Hurricane 2025
❗ That's not a typo: even while 2024 is still very much on everyone's minds, it's also time to be thinking about 2025.
🕰 Back in 2006 and 2007, my co-authors and I published a bunch of articles about how to estimate hurricane rates in the current climate (you can find them on my website by going to publications and filtering on "hurricanes"). Some were peer-reviewed, and others we just posted on arxiv and left it at that. The articles are a weird mixture of climate physics and statistics, which is what you need for thinking about this problem.
😵 ❓ The biggest challenge was trying to understand hurricane decadal variability, given the limited data available. A secondary issue was trying to understand the possible affects of climate change.
👴 18 years later, these are still the two big questions (and still in that order). But what's changed is that we now have 18 more years of data to go on, and a better understanding of the possible impacts of climate change. Also, I've got better at statistics. Given all that, I figured the whole question of how to estimate current hurricane climate was due a reassessment.
💨 I've just finished part one of the refresh project. Reading back over the old articles we wrote, it turns out that some things we assumed about hurricane behaviour were reasonable, but others not so much. I've put together new estimates for landfall frequencies, for 2025, based on learnings from the new data.
📖 Nowadays, I publish all my research in peer-reviewed journals, so that you know it's legit. So I'll be making the results and methodology public at that point. My research is sponsored by a group of insurance companies, and prior to publication the results are only available to them. That's the fine balance I tread between wanting to publish and also wanting to give my clients exclusivity. As well as the projections, I provide software that makes it easy to make the model adjustments required, and calculates the impact on various metrics. The software incorporates various algorithms I've published over the last few years. It uses new technology that my kids taught me about, so you can even run it on your phone (and it's surprisingly fast).
☕ Any questions about any of that, let me know, otherwise I'll get some tea and go back to writing the paper.
thanks,
Steve
🤔 ps: have a look at the chart and see what you think about decadal variability, climate change, and what the distribution of possible values for the next year should be
🙏 pps: many thanks, as always, to the companies that fund my research
#hurricane #insurance #risk #catastrophemodeling