I promised a quick return to recent work on the Drake Equation, which helps us estimate the number of communicating civilizations in the galaxy, but a BBC story on Duncan Forgan has me back at it even sooner than I had intended. It’s no surprise that the matters encapsulated in Drake’s thinking should be in the news. After all, the era of Fermi and Drake was without firm knowledge of extrasolar worlds, of which we now know over three hundred. For that matter, the concepts of habitable zones around both stars and the galaxy itself had not come to fruition, nor had anyone ever heard of the ‘rare Earth’ hypothesis.

We also work today with knowledge of Charles Lineweaver’s studies of the median age of terrestrial planets in the Milky Way, which point to civilizations around other stars having had as much as two billion years-plus to emerge before our own Earth had even coalesced. Until we know more, I suspect we’ll be adjusting Drake parameters for some time, as Duncan Forgan (University of Edinburgh) does in his new paper.

Hypotheses and Conclusions

Forgan’s simulations, which take current exoplanet findings into account, work with several contrasting hypotheses, for each of which statistical methods were applied. The BBC explains:

The first assumed that it is difficult for life to be formed but easy for it to evolve, and suggested there were 361 intelligent civilisations in the galaxy.

A second scenario assumed life was easily formed but struggled to develop intelligence. Under these conditions, 31,513 other forms of life were estimated to exist.

The final scenario examined the possibility that life could be passed from one planet to another during asteroid collisions – a popular theory for how life arose here on Earth. That approach gave a result of some 37,964 intelligent civilisations in existence.

Forgan’s statistical approach relies on so-called Monte Carlo methods that use repeated random sampling to compute results in complex systems. I note this from the Wikipedia article on these algorithms: “Monte Carlo methods are useful for modeling phenomena with significant uncertainty in inputs, such as the calculation of risk in business.” And nothing would suggest uncertainty in inputs more than the classic Drake Equation, especially in terms of the biological factors relating to how life develops and turns into civilizations.

Problems with Exoplanet Modeling

And, of course, there is still much we don’t know about exoplanets, considering that we have yet to observe a single terrestrial world around another star. But in our simulations we can use existing exoplanet data to provide a distribution of planetary parameters. These can then be applied in sampling, including factors such as the planetary mass function, the distribution of planetary orbital radii, and the metallicity of the host stars.

Note what Forgan has to say about this challenge as he threads his way through this statistical analysis of Drake:

…as with the stellar parameters, a population of planets can be created around the parent stars, with statistical properties matching what can be observed. However, this statistical data is still subject to strong observational bias, and the catalogues are still strongly incomplete. There is insufficient data to reproduce a distribution of terrestrial planets: therefore it is assumed that life evolves around the satellites of the planets simulated here. In essence, this constitutes a lower limit on the number of inhabited planets: the work of Ida and Lin… shows that, as a function of metallicity, habitable terrestrial planets are comparable in frequency (or higher) than currently detectable giant planets. This data is hence still useful for illustrating the efficacy of the Monte Carlo method (at least, until observations of terrestrial exoplanets become statistically viable). All this should be borne in mind when the results of this work are considered.

The ‘Single Biosphere’ Issue

And in biological terms, we are even more up the creek, since we base our thinking on observations of a single biosphere, our own. To keep the number of free parameters to a minimum, Forgan works with “a biological version of the Copernican Principle,” the notion that our Terran biosphere is not special or unique, so that we can think about life on other worlds as sharing many of the same characteristic parameters. The thinking on these matters — and the statistical methods used to explore them — forms the most absorbing section of the paper.

About half of all emerging civilizations destroy themselves under two of Forgan’s three hypotheses, the panspermia approach and what he calls the ‘rare life hypothesis,’ while self-destruction is a bit more likely still in the ‘tortoise and hare hypothesis,’ where life evolves easily but evolution toward intelligence is more difficult. We also have to factor in possible ‘reset’ events that may annihilate a biosphere before a civilization can emerge, and make estimates on the amount of time needed between life’s appearance and the first civilization.

There is so much we simply don’t know. But Forgan is no ideologue. He knows he’s working with outputs that are only as accurate as their inputs will allow. Thus this, on planetary modeling:

Current data on exoplanets, while improving daily, is still insufficient to explore the parameter space in mass and orbital radii, and as such all results here are very much incomplete. Conversely, as observations improve and catalogues attain higher completeness, the efficacy of the Monte Carlo Realisation method improves also. Future studies will also consider planetary parameters which are sampled as to match current planet formation theory, rather than current observations.

Where Statistical Analysis Is Taking Us

How much work remains to be done on the ever complicated Drake Equation? Quite a bit, in the opinion of this researcher. Forgan notes that we need an improved three-dimensional galaxy model that incorporates the evolution of the Milky Way over time and takes into account its components, such as the bulge and the bar. We also will continue to plug in better models for star formation and the spatial distribution of stars. New input from our space-based observatories should gradually strengthen our statistical models as we tune what Frank Drake started into an ever more sophisticated instrument for SETI research.

The paper is Forgan, “A Numerical Testbed for Hypotheses of Extraterrestrial Life and Intelligence,” published online by the International Journal of Astrobiology (January 23, 2009) and available here.