Climate Central has a model that can show how much more intense hurricanes are based on how much warmer the ocean is.<p>This is an advance for <i>attribution science</i>, which aims to show how much of a natural disaster is attributed to climate change.<p>In the future I expect a party, perhaps an insurance firm, or reinsurance firm sue oil companies for their role in accelerating climate change to pay for the cost of natural disasters.
The empirical data is that we don't have more hurricane (in total or per category) per decade now than we did 150 years ago [1]. So if climate change (which is real, to be clear) is intensifying hurricane, shouldn't that be reflected in the data?<p>Likewise, when Helene hit North Carolina, this too was attributed to climate change except the <i>exact same thing</i> happened a century ago [2].<p>When we talk about the impact of hurricanes on infrastructure, people and buildings, we forget that there are an awful lot more people now than there was a century ago. 100 years ago, the population of Florida was less than a million.<p>Calling every storm a once in a century storm or saying how once a century events now happen every year (you'll hear both of these claims often) does nothing but discredit climate change.<p>Move a normal distribution half a standard deviation left or right and you have real impact but it will take you a lot of data points to figure out that's really happened. Extreme outliers in either case will tell you, quite literally, nothing.<p>[1]: <a href="https://www.nhc.noaa.gov/pastdec.shtml" rel="nofollow">https://www.nhc.noaa.gov/pastdec.shtml</a><p>[2]: <a href="https://www.usatoday.com/story/graphics/2024/10/07/hurricane-helene-great-flood-asheville/75456390007/" rel="nofollow">https://www.usatoday.com/story/graphics/2024/10/07/hurricane...</a>
> the potential damages caused by a storm increase exponentially—by a power of eight—with increases in wind speed<p>As any CS undergraduate will be able to point out, this is not exponential, it's polynomial. Exponential means that damage ∝ a^speed<p>A good CS undergraduate will be able to point out that this doesn't matter, as the constants involved may well be more important.
Glad we've finally figured out how to isolate the impacts of a single variable on a massive and incredibly complex system without even enumerating all of those other variables. The Science (tm) is so cool!