Hurricane forecasting has never been better. We say that a lot, but the statistics bear that out, and last hurricane season was literally the best one yet.
This week, the National Hurricane Center released a preview of their annual verification report. They rigorously verify their forecasts each year and compare them to the key tools that they utilize to assist in forecasting. And this year was a doozy in a good way. Put one way: The track forecast 5 days out in 2024 was equivalent to the error of a 2-day track forecast as recently as 2000-2005. Think about that for a second. They have improved track forecast errors by a whopping 3 days on average in 20 years. A 3-day track forecast today performs better than a 1-day track forecast did on average in the late 90s and early 2000s as well. The amount of improved lead time you have on storms today compared to less than 20 years ago is incredible.
Intensity forecasting is another matter. While it too has improved over the last 20 to 30 years, the rate of improvement is not quite to the level of track error. That said, a lot of research has been coming out in recent years with respect to rapid intensification, and one would expect to see these forecasts pick up further improvement in the years ahead. This statement will likely put off some folks, but sometimes the truth can do that: It is likely that climate change plays some role in the idea of chasing a moving target. This is especially true in recent years with extremely warm water temperatures in the Atlantic basin. How has this changed the behavior of rapid intensification? That in and of itself makes things difficult. So while the long-term average continues to go in the right direction, we need to continually invest in research and understanding of the mechanics of hurricanes to improve both lead time and forecast accuracy even more.
Within the report, the probability of detecting a rapidly intensifying hurricane in 2024 sat at around 70 percent. I don’t know what that was 20 years ago, but it was nowhere near that level. Per the report, “although RI remains one of the biggest challenges at NHC, it is worth noting that advancements in hurricane modeling and understanding of the science are making a difference in improving forecasts for even the most challenging cases.” Investment in research is literally paying dividends. As Jeff Masters noted in his discussion on this topic, these research improvements likely saved billions of dollars in 2024 itself, let alone over the last 10 years or so.
Some of the work that’s gotten us here has happened organically through research to improve forecasts and models, often funded by federal programs, but the bulk of this specific for hurricanes is likely attributable to the Hurricane Forecast Improvement Project (HFIP). HFIP was established within NOAA back in 2007 after the very bad 2004 and 2005 hurricane seasons. HFIP goals are to improve hurricane forecasts, extend lead time, extend forecast confidence, and more. It’s interesting to read comments from Bill Read in 2008 about the HFIP and then to see the results we have in the 15+ years since. Our Houston audience should know Bill Read who is about as legendary as a meteorologist can be in a community. He was also the director of the National Hurricane Center from 2008 through 2012 as HFIP began being implemented.
The problem is always funding. HFIP got a nice boost back in the first Trump administration when the Weather Research and Forecasting Innovation Act was passed by Congress. Millions of dollars were allocated to weather research with that law. HFIP was specifically cited in the bill as an area to focus on. And indeed all this money and hard work to improve these forecasts is now bearing fruit.
But we sit in a perilous moment. At the risk of getting too political here, the current “slash and burn” philosophy of cutting government spending risks halting the progress that has been made. Meteorology is sitting at a crossroads right now as it is: Physics-based modeling and AI-based modeling are now both entering operational phases together, and the work being done to marry the best parts of these different technologies is just beginning. The potential results are tantalizing. Forecasts will never be perfect. But if we can continue to reduce errors and improve forecast lead time, we can save lives, save property, and reduce the economic impact from storms. If the couple billion that was put into this effort led to the economy saving tens of billions of dollars from hurricane impacts in recent years, then how does cutting a few hundred million or a billion help improve efficiency, as is the stated goal of Department of Government Efficiency? It’s a simple math problem, and the math doesn’t check out, no matter how much of a deficit hawk you may be. If these programs added no value and cost more than they saved or added to the economy, that would be another thing entirely.
Bill Proenza was director of the NHC back in 2007. He had an interesting tenure, leading the NHC through a tumultuous 6 months. He was known for being a bit brash and for publicly criticizing his bosses, the combination of which led him to be let go from that job. That style may work in some sectors, but in the NHC, the best leaders usually cause fewer distractions. Regardless of that, he made a very good point back in 2011 during talk of budget cuts in that Congress He told the Miami Herald at the time that the short-term savings of cutting hurricane research flights (allocating $17 million in funding to the program it was under versus the $29 million it had been funded at) would come with higher long-term costs. This is exactly the problem. Your political ideology may color how you feel about cutting spending, but the simple reality is that when people without expertise in matters begin to cut programs without understanding them, they’re accepting higher long-term costs in exchange for the short-term gratification of “hey, we did something.” If one is truly concerned about budgets and deficits, they would look at this problem through nuanced lens. Could efficiency be improved in certain areas of programs or research and development? Absolutely. I’m sure it could be. Does the possibility of taking an axe to a program through significant personnel losses or funding cuts put us all at risk of poorer long-term outcomes both physically and fiscally? Yes. Is the prudent way slower? Yes. Is the slower way smarter? Also, yes.
The forecast improvements in recent years have produced amazing outcomes, saved lives, and saved the economy billions of dollars. If we pull the rug out from under this program right when it’s literally at its best, we risk halting that progress, increasing costs and burdens on federal, state, and local governments, not to mention people and small businesses in the path of increasingly complex and dangerous storms. It’s simply not the fiscally prudent thing to do. Many folks in hurricane alley share the priorities of the current administration, which is fine but it’s important to understand the potential ramifications and impacts to us all in certain instances and to make sure your elected officials understand that. Given what we’ve witnessed between Texas and Florida since 2017 or so, it’s in our best interest to continue to work to improve hurricane forecasts as much as possible.
