Boffins reckon Mars colony could survive with fewer than two dozen people

Taming that unforgiving dust world may be significantly less expensive than anticipated

When humans head off to colonize Mars, it's not entirely clear how many people will be needed to keep everyone alive.

SpaceX president and chief operating officer Gwynne Shotwell last year speculated that humans could land on the Red Planet within this decade, so it's not too soon to start thinking about the feasibility of a Mars colony or how many people would need to be present to keep the settlement functioning. No one wants to re-enact the desperation depicted in 2015 sci-fi flick The Martian.

While the technology necessary to make Mars habitation plausible remains an ongoing project, some have already done the math. In a 2020 research paper published in Nature, Jean-Marc Salotti, professor of computer science at Ecole Nationale Superieure de Cognitique, Institut Polytechnique de Bordeaux in France, concluded that 110 humans would be required for a self-sustaining colony on Mars.

There have been other estimates. A 2003 study put the low-end colony size at 100. A separate 2001 paper suggested a self-sustaining colony of 500 – not cited as a minimum – could work on the North Polar Cap of Mars. But the 110 figure is more recent.

That may be more people than necessary. A group of researchers from George Mason University in Fairfax, Virginia, has concluded that a Mars colony could get by with just 22 colonists.

The data scientists – Edgar Arguello, Sam Carter, Cristina Grieg, Michael Hammer, Chris Prather, Clark Petri, and Anamaria Berea – describe their findings in a preprint paper titled "An Exploration of Mars Colonization with Agent-Based Modeling."

"We started our study after we saw [Salotti's] paper and we wanted to verify that number," said Anamaria Berea, associate professor of computational and data sciences at George Mason University and one of the co-authors of the research, in an email to The Register.

Berea said the Nature paper made assumptions that didn't take into account the reality of social and psychological behavior and the continuity of human interactions, even at a distance in space.

"We tend to often treat humans as just numbers or particles devoid of personal incentives, heterogeneity and adaptability," said Berea. "Human groups are complex systems where the outcome is not the sum of its parts, but synergistic. Any social system exhibits properties of adaptability, emergence, and non-linear dynamics."

"Additionally, what we included in our model – and it was not included in the other paper – is the relationship between Earth and the habitat on Mars. It is hard to imagine a habitat that will be completely cut off from Earth supplies and independent from the get-go, especially in an environment that is so unsustainable for human life as Mars, even if we would have the best technology to help us there."

"The scenario where you can simply send X number of people somewhere and just let them survive is very unlikely, also because it actually costs much more to send tens and hundreds of people into space than sending resupply shuttles."

In an email to The Register, Salotti wrote, "Twenty-two is in fact compatible with my 110 as different problems are addressed. Twenty-two is acceptable when shipments from Earth are possible for resupply. It might even be lower. The problem with this kind of approach is that the results highly depend on the set of parameters, which are arbitrary."

Berea and her co-authors made their Mars colony population calculations using agent-based modeling in an open source software application called NetLogo.

"Agent-based modeling (ABM) is a class of computer simulations that encode 'agents'/individuals with the relevant attributes similar to the individuals in real life, and also the rules of interactions between these agents," explained Berea.

"In general, with ABMs we create simulations that are representative of a complex system and phenomenon from reality, where we can track the emergence of large patterns or phenomena out of simple rules of behaviors and interactions at the micro level."

"Additionally, we can explore a wide range of scenarios and we can isolate relevant initial conditions and parameters that are more likely to render the phenomena we observe in reality. This is especially useful for systems where direct causal phenomena is not easily deduced from data or experiments, and for systems where we have interactions between agents and various environments."

Using past research about how crew perform under stress (e.g. submarine crews, Antarctic exploration teams, soldiers in Iraq), the boffins created data models for Martian settlers (e.g. coping capacity, resilience, skill level, metabolism, stress level, etc.) and assorted environmental variables and fed them into the NetLogo software to create a simulation. The result is something like The Sims but without the fun graphics.

Each Martian colonist was assigned one of four possible psychological traits: neurotic, reactive, social, or agreeable, and then simulation was run. The modeled Martians would move, sleep, work with others, produce or consume resources, and interact. Without sufficient resources, the Sim Martians lose health and those reduced to zero health die and get removed from the simulation.

Every 78 weeks, a shipment from Earth may be received bringing four new colonists, each with their own personality traits, unless an accident occurs in transit (creating stress for the colonists).

One of the findings of running this ABM is that mission planners may want to minimize the number of neurotics sent to Mars. "Martians with the 'neurotic' psychology die at a much higher rate than those of other psychologies, and once their population reaches a low enough level, the settlement population stabilizes," the paper observes, and notes that those with the agreeable personality type fare the best over time.

Based on five runs of the model for a period of 28 years – seeded with initial population sizes ranging from 10 to 170 in steps of ten – "an initial population of 22 was the minimum required to maintain a viable colony size over the long run," the paper says.

"We wanted to show that if we neglect the social, behavioral and psychological aspects of space explorations, we can err grossly in our estimations, predictions and projections," said Berea.

"Astronauts or regular humans that will explore space in one way or another are not amorphous entities, and our sociality and psychology will have a crucial role for these missions. Our social and behavioral aspects as humans can literally cut the line between the success and failure of a mission, especially a long term one, and can also be reflected in the economic costs of such missions." ®

More about


Send us news

Other stories you might like