Just a few months ago, Houston and parts of the rest of Texas were devastated by Hurricane Harvey, which produced the most rainfall of any previously recorded hurricane in the United States. On top of Harvey, there have been 9 other hurricanes in the Atlantic this season, causing more than $300 billion in damage. Hurricane Maria, the strongest storm this year, had maximum sustained winds of 175 mph (280 km/h) – for context on just how fast that is, see this video of a car going the same speed. As we begin recovering from these storms, many people are wondering how often major hurricanes like these will make landfall in the coming years. In other words, should we expect more events like Harvey?
Was Hurricane Harvey a fluke?
Dr. Kerry Emanuel, professor in the Earth, Atmospheric, and Planetary Sciences (EAPS) department at MIT, wrote a new paper on exactly that question. Emanuel uses computer
Running a hurricane model is a lot like taking a video of an ant with your cell phone to figure out how fast the ant is walking, where it lives, and where it is going. To have a good handle on the statistics of future storms, models need to be run for a long time; this is like taking a really long video that takes up a lot of space on your phone. You might have to follow the tiny ant around for hours just to figure out where it is going. While that might work for one ant (or one hurricane), you need to take videos of a lot of ants to learn anything about ants in general. On the other hand, models that have the capability to simulate possible weather decades into the future don’t have the spatial resolution required to study hurricanes – they are too zoomed out. This is like taking a time lapse video from out your bedroom window: you might be able to take a longer video, but you can’t see the ants.
Emanuel’s new approach combines the zoomed-out time lapse video and the zoomed in video of each individual ant into a powerful and fast model to analyze a large number of simulated hurricanes. He “seeds” global climate models with thousands of hurricanes that develop along with the rest of the model.
With this new method, Emanuel calculated the probability of a Harvey-like amount of rain (more than 500 mm) for the end of the 20th century and the end of this century to assess the impact of climate change on hurricane frequency. The model predicted that rainfall like Harvey would only have occurred once every 100 years at the end of the last century, but could happen once every 5.5-years by 2100!
How else can we learn about hurricanes?
There is a lot of room for more research about storms, from modeling future hurricanes to finding evidence of hurricanes that made landfall hundreds or thousands of years ago. Katie Castagno, a PhD student in the MIT-WHOI Joint Program in Oceanography, studies paleotempestology, or past storms. She studied geology as an undergraduate at Smith College, and though she enjoyed her research, she struggled with committing to a field that wasn’t immediately applicable to everyday life. That’s where hurricanes came in.
Castagno uses sediment cores to study past hurricane events. Her research helps create databases of past hurricanes that can be used in models like Emanuel’s. “I love that the research I do now is so straightforward and so relevant. I can explain the mechanics to family and friends of all ages (5-95), and they understand it and its importance,” Castagno said. Like Emanuel, Castagno emphasizes the importance of creating a better understanding of future storms. “I enjoy using geologic tools to work toward studying policy problems, and understanding how storms have affected our coasts in the past is imperative to understanding how they may affect our coasts in the future.” If you want to contribute to scientific research on hurricanes, too, check out this citizen science tool from Zooniverse.