Global Warming: How Quantum Computers Could Improve Climate Predictions
Book: Quantum Supremacy: How the Quantum Computer Revolution Will Change Everything Author: Michio Kaku ISBN: 978-0385548366
Chapter 14 opens with Kaku visiting a university in Reykjavik, Iceland. He tours their ice core research lab, basically a giant freezer room with long metal rods containing ice samples drilled from deep underground. Some of this ice fell as snow thousands of years ago. Inside the ice, there are microscopic air bubbles, snapshots of the ancient atmosphere. Scientists measure CO2 levels from those bubbles and calculate temperature using the ratio of heavy water molecules to normal ones.
When you plot temperature and CO2 content over centuries on the same graph, they track each other almost perfectly. Two roller coasters going up and down together.
The Spike That Should Worry Everyone
Looking at these ice core records, temperature has been slowly rising since the last ice age ended about 10,000 years ago. Then there is a sudden spike in the last 100 years, right when the Industrial Revolution kicked in and we started burning fossil fuels at scale. The years 2016 and 2020 were the hottest ever recorded. The period from 1983 to 2012 was the hottest thirty-year stretch in the last 1,400 years.
NASA satellites confirmed this from a different angle. They measured energy coming in from the sun and energy going back out from Earth. If things were in balance, those numbers would match. They don’t. Earth absorbs more energy than it radiates back, and the difference roughly equals the energy generated by human activity.
Kaku mentions that submarine visits to the North Pole since the 1950s show Arctic ice has thinned by 50 percent during winter months. According to NASA, the Arctic Ocean will be completely ice-free in summer by mid-century.
Methane and the Feedback Loop
CO2 gets most of the attention, but Kaku points out that methane is over thirty times more potent as a greenhouse gas. The scary part: vast stretches of tundra in Canada and Russia are thawing out, releasing trapped methane. The more Earth heats up, the more tundra melts. The more tundra melts, the more methane releases. The more methane releases, the more Earth heats up.
A positive feedback loop. Many current computer projections might actually underestimate the real magnitude of global warming.
Kaku shares a detail from his lecture in Krasnoyarsk, Siberia. Locals told him they didn’t mind the warmer temperatures. They also mentioned mammoth carcasses from tens of thousands of years ago emerging from the melting ice.
Military Implications and the Polar Vortex
The Pentagon once drafted a worst-case scenario for uncontrolled global warming. One of the deadliest hotspots: the border between Bangladesh and India. Sea level rise and flooding could force millions to rush the Indian border. The scenario ends with India potentially using nuclear weapons against waves of climate refugees. Extreme, but it shows military planners take this seriously.
Kaku also explains something useful about the polar vortex. People who point to massive winter storms as evidence against global warming are actually seeing a consequence of it. The polar vortex is a spinning cylinder of super-cold air centered on the North Pole. As the Arctic warms faster than other regions, the temperature difference between the pole and lower latitudes shrinks. This weakens the vortex, making it unstable. It wanders south, pushing the jet stream with it, and you get freezing weather in Texas. Global warming can produce extreme cold snaps. Counterintuitive, but the physics makes sense.
What Can We Do About It?
Kaku runs through six approaches. Most are geoengineering options for worst-case scenarios.
Carbon sequestration separates CO2 at refineries and buries it underground. Works in theory, but the economics are not there yet. Companies are in wait-and-see mode.
Weather modification uses volcanic ash or particles to reflect sunlight. Mount St. Helens showed this works, but the scale needed is enormous and testing is basically impossible.
Algae blooms could absorb CO2 if you seed oceans with iron. Algae reproduces in unpredictable ways though. You can’t recall a life form like you recall a faulty product.
Cloud brightening seeds clouds to reflect more sunlight. Weather modification is very local though and the track record is not great.
Planting trees or genetically altering plants to absorb more CO2 is the safest approach. Not enough at planetary scale though, and it requires coordination across many countries.
Computing virtual weather is where quantum computers come in.
Where Quantum Computers Fit
All climate models start by dividing Earth’s surface into grid cells. In the 1990s, each cell was about 311 miles on a side. By 2007 for the IPCC Fourth Assessment Report, grid size was down to 68 miles. These grids extend into the third dimension, slicing the atmosphere into about ten vertical layers. For each slab, the computer calculates humidity, temperature, pressure, sunlight, and how they change across neighboring cells.
The problem is data volume. Classical computers are already struggling with this. And there are big uncertainties that make predictions unreliable.
The biggest uncertainty is clouds. Up to 70 percent of Earth’s surface is covered by clouds at any given time, and cloud formation changes minute by minute. The jet stream is another wildcard. Its path is hard to predict, so meteorologists use rough averages.
You see this when hurricane predictions appear on TV. Different computer models show different paths, sometimes varying by hundreds of miles. Those hundreds of miles mean the difference between a city evacuating or not.
Quantum computers could help in two concrete ways. First, they can handle much smaller grid sizes, capturing weather changes that happen over just a mile instead of dozens of miles. Second, they can model variable parameters like clouds and jet stream paths instead of using fixed estimates. You could turn knobs on the simulation and see how changing cloud coverage or jet stream behavior affects the forecast.
Kaku also mentions hindcasting, running models backward to “predict” past weather we already know. Current models pass this test for the last fifty years. They are at the limit of what classical computers can do though.
My Take
One of the more grounded chapters in the book. Climate modeling is genuinely a problem where more computing power directly translates to better predictions. We already have the physics and the models. We just don’t have enough computing power to run them at the resolution we need.
The geoengineering section is sobering. None of the options are great. Carbon sequestration is too expensive, weather modification is too unpredictable, biological approaches carry their own risks. The honest answer seems to be renewables, possibly fusion power (which Kaku covers in the next chapter), and much better predictive models so we can plan for what is coming.
As an engineer, the grid cell explanation resonated. Same problem we face in any distributed system. More granularity means more accuracy but exponentially more computation. Quantum computers don’t change the fundamental problem, but they push the boundary of what resolution is achievable. For climate science, that resolution matters.