An international team of researchers has found that artificial intelligence (AI) can help identify hidden patterns in geographic data that could indicate life on Mars.
Since there are few opportunities to collect samples from Mars when searching for life beyond Earth, it is crucial that these missions target locations that have the best chance of harboring extraterrestrial life. The new study, led by an international team of over 50 researchers, ensures this can be supported through the use of artificial intelligence and machine learning methods. This technology can be used to identify hidden patterns in geographic 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” was published in Nature Astronomy.
The resulting model was able to locate 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 on the border of Chile’s Atacama Desert and Altiplano in South America. This mapped the distribution of photosynthetic microorganisms. Gene sequencing and infrared spectroscopy have also been used to reveal unique markers of life, called “biosignatures”. Aerial imagery was then combined with this data to train a machine learning model to predict which macro and micro habitat types would be associated with biosignatures that could indicate life on Mars and elsewhere.
The resulting model was able to locate and detect biosignatures up to 87.5% of the time on data it was not trained on. This has reduced the search area required to find a positive result by up to 97%. In the future, life on Mars could be detected by identifying the areas most likely to contain signs of life. These can then be searched extensively by rovers.
dr Freddie Kalaitzis from the University of Oxford Computer Science department led the application of machine learning methods to microhabitat data. He said: “This work demonstrates an AI-guided protocol to search for life on a Mars-like terrestrial analogue on Earth. This protocol is the first of its kind to be trained on actual field data, and its application can in principle be generalized to other extreme living conditions. Our next steps will be to further test this method on Earth, with the aim that it will eventually support our exploration of biosignatures elsewhere in the solar system, such as Mars, Titan and Europa.”
On Earth, Pajonales, a four-million-year-old lake bed, is one of the closest analogs to Mars. This area is considered inhospitable to most life forms. Comparable to the evaporation basins of Mars, the high-altitude (3,541 m) basin is exposed to exceptionally high levels of ultraviolet radiation, hypersalinity, and low temperatures.
Water availability is probably the key factor in the location of biological hotspots
The researchers collected over 7,700 images and 1,150 samples, testing for the presence of photosynthetic microbes that live in 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.
Soil sample data and 3D topographical mapping were combined with the drone imagery to classify the regions into four macro-habitats (meters to kilometers) and six micro-habitats (centimeters). The team found that although the Pajonales had a nearly uniform mineral composition, the microbial organisms were clustered in different regions at the study site.
Follow-up experiments showed that water availability is the most likely factor in determining the location of biological hotspots, rather than environmental variables such as nutrient or light availability.
The combined data set was used to train convolutional neural networks to predict which macro and micro habitats are most strongly associated with biosignatures.
“Both for the aerial imagery and for the centimeter-scale ground-based data, the model showed a high predictive ability for the presence of geological material that is highly likely to contain biosignatures,” said Dr. Kalaitzis.
“The results agreed well with the ground truth data, with the distribution of biosignatures being strongly associated with hydrological features.”
The model is used to map other harsh ecosystems
Now the researchers want to test the model’s ability to predict the location of similar but different natural systems in the Pajonales Basin, such as ancient stromatolite fossils. The model is also used to map other harsh ecosystems, including hot springs and permafrost. The data from these studies will inform and test hypotheses about the mechanisms that living organisms use to survive in extreme environments.
“Our study has once again demonstrated the power of machine learning methods to speed up scientific discoveries, thanks to their ability to analyze vast amounts of disparate data and identify patterns that would be unrecognizable to a human,” added Dr. Added Kalaitzis.
“Ultimately, we hope the approach will facilitate the compilation of a database of biosignature probability and habitability algorithms, roadmaps and models that can guide research into life on Mars.”