Evidence of this warm, watery past has been preserved all over the planet in the form of lakebeds, river valleys, and river deltas.

For some time, scientists have been trying to answer a simple question: where did all that water go? Did it escape into space after Mars lost its atmosphere, or retreat somewhere?

According to new research from Caltech and the NASA Jet Propulsion Laboratory (JPL), between 30% and 90% of Mars’ water went underground. These findings contradict the widely-accepted theory that Mars lost its water to space over the course of eons.

Until recently, scientists theorized that atmospheric escape was the key, where water is chemically disassociated and then lost to space.

This process is known as photodissociation, where exposure to solar radiation breaks down water molecules into hydrogen and oxygen.

At this point, the theory goes, Mars’ low gravity allowed for it to be stripped from the atmosphere by solar wind. While this mechanism is sure to have played a role, scientists have concluded that it cannot account for the majority of Mars’ lost water.

The team analyzed data from Martian meteorites, rover, and orbiter missions to determine how the ratio of deuterium to hydrogen (D/H) changed over time. They also analyzed the composition of Mars’ atmosphere and crust today, which allowed them to place constraints on how much water existed on Mars over time.

Deuterium (“heavy water”) is a stable isotope of hydrogen that has both a proton and neutron in its nucleus, whereas normal hydrogen (protium) is made up of a single proton orbited by one electron. This heavier isotope accounts for a tiny fraction of hydrogen in the known Universe (about 0.02%) and has a harder time breaking free of a planet’s gravity and escaping into space.

Because of this, the loss of a planet’s water to space would leave a telltale signature in the atmosphere in the form of a larger-than-normal level of deuterium. However, this is inconsistent with the observed ratio of deuterium to protium in Mars’ atmosphere, hence why Scheller and her colleagues propose that much of the water was absorbed by minerals in the planet’s crust.

 Since Earth is tectonically active, hydrated minerals are endlessly cycled between the mantle and the atmosphere (through volcanism). 

But since Mars is tectonically inactive (for the most part), its surface water was sequestered early on and never cycled back out. Thus, the features that indicate the past presence of water were preserved by the permanent drying of the surface. Meanwhile, a significant portion of that water was preserved by becoming absorbed beneath the surface.

This study not only addresses the question of how Mars’ water disappeared billions of years ago. It could also be good news for future crewed missions to Mars, which will depend on locally-harvested ice and water.