(Bloomberg) -- How much does climate change actually cost? It’s one of the more important questions, and also one of the more difficult ones. The reason often hinges on one word: attribution.

It’s one thing for science to conclude that global warming generally leads to more intense hurricanes, droughts, wildfires, and all sorts of other extreme weather events. There’s no doubt climate change will fuel the upcoming wildfire season in California and a busy summer for Atlantic hurricanes. But it’s another thing to link any single storm to climate change and calculate the cost as a direct consequence.

Attribution science had its coming-out moment with the 2003 European heatwave that killed over 70,000. It’s relatively easy to read temperatures off a thermostat and conclude that the summer was probably the hottest in Europe since at least 1500. How much of it was caused by climate change? A 2004 Nature study concludes with over 90% confidence that human influence was to blame for a doubling of the risk of such a heatwave in 2003, conditions not met in any other year since 1851. That paper, co-authored by Peter Stott, who heads the Climate Monitoring and Attribution team at the U.K. Met Office, is now one of the most cited attribution studies.

Stott’s team inside the British forecasting agency has been churning out rapid-response analyses for any number of extreme climatic events, placing percentages on the likelihood that any one drought, flood, or wildfire was caused by climate change. The ever-growing human influence on the climate, coupled with more data and more sophisticated statistical techniques, now allows for better and crisper conclusions. The 2010 Moscow heatwave? Climate change has increased the probability of such record heats fivefold, which translates into an 80% probability that the heatwave would not have occurred without climate change.

That doesn’t exclude the possibility that climate change played no role, but it’s clear that warming temperatures are loading the dice in one direction. A recent analysis took things one step further still: If we can say with probabilities that a single event was caused by climate change, why can’t we express climate’s role in dollars and cents?

In September 2017, Hurricane Harvey made landfall in Texas, resulting in massive flooding in and around Houston. Three years is still too soon for a full accounting of all the economic costs—a single storm often has visible fallout for many years, even decades—but even just the immediate economic costs of Harvey amount to at least $90 billion

Even if climate change made Hurricane Harvey more likely, that doesn’t mean it’s appropriate to attribute the full $90 billion cost to climate change. That’s where we’re back to attribution. The best estimates put the fraction of the risk attributable to climate change at over 70%, resulting in direct climate damages of around $67 billion, with $30 billion as the “likely lower bound,” per the study.

Blaming even $30 billion on one throw of loaded dice, in a single climatic event, is an enormous sum. To put that estimate in perspective, a prominent climate-economy model created by Nobel Prize winner Bill Nordhaus puts the total climate damages for the U.S. over the entire year of 2017 at less than $30 billion.

The Dynamic Integrated Climate-Economy (DICE) model created by Nordhaus is a key tool for calculating the “ social cost of carbon,” what one ton of CO₂ emitted into the atmosphere ought to cost. One of the main inputs into that calculation is a damage function that tallies the costs of unmitigated climate change. The model has long been considered conservative, making a number of assumptions that bias the cost of carbon downward, and the damage function is among its most conservative inputs. By and large it only accounts for the “known knowns” of climate change.

DICE also uses one global climate damage function, making it a top-down exercise. The bottom-up calculation for Hurricane Harvey’s costs that can be attributed to climate change exceeds the total top-down estimate by DICE for the U.S. by almost 50%.

To be clear, Harvey causing $30 billion or more in climate damage doesn’t prove DICE wrong. Risk, after all, equals impact multiplied by probability. Attribution science is all about specifying the impacts and connecting them to climate change. While the probability of Harvey-sized hurricanes has been increasing—and this coming hurricane season will probably be busier than most—it is, fortunately, far from certain that a hurricane will hit and cause high damage.

By now there’s a cottage industry of economists trying to tally bottom-up economic costs of climate change. There’s still a long way to go to get to anything close to a full accounting. But the verdict is already clear: the more climate damages move from “known unknowns” into the “known knowns” category, and the more weather events can be attributed to climate change, the costlier we'll know climate change to be. That should shape the scale of our ambitions for climate action to prevent weather disasters.

Gernot Wagner writes the Risky Climate column for Bloomberg Green. He teaches at New York University and is a co-author of Climate Shock. Follow him on Twitter: @GernotWagner. This column does not necessarily reflect the opinion of Bloomberg LP and its owners.

©2020 Bloomberg L.P.