I was working on a project involving the search for biosignatures (signs of life) on exoplanets and the probability of that planet harboring life. In this case, we know that Earth is the only planet confirmed to have life on it. So the parameters of atmospheric conditions, radius, temperature, distance from the star for planets confirmed to have life is one (Earth).

Is there any way to use NNs to predict the probability of an exoplanet harboring life if we have the data of all these parameters for that planet?

  • $\begingroup$ I would very much doubt it. The problem in its most simple form is trying to generalize based only on a single data point i.e the Earth. What I think need to be done is finding data on deviations of radiation, temperature, air that a particular living being can survive and then construct a polygon/a closed surface around the data point i.e the earth based on the findings. If the uknown planets parameters fall in this region of closed surface it might harbour that life form (note that different life forms have different survivability under deviations from normal) $\endgroup$ – DuttaA Sep 8 '20 at 14:21
  • $\begingroup$ Also there is a basic presumption that carbon based life form is the only life form you are trying to find. And also presumption that the evolutionary process is invariant i.e it is not possible to evolve into something resistant to extremely harsh conditions. $\endgroup$ – DuttaA Sep 8 '20 at 14:23
  • $\begingroup$ @DuttaA thanks for the insight. I too think the same. I doubt that there's any substantial way to achieve it, given that the data we have is insufficient $\endgroup$ – Niranjan Dindodi Sep 8 '20 at 17:18
  • $\begingroup$ exoplanetarchive.ipac.caltech.edu/cgi-bin/FDL/nph-fdl?atmos Turns out that astrobiologists at SETI and NASA have come up with a dataset with simulated stable atmospheres for Earth-like exoplanets that might allow them to harbour life. This would probably be the only way to solve the one data point problem $\endgroup$ – Niranjan Dindodi Sep 8 '20 at 18:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.