An Agnostic Biosignature Based on Modeling Panspermia and Terraforming

TL;DR

This work uses an agent-based model to show that if life can spread between star systems, and affect the observable properties of a planet, then a robust signature of life can emerge, defined by correlations between planet characteristics and their locations.

Abstract

The search for a second instance of life is one of the greatest problems of modern science. Outside of creating an artificial origin of life on Earth, the primary targets for the search for life are planets inside or outside the solar system. Realistically, there are just a few locations to search for alien life within the solar system. Outside the solar system, opportunities are nearly unlimited, but there's a catch: it is difficult to attribute, with certainty, features of exoplanets to extraterrestrial life. Simple spectral biosignatures are susceptible to false positives; technosignatures reduce this susceptibility at the expense of strong assumptions about potential underlying life and its technologies. We have developed an agnostic approach to exoplanet life detection that overcomes these limitations by using properties that emerge on the scale of groups of planets, without the need for a "smoking-gun" single-planet level biosignature. We use an agent-based model to show that if life can spread between star systems, and affect the observable properties of a planet, then a robust signature of life (with very few false positives) can emerge, defined by correlations between planet characteristics and their locations. By clustering planets based only on their observed characteristics, and retaining clusters localized in space, we demonstrate (and evaluate) a way to prioritize specific planets for further observation, based on their potential for containing life. We consider obstacles that must be overcome to practically implement our approach, including identifying specific ways in which better understanding astrophysical and planetary processes would improve our ability to detect life.

Figures (21)

• Figure 1: Target planet selection and terraforming. A. The objective function, used for determining the destination of life from a terraformed "parent" planet. Candidate destinations are first constrained by a maximum positional distance threshold; among these candidates, the planet closest in composition to the parent planet is chosen as the target. B. Simulations are initialized with 1 origin of life, causing the initial distribution of planet compositions (seen in A) to become correlated. C. An example of how we determine target planet composition when retaining 10% of the pre-terraformed planet composition. Note that while our simulations use a 3D space, the concept figure shows only a 2D space for clarity.
• Figure 2: Mantel coefficient and p-value as a function of the ratio of planets terraformed. The earliest we observe a p-value $\leq 0.01$ is at a terraformed ratio $\approx 7\%$ (here, 70 planets).
• Figure 3: Mantel contribution as a cluster selection criterion. Clusters with a Mantel contribution $> 0$ (black) meet this selection criterion, indicating that their removal is a detriment to the residual space's Mantel coefficient (causing it to decrease). Clusters with a negative Mantel contribution shown in grey. The inset shows the full range of negative values in all clusters.
• Figure 4: Selection criteria used on clusters of planets. Spatial localization of clusters of planets is shown on the y-axis, as measured by the Interquartile Range (IQR) of each cluster. Horizontal dashed line (at $\text{IQR}=25.2$) denotes the threshold used, below which we selected clusters for being spatially localized. This corresponds to approximately the average IQR of planets in a cube the size of 1/8 the model space. Color bar shows the Mantel Contribution (MC) of clusters, with a high MC indicating a cluster as being important for raising the full space's Mantel coefficient.
• Figure 5: The earliest detected cluster in our simulation, at a terraformed ratio of $0.04$. This is a projection of 3D planet locations in the 2D X-Y plane, and the earliest time step where we detect a cluster of planets meeting our selection criteria. True terraformed planets ($n=40$) have blue fill, while planets detected by our selection method ($n=19$) have a red outline.