Mapping The Ocean With Big Data

Creating a digital seafloor with AI and Big Data


It is often said that the sea floor is still 95% unexplored, and that we know more about outerspace than we do the depths of the ocean. This is not entirely true, but nearly.

Nabokov once famously said that the breaking of a wave cannot explain the whole sea. However, scientist David Sandwell and his team at the Scripps Institute of Oceanography in San Diego are doing a good job of proving him wrong. Last year, they published a global map of the seafloor which set out the contours of the entirety of the ocean floor. They used satellites and radar to measure the height of the sea’s surface, subtracting the effects of waves and tides, and looking at bumps and dips in the surface to understand the underlying landscape of the ocean floor. From this, they established where there are underwater mountains or trenches beneath. For example, a large underwater mountain results in a tiny local increase in gravity that comes from its mass, which pulls sea water into a slight bump above it. An ocean trench, on the other hand, is revealed by a dip in the ocean surface which occurs as a result of weaker local gravity.

This all may sound impressive, but the challenges of penetrating sea water with radio waves means that the map can only see features larger than 5km. This may be a marked improvement on the previous global map of the ocean floor from 1997, which could only see those of more than 20km, but still means that just 0.05% of the sea floor has been mapped to a level of detail useful for detecting items such as undersea volcano vents. On the other hand, scientists may have mapped out just 60% of Mars, but it is to a resolution of around 20m, while 98% of Venus has been mapped out to a resolution of around 100m.

The National ICT Australia (NICTA) is also making strides in better understanding the seafloor. They have collaborated with the University of Sydney to try and better understand its geology. They used Big Data and Artificial Intelligence (AI) to create a computer algorithm that converted 15,000 seafloor sediment observations into a continuous digital map. It is the first time anyone has built up a composition of the seafloor since the 1970s, and scientists can now use this map to gain a better understanding of how our oceans respond to environmental change.

One of the most important discoveries from this research so far relates to what is preserved in the sea floor. It largely consists of the remains of microscopic sea creatures called phytoplankton, a subset of which - called ‘diatoms’ - produce roughly a quarter of the oxygen we breathe. When it dies, it sinks to the bottom, where it’s carbon is locked away. Diatoms actually make a greater contribution to fighting global warming than most plants on land. The digital map found that diatom accumulations on the seafloor are nearly entirely independent of diatom blooms in surface waters in the Southern Ocean. This indicates that we understand the carbon source, but not the sink.

Dr. Simon O’Callaghan, lead researcher at NICTA, notes that: ‘Big data technologies, paired with old and new data, now allow us to better predict the impact changing climate and ecosystems may have on the life cycle of phytoplankton, by jointly analysing a multitude of sea surface and seafloor observations. Now that Australia has a brand-new marine research vessel, the Investigator, the new seafloor data set opens the door to future voyages aimed at better understanding the workings and history of the marine carbon cycle.’

The seafloor holds within it a wealth of information that could be leveraged to save the environment. While the idea of exploring space may be more glamorous than a deep sea dive, the importance of the insights that could be garnered from its composition should not be underestimated.

University lecture small

Read next:

How Are Higher Education Institutions Using Analytics?