#88 - Qubits and Ocean Circulation

I came across a paper this week titled “Algorithm for Simulating Ocean Circulation on a Quantum Computer” by Ruimin Shang et al. (Science China Earth Sciences, Vol. 66 No. 10, 2023; DOI: 10.1007/s11430-023-1162-x).

It’s a heavy read (my monkey brain can't digest) dense with equations and quantum terminology, but it caught my eye because it bridges two worlds that rarely meet: marine science and quantum computing (both fascinate me).

I thought I’d use a bit of tech help to break it down into layman’s English, partly because I’m fascinated by where quantum (and its more approachable cousin, the probabilistic “p-bit”) might eventually show up in real-world applications.

It’s still early-stage and mostly theoretical, but interesting nonetheless to see how ideas like this could one day help us better model, understand, or even protect our oceans. Enjoy!

Why even bother simulating the ocean?

If you’ve ever looked at a weather forecast, a sea surface temperature map, or even a shipping route prediction, you’ve already benefited from ocean simulation. These models are the invisible engines behind much of modern climate and marine science.

But simulating the ocean isn’t like rendering a map. It’s about representing how heat, salinity, currents, and pressure all interact in three(+) dimensions, over time, across an entire planet (very hard to accurately model, lot of assumptions). Each of those variables depends on the others, and small changes ripple through the system. That’s what makes it both fascinating and frustrating to model.

Here are some uses and why it’s such a big deal:

• Climate forecasting: Ocean circulation stores and redistributes heat and carbon. If you can simulate that accurately, you can better predict how climate change will unfold region by region.

• Weather prediction: The ocean drives atmospheric patterns. Better ocean models feed into better forecasts for storms, rainfall, and extreme weather.

• Marine operations: Offshore energy, aquaculture, and shipping rely on local current models to plan routes, manage risk, and optimise operations.

• Environmental response: Simulations underpin decisions on oil spills, pollution tracking, and habitat restoration, essentially predicting where the ocean will “move” next.

The problem? These systems are massive. To capture global circulation, scientists divide the ocean into billions of small 3D cells and compute how each interacts with its neighbours, second by second.

That means solving an astronomical number of equations, something even today’s supercomputers struggle to do at high resolution or in real time.

This is where quantum computing becomes interesting. In theory, it could help process those calculations not by brute force, but by exploiting probabilities and parallel states, cutting through the complexity in entirely new ways.

In short, simulating the ocean matters because it underpins nearly everything we know (and need to know) about the planet’s future. And the search for faster, smarter ways to do it, whether through classical AI, hybrid models, or now, quantum algorithms, is really a search for understanding Earth itself.

What the researchers actually did

The team behind the paper: Ruimin Shang, Zhimin Wang, Shangshang Shi and colleagues at the Ocean University of China, set out to test whether a quantum computer could handle one of the toughest parts of ocean modelling: solving the giant webs of equations that describe how water moves.

Here’s the gist, minus the maths:

• They started with the real physics.

  The researchers took the primitive equations, the standard set used in most global circulation models, and split the ocean into a 3D grid. Each grid cell interacts with its neighbours, producing a cascade of relationships

  between temperature, salinity, pressure and motion.

  
  

• They translated those relationships into algebra.

  Once the equations were discretised (broken into steps of space and time), they became linear systems — large but structured sets of equations of the form A × x = b. These are the workhorses of computational fluid dynamics and the main reason models are so expensive to run.

  
  

• They handed that step to a quantum algorithm.

  Instead of letting a classical solver grind through those systems, they used something called the Variational Quantum Linear Solver (VQLS).

  
  

• Think of it as a hybrid approach: the quantum chip proposes an answer, the classical computer checks and fine-tunes it.

  
  

• It’s built for the noisy, limited quantum hardware that exists today rather than the fault-tolerant machines still on the horizon.

  
  

• They designed a data-exchange bridge between the two worlds.

  Because classical and quantum machines speak very different languages, they proposed a way to move data in and out using quantum random-access memory (qRAM) ideas and a tomography trick to read quantum states efficiently. These are theoretical constructs for now, but useful for sketching what a future workflow could look like.

  
  

• They ran miniature tests.

  The algorithm was trialled in three environments — MATLAB, a quantum simulator, and an actual small-scale quantum device.

• Accuracy dropped when real hardware noise crept in, but error-mitigation methods like zero-noise extrapolation recovered much of it.

In essence, the researchers weren’t building a working ocean model, they were proving a concept: that a quantum-assisted solver can handle the kind of linear-algebra backbone that underpins global circulation models.

It’s early and limited, but it’s a meaningful first step toward blending quantum logic with earth-system science, two fields that, until recently, had almost nothing to do with each other.

What this could mean (someday)

Right now, this research is more of a proof of concept than a practical tool, the kind of experiment that lays groundwork rather than delivers outcomes. But if (and it’s a big if) quantum computing keeps improving, there are some interesting directions this could lead.

• Faster, finer models.

  The main barrier to improving ocean simulations is computing time. If quantum algorithms can one day solve certain types of equations exponentially faster, we could see models that run at far higher resolutions — capturing local eddies, mixing layers, or small-scale dynamics that current systems gloss over.

  
  

• More efficient forecasting.

  Speed doesn’t just mean prettier visualisations — it means faster turnaround for weather, current and climate forecasts. That could translate to real-world advantages: better storm prediction, improved fisheries management, or more precise coastal planning.

  
  

• Hybrid workflows.

  The likely scenario isn’t all-quantum modelling but hybrid setups, where specific heavy tasks (like matrix inversion or data assimilation) are handled by quantum accelerators, while the rest runs on classical supercomputers. This mirrors how GPUs transformed AI — not by replacing CPUs, but by specialising.

  
  

• Cross-domain insights.

  Techniques developed here could ripple outward — into areas like fluid simulation for renewable energy, pollutant dispersion, and offshore engineering. Essentially, any problem that relies on complex, dynamic flow systems could benefit.

Of course, all this depends on quantum hardware catching up. Current devices are small, noisy, and limited by qubit stability. But research like this helps define the problems quantum computing should be aiming to solve once it matures — grounding the hype in practical scientific use.

So while we’re nowhere near a “quantum ocean model” running live forecasts, the idea of ocean physics and quantum algorithms working together isn’t science fiction anymore. It’s early and speculative but it’s a glimpse of what might come when our computing power finally matches the complexity of the sea itself.

Where p-bits fit in

While qubits get most of the attention, there’s a quieter, more practical cousin worth mentioning: the p-bit, or probabilistic bit.

A qubit lives in a quantum state, it can be 0 and 1 at the same time, governed by superposition and entanglement. That makes it powerful, but also fragile, which is why quantum computers need vacuum chambers, extreme cooling, and error-correction just to function.

A p-bit, on the other hand, stays in the classical world. It flips randomly between 0 and 1 based on probability — no superposition required. Think of it as a coin that’s constantly being tossed, but with a bias you can control.

So why do p-bits matter here?

• They’re easier to build.

  P-bits can be made using conventional electronics — magnetic devices, transistors, or memristors — meaning they could slot into existing hardware without the fragility of quantum systems.

  
  

• They can model uncertainty naturally.

  Ocean systems are full of probabilities: turbulent flow, variable salinity, chaotic weather influence. P-bit architectures are good at sampling and optimising within uncertain systems, which might one day make them useful for coarse-grained ocean or climate models that don’t need full quantum horsepower.

  
  

• They bridge the gap.

  You can think of them as a stepping stone between classical and quantum computing — giving us some of the flexibility of quantum logic without requiring quantum hardware.

While the Ocean University study didn’t use p-bits directly, the same logic applies: using new forms of computation to handle complex, non-linear systems that traditional methods struggle with.

In that sense, p-bits represent a “good enough” version of the quantum dream — a way to experiment with probabilistic computation today, while qubits continue their long march from lab to application.

Closing Remarks

I like papers like this not because they promise an immediate breakthrough, but because they show where boundaries might start to blur.

We’re still years away from anyone running a full ocean model on a quantum machine. But the fact that researchers are already sketching how that workflow could look is what makes it exciting. It’s a small but symbolic step toward rethinking how we handle planetary-scale complexity.

What stands out is the mindset behind it: take a hard, practical problem — ocean circulation — and see if a frontier technology can chip away at its limits. Even if quantum never fully takes over, the process of exploring it often sparks fresh approaches on the classical side, too.

And if p-bits or other probabilistic systems become the bridge in between, even better.

At the end of the day, it’s not really about quantum supremacy or exotic hardware. It’s about finding smarter ways to model the one system that keeps everything else alive, the ocean itself.

OTI-H

Reference: Shang, R., Wang, Z., Shi, S., Li, J., Li, Y., & Gu, Y. (2023). “Algorithm for simulating ocean circulation on a quantum computer.” Science China Earth Sciences, 66(10), 2254–2264. https://doi.org/10.1007/s11430-023-1162-x