What Spectroscopic Capabilities Does A Terrestrial Planet Finder Need?
Abstract
We employ artificial neural networks (ANNs) to develop a new, holistic approach to determining the trade-offs among the instrument parameters for nominal visible and infrared Terrestrial Planet Finder missions. Our results are aimed at demonstrating the technique rather than making precise instrument comparisons because any quantitative results will need to be refined based both on more realistic model instruments and more finely tuned mission requirements for distinguishing among planetary types. Nonetheless, in our preliminary study we find that an Earth-like planet with atmospheric oxygen is distinguishable by either a visible or infrared mission. For example, using ANNs trained to differentiate among Earth models with and without oxygen, as well as Mars-like, Venus-like, and Jovian planet models, we find that a spectral resolution of R=40 and signal-to-noise=10-20 is sufficient for at least 95% certainty in classifying these planetary atmospheres in the visible. The same problem requires R=15 and S/N=10-20 in the infrared. For the more demanding case where the ANNs must also distinguish among Earth-like models ranging from 0.01% to 100% of the Earth's current O2 abundances, we find that a visible-light TPF would require S/N > 85 at R=30 or S/N = 25-55 at R=200. The comparable problem in the infrared requires S/N=10-40 at R=20. These determinations can be refined based on different statistical thresholds of certainty, specific planet types of interest, and more developed instrument concepts.
- Publication:
-
Search for Life Beyond the Solar System. Exoplanets, Biosignatures & Instruments
- Pub Date:
- March 2014
- Bibcode:
- 2014ebi..confP2.37H