Climate Change and the Terrible, No-Good Odds of Bad Weather
About three weeks ago, a three-day storm came through southern Louisiana and flooded the state capitol. Baton Rouge itself had over 17 inches of rainfall in the 72-hour period. Some nearby parishes got more than two feet. The storm was historic, dumping three times as much rain as Hurricane Katrina.
So, was it climate change? A new study from NOAA and the World Weather Attribution says yes, sorta. It’s complicated. Deep breath: With climate change, the odds of a storm with this magnitude hitting this region during this time this year are one in 30. In an alternate world in which the Industrial Revolution had never happened, the odds would have been one in 50. An increase of 40 percent, all thanks to climate change. In terms of actual rainfall, the study says global warming added about 10 percent to what would have fallen anyway—a few inches.
These results aren’t easy to explain. But these dense little statistics are important: Understanding them will help as engineers adapt infrastructure to the altered weather regimes civilization has baked into the atmosphere.
Climate attribution is a relatively young science. Its debut was a 2004 study published in Nature that tried to decipher whether human-caused climate change contributed to a freak 2003 European heat wave. In general, the methods haven’t changed much. Begin by looking at the historical record. How common is this type of weather event? Europe hadn’t experienced a summer that warm since the 1500s. Do a little math, and you have the baseline odds of recurrence.
The climate connection is trickier. Scientists first need valid weather models. “When you run the model, does it look anything like the real world?” says Karin van der Wiel, a research associate at NOAA’s Geophysical Fluid Dynamics Laboratory and lead author of the Baton Rouge paper. Getting it to accurately predict (or hindcast) past weather events in the real world is an iterative process—basically, the scientists add new data and tweak parameters until the model fits.
Once they’re satisfied the model reflects reality, the scientists run it head to head against another model without a climate-changed atmosphere. “We do separate experiments where we target current conditions of greenhouse gases, aerosols, and ozone and run similar simulations with those variables set to pre-industrial revolution levels,” says van der Wiel. Then they watch. Does the storm, wildfire, or drought show up more frequently in the climate-changed model? Are the storms wetter? Wildfires bigger? Droughts drier?
Don’t expect a binary answer. “Climate attribution looks for signals in weather events, it can’t say for sure if this rainfall event, this fire, or this drought is caused by climate change, because weather has always happened,” says van der Wiel. What they look for are probabilities. Like: The 2004 study found (with 90 percent certainty) that industrialized human civilization doubled the odds of that 2003 European heat wave.
Early attribution studies took a long time, up to a year. This time, though, van der Wiel and her colleagues ran the models, and did the science in less than a month. Part of that speed comes from the experience of the scientists, and part of it is the World Weather Attribution initiative, a way for research groups around the world to share data and workload.
Is there a point in moving so fast? Sure, nailing down the climate connection might offer victims some catharsis. And partisans in the (uniquely American) debate over whether humans are responsible for climate change are always happy to have a new data point to lob.
But really, the point is preparedness. “These researchers are saying that the likelihood of a storm crossing a red line into historic levels has increased by 40 percent,” says Marty Hoerling, research meteorologist at NOAA’s Earth System Research Laboratory in Boulder. City, state, regional, and federal planners need to know whether that means climate change has reset the bar for everything from flood insurance to property values.
And what if this storm made critical flood infrastructure like dams and levees obsolete? Every bit of flood protection infrastructure comes at a cost. Ideally, that cost reflects the real danger of an extreme event recurring (Hurricane Katrina reminded us to be skeptical of that ideal, but let’s just pretend such a thing exists). The levees protecting the Amite River near Baton Rouge were designed to handle 14.6-foot wave crest, a record set by a nasty 1983 storm. This year’s August storm beat that level by nearly three feet. Is an upgrade to protect against this storm’s high water mark worth it? “Nobody wants to go bankrupt for an event that might only happen every 500 years,” says Hoerling.
So what do you do if you are some sort of southern Louisiana god-planner and someone walks in your office and tosses this study and a fat wad of taxpayer money on your desk? It’s not rocket science: It’s actually far more difficult than that. “What looks like a miracle of launching a spacecraft into the solar system and having it arrive at a certain location relies on forces that are extremely deterministic and well known,” says Hoerling. Weather, climate, and yeah, human nature are full of unmeasurable and unknown forces. When it comes to upgrading infrastructure to protect from climate change, no decision will be easy. But at least numbers like the ones from this study give you something to work with. So, how about those levees?