The 0.38 g Question

The gap

Two Data Points, and a 38% Answer Between Them

The Moon at 0.165 g and Mars at 0.38 g both sit in territory no human has ever occupied for more than three days.

Long-duration human physiology is well characterized at one g, because that is where everyone lives, and increasingly well characterized at effectively zero g, because Mir and the International Space Station have accumulated decades of continuous occupancy. Between those two points there is nothing. The Apollo surface stays were up to three days at 0.165 g, which is long enough to measure nothing chronic.

One small biological exception exists and is worth naming precisely because it is so small. JAXA has flown rodent habitats aboard the station with onboard centrifuges capable of partial gravity, and results at lunar-equivalent levels suggest that partial gravity gives partial protection — more than weightlessness, less than one g. That is a handful of mice, for weeks, at one gravity level, in one species. It is the only partial-gravity biological dataset in existence, and no human, and no full life cycle, appears anywhere in it.

The shape of the curve

Is the Response Linear in Gravity, or Is There a Threshold?

Two plausible curves fit the two points, and they give opposite answers about Mars.

Everyone modeling this problem draws a line between the two known points, and there is no evidence the underlying relationship is a line. Two hypotheses fit the data equally well.

The linear hypothesis: physiological response scales with gravitational load, so Mars at 0.38 g delivers roughly 38% of the protection one g provides. Bone and muscle loss proceed at reduced but significant rates, countermeasures remain mandatory, careers are limited, and a permanently Mars-resident population is permanently deconditioned relative to Earth.

The threshold hypothesis: most of the biological signaling that drives adaptation is triggered by mechanical loading above some modest level, so any meaningful gravity captures the bulk of the benefit. Under this reading 0.38 g is nearly as good as one g for most systems, Mars residency is close to physiologically normal, and the entire problem shrinks to a manageable countermeasure question.

The two readings differ by enough to change program decisions in opposite directions, and the difference between them is not a matter of opinion or modeling sophistication. It is a missing measurement, and it is the only kind of question that cannot be resolved by more analysis.

Systems at risk

What Weightlessness Does, and What Partial Gravity Might Not Prevent

The microgravity effects are the catalogue; the open question is which ones 0.38 g suppresses.

The column on the right is the page in miniature. Every entry reads “unknown”, and the reasoning offered for each is mechanism-based speculation rather than measurement. Some of that speculation is well grounded — fluid shift really should scale with gravity — but a settlement is not built on a mechanism argument.

One interaction deserves separate mention because it is rarely modeled. On Mars, partial gravity and chronic radiation act on the same tissues at the same time over years, and the terrestrial and orbital literature treats them as separate hazards studied separately. Whether they combine additively, or worse, is unexamined.

The ring

A Partial-Gravity Ring Is a Modest Piece of Engineering

18 m of radius at about 4 rpm produces Martian gravity. The hard part is not the physics.

Centripetal acceleration is radius times angular rate squared, so Martian gravity at an 18 m radius needs roughly 4.3 revolutions per minute. That rate sits near the comfort boundary suggested by terrestrial rotating-room studies, which is why the radius matters: halving it to 9 m would require about 6.1 rpm, where cross-coupled head movements produce significant illusory motion and nausea until the subject adapts.

Two artifacts come with any rotating habitat and both scale with radius. Coriolis acceleration deflects anything moving relative to the floor, which affects reaching, pouring and walking in ways that are noticeable rather than disabling. And the gravity gradient across a standing person — head to foot — is about 11% of the local value at 18 m, against 22% at 9 m. A larger ring is a gentler ring, and 18 m is roughly where the artifacts stop dominating the experiment.

What makes it a real program rather than a sketch is not the structure but the qualification. A spinning habitat has no graceful degradation: between operating normally and tumbling there is no intermediate state. Bearings, seals, despin control, mass balance as crew move around, and the response to a jammed rotation joint all have to be demonstrated, which means the vehicle must be spun up, deliberately failed, and recovered before anyone builds a settlement on the answer. That qualification requirement, rather than the ring itself, is what the outer-system architectures are actually waiting on.

Two experiments

These Are Two Experiments, Not One, and the Smaller One Matters More

The adult physiology ring answers more questions. The developmental ring answers the one that gates the settlement.

The second cannot ethically be run on humans first, which is what makes it an animal study by necessity rather than by convenience. It is also years long by construction, because a life cycle is a life cycle — conception, gestation, birth, growth to maturity, and ideally a second generation.

The recommendation is to rank the developmental ring first, even though the physiology ring answers more questions, because it is the long pole. An adult physiology result that arrives late delays crew planning. A developmental result that arrives late delays the settlement itself, and no amount of funding afterward compresses a life cycle.

Reproduction

Every Stage of Development Is Gravity-Sensitive on Earth

And the first child born on Mars is an experiment whose subject cannot consent.

Terrestrial development uses gravity as an input at several stages. The vestibular system calibrates against a gravity vector during a critical window. Skeletal architecture develops in response to mechanical loading. Cardiovascular regulation assumes a hydrostatic column. Rodent work in orbit has shown that pups gestated in weightlessness exhibit vestibular abnormalities that resolve only partially, and no mammal has ever completed a full life cycle at any gravity level other than one.

A child conceived, gestated and raised at 0.38 g is therefore not a smaller version of a terrestrial child with known deficits. It is a developmental trajectory nobody has observed at any scale, in any species, and the result is permanent for that individual. It is also the point at which the settlement stops being reversible in the sense discussed on Settlement & Population — a resident who cannot adapt to one g cannot be repatriated.

That is the reason the developmental ring is on the settlement’s critical path rather than on its wish list. A program that intends permanent habitation and has not run that experiment is proposing to run it on a child instead.

Countermeasures

Some Answers Do Not Require the Measurement

A surface centrifuge hedges the whole question, and it is far cheaper than the ring that would answer it.

Three interventions work regardless of where the true curve sits. Resistive exercise is already the workhorse: the station’s resistance device preserves most bone and muscle at zero g, and it should perform at least as well at 0.38 g, so a Mars habitat carries one whatever the physiology turns out to be. Pharmacological support — bone-density agents in particular — is mature terrestrial medicine and transfers directly.

The interesting one is a short-radius centrifuge inside the habitat: a spinning bed or exercise station delivering an hour or two of Earth-equivalent loading a day. At 2 m radius and roughly 21 rpm it produces one g at the feet, it fits inside a buried hull, and it costs a few tonnes rather than a station program. If the threshold hypothesis is right it is unnecessary; if the linear hypothesis is right it is the difference between a deconditioned population and a healthy one. It is the cheapest hedge available against the largest unknown in the settlement case, and it appears in no published Mars architecture.

None of this substitutes for the measurement where reproduction is concerned. A centrifuge protects adults who choose to use it. It does not answer whether a fetus develops normally at 0.38 g.

Waiting

Six Analyses in This Reference Depend on the Answer

Which is unusual — most open questions affect one thing.

The last row is the sharpest illustration. Venus’s cloud layer offers close to Earth gravity, which is either a decisive advantage over Mars or an irrelevance — and which one it is depends entirely on a measurement nobody has taken. A single experiment moves the relative ranking of two destinations.

Status

Nobody Is Building Either Ring

The station’s centrifuge module was cancelled 20 years ago and never replaced.

The International Space Station was designed with a Centrifuge Accommodations Module intended to provide variable artificial gravity for biological research. It was cancelled in 2005 during budget restructuring, the hardware was never flown, and no replacement has been manifested since. JAXA’s small rodent centrifuges are the surviving remnant of that capability, and they operate at the scale of a laboratory animal habitat rather than a research facility.

Several commercial station developers have discussed artificial-gravity demonstrations as future capability, and none is funded as a partial-gravity physiology program with human subjects. A question that six independent analyses depend on, that no amount of modeling can resolve, and that gates the settlement case for an entire planet, has no program anywhere.

It is also cheap relative to what it unblocks. An 18 m ring in low Earth orbit is a small fraction of a crewed Mars program, and it can be built with existing structures, existing life support and existing launch capability. The reason it does not exist is not technical difficulty. It is that no single program owns the question.

Record

Provenance and Data Sources