Climate change is complicated. Every part of our daily lives can play a role in causing it, from electricity, to transportation, the homes we live in, the food we eat, even the healthcare services we rely on. And all of those aspects of our lives are also affected by climate change.
Much of what we know about the impacts of climate change comes from sophisticated computer models. Now, a group of computer scientists is calling on their colleagues to put advanced computing and artificial intelligence to work to solve the climate problem.
“A.I. can help pinpoint where deforestation is happening using satellite imagery or aerial imagery,” said David Rolnick, lead author of a new study outlining how artificial intelligence could help with climate change. “Or [we can] gather data on where buildings and infrastructure are, which is essential to policymakers in developing appropriate policies around the world.”
Artificial intelligence could also assess damage after disasters. It could accelerate the process of scientific discovery by helping suggest new materials to use for batteries or carbon capture technology. It could also help improve efficiency in systems, such as consolidating freight, helping design carbon markets, or reducing food waste, he said.
Rolnick, who is a National Science Foundation Mathematical Sciences Postdoctoral Research Fellow at the University of Pennsylvania, said artificial intelligence could also help with “precision agriculture.”
“It's not generally appreciated just how much carbon emissions come from agriculture,” Rolnick said.
A.I. could help by helping determine exactly when an individual plant need fertilizer or water, thereby reducing unnecessary treatments on the field.
Despite all the promise, Rolnick warns that we shouldn’t expect a computer to be sitting at the table at the next international climate negotiations.
“I think that the most valuable thing that A.I. can do in that situation is to provide the necessary data for policymakers to make informed decisions,” he said. “So, for example, providing data on infrastructure, providing data on how disasters will affect parts of that infrastructure and different parts of the world, and providing ways that policymakers can potentially improve the systems that exist using the efficiency gains that A.I. can, in some cases, recommend.”