Like many cities around the world, Austin is now facing questions about how to build and adapt for a changing climate. A growing number of these cities—as well as insurance companies, home builders, and farmers—are turning to climate modelers for answers. But despite decades of effort, forecasting how global warming will play out on a local scale remains a stubborn challenge, riven with uncertainty.

There is little agreement on how best to convert climate models, which simulate the entire world at coarse resolutions, into the detailed local forecasts of temperature and rainfall that planners crave. Different methods lead to drastically different projections, especially in terms of rainfall—even when using the same climate model. “It really is a mess right now,” says Alex Hall, a climate scientist at the University of California, Los Angeles.

The problem has become more pronounced with the discovery that global climate models, good at the big picture, often miss local impacts that are already painfully evident. For example, despite nailing the overall pace and intensity of global warming, models did not predict the more frequent heat waves that are plaguing Europe.

In order to make local forecasts, modelers must not only do the global-to-local conversion, but also adjust for these model biases. Tiffany Shaw, a climate dynamicist at the University of Chicago, calls this “the other climate crisis.” “The cat’s going to get out of the bag, and the question will be, what value does this information have?”

While scientists fret about the limitations of their models, others are not as cautious. Dozens of firms have sprung up selling climate risk products, often promising unrealistic levels of detail, down to a city block. “It’s unfortunately closer to a Wild West scenario right now,” says David Lafferty, a climate scientist at Cornell University. “There is a real need to evaluate those products.”

https://archive.ph/AD4rV