Our brains are scary good at finding patterns where none actually exist. It’s why we see a face on the surface of Mars, and it’s the basis for the cloud gazing game many of us played as kids. The shifting nature of condensing water vapor in the sky makes perfect fodder for finding silhouettes of passing pirate ships or rampaging dinosaurs, if you’re willing to let your imagination loose. It’s even good inspiration for SNL skits.
Cloud gazing has very likely been a favored pastime for as long as humans have been around and looking up. In most climates, clouds can be counted on to streak slowly across the heavens, or else gather ominously in advance of a storm. As climate change progresses, however, that might change. While humans are great at finding patterns in individual clouds, understanding their long-term patterns and their contribution to the larger climate isn’t quite as easy.
Getting a decent picture of the world at large is comparatively simple, while getting the granular data of shifting clouds and other high-definition, ever-shifting processes is difficult. When it comes to modeling the climate, it seems we can either get the big picture or high definition on a small scale, but not both. To solve the cloud puzzle we’re going to need new tools.
Researchers from the University of California, Irvine and colleagues are working to build new climate models which will give us both the big and small pictures we’re going to need in order to more accurately predict how climate will change in the future. Their findings were published in the Journal of Advances in Modeling Earth Systems.
Anthropogenic climate change is drastically changing the face of our planet and our ability to live comfortably on it. Seas are rising, ecosystems are shifting, and species are changing or disappearing. Above it all, the clouds continue to drift idly by, but that might not always be the case. Climate change will very likely leave its mark on the sky, we’re just not sure how.
As average global temperatures continue to rise, it’s likely that the formation of clouds will be impacted. They could shrivel and disappear, or they could become even denser and more abundant. Which of these two scenarios is borne out is of critical importance for understanding how climate change will evolve in the future. Fewer or thinner clouds would mean more sunlight reaching the surface of the planet and, potentially, more rapid heating. More and denser clouds would block sunlight, be more reflective, and have the opposite effect of delaying some of the effects of climate change.
The problem is that accurately reproducing cloud formation in our climate models is currently beyond the ability of even our most powerful computers. Instead, climate scientists approximate the presence and shifting behavior of clouds to make best guesses. Our current models have a resolution — you can think of it as pixels on a massive planetary screen — of about 4 kilometers. That means that for anything happening on scales of less than 4 kilometers, the data just isn’t there. When you back up and look at the model as a whole, it looks pretty good, but we’re missing a lot of the granular detail which feeds that picture.
In order to get the data we would really need in order to know how clouds will evolve and contribute to future climate change, we’d need to have a resolution of about 100 meters, about 40 times more definition than we currently have. Taking current technological trends as a guide, we’ll eventually get there, but it might be too late for us to make good use of the data by then.
The process outlined in this new paper involves running two separate models and getting them to talk to one another. They start with a low-resolution model which looks at the planet with 100-kilometer resolution to get the broad strokes. A second model builds patches of information at 100-to-200-meter resolution. This allows scientists to capture the complex mechanics at work in cloud formation. The two models then trade information every 30 minutes to make sure neither of them has drifted too far from reality and course correct.
This lets a supercomputer use its limited resources more efficiently by splitting the cost between the large and small, gaining the best of both worlds.
The fate of the clouds and, consequently, the rest of our planet remains unclear, but now scientists have additional tools for glimpsing into the future. Hopefully the clouds will part — or thicken — before it’s too late.