If you’d like to read this issue on my website, click here! If you’d like to sign-up, and receive this in your inbox each week, click here! Read past issues here.
Good Friday Morning! This is my third missive from quarantine. Like most of you, I no longer remember what pants are. The days are blurring together, and I legitimately caught myself doing the old college “sniff test” one day — which is a decent sign I need to work on the more day-to-day structure. I also need a break from reading, writing, or thinking about the coronavirus. Fortunately, one of my favorite topics is delivering new data, information, and predictions: the Atlantic hurricane season.
In a column recently, I talked about how people were scaremongering around COVID-19 models, which have a wide range of possibilities. Two things go into models: 1) data and, 2) assumptions. I typically read and study weather models and electoral models. Hurricane season is warming up, and there’s some exciting stuff out there right now that if you know, it makes all the freakouts during hurricane season look like the unprepared freakout we’re witnesses with the coronavirus. So while this newsletter is going to be about hurricane season, it’s also about models and how we predict the future. Links to follow.
Quick hits:
- We know China and Iran are lying about the number of infections and deaths related to COVID-19. Iran has mass graves we can see by satellite. And if this virus in places like Montana and Idaho — you can bet easy money it’s in places like Russia, North Korea, and other areas that regularly lie. Back to China. There’s evidence they never beat the virus as they claim, and it is coming back. They’re blaming “foreigners” for bringing the virus back to Wuhan and causing more cases. What is more likely: they never got it under control, they stopped reporting numbers, and now they’re going to try and flip the script and blame the US for new infections.
- There’s one more mystery here. Chinese telecom carriers are reporting, according to Bloomberg, that they’ve lost 21 million cellphone users in the last few months. The telecoms claim this is because of the quarantine, causing people to cancel accounts. Cellphones are required for nearly everything in Chinese life. The odds people are simply canceling seems far-fetched. The Epoch Times put together a well-sourced argument that these cancellations are at least partly because the death rate in China was far higher than reported. I can’t say one way or another they’re right in that argument — but I also can’t take anything China says at face value.
Where you can find me this week
Please subscribe, rate, and review my podcast on iTunes, Spotify, or Google Play — the reviews help listeners, and readers like you find me in the algorithms. Make sure to sign up for the Conservative Institute’s daily newsletter and become a subscriber at The Dispatch, where I’m a contributor.
Stop scaring people with COVID-19 models. Focus on real data. – The Conservative Institute
COVID-19 brings out the exceptional courage of everyday Americans – The Conservative Institute
The Atlantic Hurrican Season is upon us
It’s not hurricane season yet. That doesn’t officially open until June 1st. But two things caught my eye this week. The first was AccuWeather’s release of their hurricane season prediction for the season. And the second was Mark Suddeth, of HurricaneTrack. He noticed an odd trend in March. Sea-Surface Temperatures (SST) for the Gulf of Mexico are already higher than usual. For the first few months of the year, temperatures were surging about 0.5 – 1.0 degrees above average. But over the last month, those temperatures have gone up closer to 2.0 degrees.
Now, it’s too early for that kind of temperature increase to lead to hurricanes. Conditions in the rest of the Atlantic aren’t in hurricane form yet. But that kind of gain can lead to increased storms and flooding over the southern United States. There’s a surge in warm moist air that can clash with cold fronts that still have an arctic chill to them, which can give us those classic intense storms. And also, come hurricane season, gulf waters will be primed to feed hurricanes.
When you’re modeling hurricanes or anything, for that matter, there are big and small data points. All these points of information interact with each other to create an overall picture. Individual storms are small data points, they interact with the overarching environment that’s in place. If you understand the background, you have a better idea of what can happen with a storm.
The same is true of modeling this virus, modeling elections, and many other things. With elections, there are fundamentals like economic growth, the unemployment rate, consumer sentiment, which party is the incumbent power, and more. With COVID-19, you’re looking at testing rate, dates when social distancing procedures went into place, geography, population density, hospitalization rates, and more.
But back to hurricanes.
This far out, when meteorologists and climatologists are modeling the hurricane season, they’re mostly trying to figure out: what are the broad, fundamental factors at work this season? If you can answer that question, you can then ask a second question: what other past hurricane seasons had similar features, and can we model them similarly? If there are similarities, you can start building a statistical model that has firm foundations that can get tweaked as you get new information.
The Accuweather model
Accuweather released its predictions for the hurricane season this month. They’re saying, “14-18 tropical storms during this upcoming season, which runs from June 1 through Nov. 30. Of those storms, seven to nine are forecast to become hurricanes, and two to four are predicted to strengthen into major hurricanes. ‘It’s going to be an above-normal season,’ Dan Kottlowski, AccuWeather’s top hurricane expert said. ‘On a normal year, we have around 12 storms, six hurricanes, and roughly three major hurricanes.'”
That’s the prediction. What are the broad features they’re pointing to? Warm waters, which I also noted above:
Early in the season, meteorologists will keep a watchful eye on parts of the Caribbean Sea and areas east of the Bahamas, where the water is already very warm. Water temperatures in the Caribbean have already hit 80 degrees Fahrenheit in late March, according to data from a NOAA station.
“Warm water is actually what drives a lot of seasons,” Kottlowski said. “So those will be areas to keep an eye on for early-season development.”
And not just any warm waters, but situations reminiscent of two previous seasons: 1980 and 2005. You might recall 2005 in particular because that was the season that gave us Hurricane Katrina, which hit New Orleans. One of the reasons they’re looking at 2005 and Katrina is because of the sea-surface temperatures. If you go back and look at the hurricane Katrina after-action report from NOAA (pg 13), sea surface temperatures were anywhere from 1.5 – 2.0 degrees above normal in the areas where Katrina strengthened.
So it’s not just warm — but unusually warm for the average of that time of year. The Gulf of Mexico is already warming up, so it’s worth noting that now. That’s part of the case for an above-average season. But that’s not all…
Wunderground hurricane report
One of the big questions every year is whether or not the Pacific Ocean will deal with an El Nino event, or a La Nina event. Broadly speaking, it’s where a specific region of the Pacific is warmer (El Nino) or colder (La Nina) than average. Even though it’s the Pacific Ocean, the size and scope of either event can impact weather across the United States and hurricane seasons.
For the Atlantic, what these two events do is affect wind flow. While water temperatures are critical in feeding a hurricane, the next, and potentially more important factor is wind shear. Basically, a storm needs a reasonably stable wind environment to form into a hurricane. If those winds aren’t right, they can rip the strength out of a storm and make it struggle to develop further. Here’s a description from NOAA:
El Niño produces stronger westerly wind at upper levels of the atmosphere across the tropical Atlantic than in normal non-El Niño seasons. This increases the total vertical wind shear, basically shearing the tops from developing storms before a healthy circulation can form. El Niño events generally suppress Atlantic hurricane activity so fewer hurricanes than normal form in the Atlantic during August to October, the peak of Atlantic hurricane season.
During La Niña, westerly winds high in the atmosphere weaken. This results in an expanded area of low vertical wind shear, allowing more Atlantic hurricanes to develop during La Niña events. La Niña increases the number of hurricanes that develop and allows stronger hurricanes to form.
The chances for the continental U.S. and the Caribbean Islands to experience a hurricane increase substantially during La Niña and decrease during El Niño.
El Niño and La Niña also influence where Atlantic hurricanes form. During La Niña, more hurricanes form in the deep Tropics from weather disturbances that originate over North Africa. These systems have a much greater likelihood of becoming major hurricanes, and of eventually reaching the U.S. and the Caribbean Islands.
What are the predictions for this year? Right now: unclear. Most models are expecting a neutral year without either El Nino or La Nina, except towards late in the season, odds for La Nina surge. Add to that, we’re almost due for another La Nina swing, as it tends to be very cyclical:
La Niña events tend to recur about every three to five years. They typically build in the northern fall and fade by spring, then sometimes rebuild for a second or even third year. The last La Niña event was in 2017-18, so it wouldn’t be premature for La Niña to return in 2020.
Why focus on it? It’s more of a trend line than anything. Even if things are mostly neutral for the season, the trend towards La Nina suggests conditions could become more favorable for hurricanes.
Finally, there’s one more prediction worth mentioning, and that’s Tropical Storm Risk. They run a far-out simulation in the winter, which lowers accuracy. But it’s worth noting that they see the lack of anything major affecting storms due to a “projected absence of any climate forcing factor.” Basically, the environment won’t, according to projections, spin up more hurricanes than average because nothing will force storms to develop as in prior years.
Different stories
You’ll notice, all these are different predictions. And that’s for a reason. Everyone, in all their modeling, tends to focus on various items that are important to their modeling. Some prefer warm water, some wind shear, and others follow trend lines. What matters is the assumptions of your model, and the data you feed it.
And I’ll note: none of these models is better than the other. People like touting specific models like Euro and other American models, but we need all of them. Every single model assumes something different, and we trust the averages to give us a rough estimate. It’s like poll averages in an election — the average tends to tell you more than individual polls. But even with that — some pollsters are better.
The same is true of any model you’re studying. Assumptions and data are weaving in and out to produce a best guess of what will happen. That doesn’t mean the models people are using right now are useless. People and governments should absolutely use worst-case scenario models and prepare accordingly. The same goes for hurricanes, and if you’re a campaign, electoral models. You can’t presume a model is wrong and design public-police like that.
That said. There’s a difference in preparing and scaremongering. I’ve seen far too many people use models to browbeat and terrify people than using them to inform. One of the more bizarre things I’ve witnessed recently is people defriending people because they push back against worst-case scenario predictions. Even though these are probabilistic scenarios, not absolutes.
Very few people in our government believe the worst-case scenario is happening in our country. We’re preparing for the worst, hoping for the best, and responding responsibly. Most of the people scaremongering with these models are not doing so because they think the virus is a terrifying ordeal, they’re doing it because they have political motivations. The same goes for people attacking and not believing in the extent of the virus. People want their political team to win, so they’re bending reality to their will.
Reality is always different. Models consistently predict various things for hurricanes, and yet we learn new things every time we investigate a storm. The same is true of this virus.
Stay safe, everyone and make wise decisions.
Links of the week
Coronavirus Disease (COVID-19) – the data – Our World In Data
America’s Pandemic Data Gap: The problem isn’t just testing — we still know far too little about what we already know. – Brendan Foht, Samuel Matlack, Ari Schulman, The New Atlantis
Chinese Propaganda Has Infected Daily Mail’s Coronavirus Coverage – Chuck Ross, The Daily Caller
[Spain update; source in Spanish] China sent 640,000 tests to Spain. They don’t work. “Rapid tests for coronaviruses purchased in China do not work well: Microbiology laboratories report that the tests acquired by the Government have a sensitivity of 30% when they should exceed 80%” – El Pais
Trump Was Probably Wrong About Coronavirus and Suicides, But the Associated Press Botched Its Fact-Check
It wouldn’t cause more deaths than COVID-19, but an economic crisis could indeed raise the suicide rate precipitously. – Robby Soave, Reason
Coronavirus direct payments: Who qualifies and for how much – Jessi Turnure, WKRN
Former Vice President Joe Biden Keeps Pushing Misinformation About Coronavirus. Here Are 5 Examples – Peter Hasson, Daily Caller
No, President Trump Did Not Make Anyone Ingest Fish Tank Cleaner – Charles C. W. Cooke, National Review
Twitter Thread(s) of the week
Alex Berenson walking through a significant walk back by COVID-19 modeling in the UK.
Satire piece of the week
Pants Sales Plummet As Everyone Working From Home – Babylon Bee
U.S.—With everyone staying home and working remotely because of the novel coronavirus, one industry is especially being hit hard: pants manufacturers.
“I’m at home all day now, so what do I need pants for?” said Carl Hampton, a computer programmer, expressing a common sentiment. “I’m starting to feel dumb that I ever wore pants.”
While shirts are still somewhat popular, pants have taken a blow from the new telecommuting paradigm and have been found to be completely unnecessary articles of clothing and are thought to even restrict creativity and productivity. “Pants are a bane to a worker’s existence,” said Spencer Neal, an accountant. “Culturally, you have to wear them around other people, but now that we’re by ourselves, goodbye, pants!”
Thanks for reading!