It's an odd thing to consider – but we know less about the oceans beneath the Antarctic ice shelves than we do about the surface of Mars.
Covering just under half of the continent's coastline, these shelves hide the least-measured ocean water on Earth, and few observations have ever been taken in the pristine waters that lie beneath them.
There is an obvious reason for this: these giant floating glaciers are hundreds of metres thick, making it extremely hard to get under them.
At the same time, the need to do so is growing increasingly urgent.
With climate change warming the planet, measurements of ocean temperatures are a crucial part of the bigger picture of climate research.
Almost all of the excess heat we've pumped into the atmosphere – 96 per cent - has gone into the world's oceans.
Of this, the Southern Ocean has absorbed more than half – acting as a vast, cold but ultimately temporary buffer against rising land surface temperatures.
"The most challenging problem facing Southern Ocean oceanographers and engineers is a lack of instruments specifically designed to make measurements of deep ocean properties beneath floating glaciers and ice shelves," Otago University's Dr Inga Smith explained.
These sensors are designed for broad use in temperate and tropical waters – but not for the extreme conditions of Antarctica.
That's where a fascinating innovation, being developed by Smith and colleagues, will come in.
A new study to take place in the McMurdo Sound over the summer of 2020/21 specifically aims to understand how to overcome two features of the ice shelves - small, free-floating ice crystals known as frazil, and what's called "supercooled water".
Beneath the Antarctic sea ice and ice shelves, the water is often colder than its freezing point temperature - yet manages to stay in liquid form.
Smith wants to learn more about frazil, and the snap-freezing of super-cooled water, as both are stopping scientists from getting high-precision measurements of some of the key ocean parameters needed for climate models.
With collaborators Dr Britney Schmidt from Georgia Tech in the United States, Professor Lars Smedsrud from the University of Bergen in Norway, and Dr Greg Leonard from the University of Otago, Smith's team will design a completely new tool they've dubbed the High Precision Supercooling Measurement Instrument, or HiPSMI.
This pumped system, fitted with a range of sensors, will take measurements very close to the interface of where ice met water, and capture factors like salinity and temperature to calculate supercooling.
Smith said the HiPSMI would then be installed into an underwater robot called the Icefin.
"By pushing ocean engineering to extreme limits, we will determine the influence of frazil crystals on measurements of in situ supercooling," she said.
"The measurements, in conjunction with numerical modelling and laboratory work, will revolutionise our understanding of supercooled waters by providing a high-precision, observational-based indicator for future climate observations beneath the vast cold cavity ice shelves of Antarctica."
Smith said the research, being supported with a $954,000 grant from the Marsden Fund, might also help overcome engineering challenges of the next frontier of polar exploration: ice-covered oceans on other worlds.