrated the power of Machine Learning methods to accelerate scientific discovery through its ability to analyse immense volumes of different data and identify patterns that would be indiscernible to a human being,” Dr Kalaitzis added.
“Ultimately, we hope the approach will facilitate the compilation of a databank of biosignature probability and habitability algorithms, roadmaps, and models that can serve as a guide for exploration of life on Mars.”