Sunday Morning: Florence decision time approaches

Things actually didn’t change too much overnight, and I’ll post a full analysis as the 11am NHC package is released since that will be used by most emergency managers for decisions today, so here’s some thoughts on information sources and the process. I want to again caution everyone using social media like Facebook: the Facebook algorithms do not show you posts chronologically, but by popularity!  Be absolutely sure you are looking at the most recent information!

As decision time approaches for taking action on Tropical Storm (likely to be a Hurricane again some time today, possibly this morning) Florence, it’s time to review some information sources and toss out some words on the overall storm response process.  First, information.  There’s really only one source of hurricane track and intensity forecasts: the National Hurricane Center. Your best bet, quick overview, TLDR product is the “Key Messages” graphic.  Here’s the one for Florence.  If your local TV folks are scaring the beejesus out of you, go check the NHC site and see how closely their big picture message matches the NHC message and if not, go somewhere  else.

Everybody uses NHC’s products as a starting point.  The problem is that increasingly other people with far less experience and expertise, especially in the news media, are using the raw track forecasts and data to put their own opinions out there and diluting NHC’s message. Was watching CNN yesterday, and they spent only about 10% of the air time talking about the NHC track and reasoning and the rest talking about track models.  That’s irresponsible – the emphasis should always be on the NHC forecast.  On social media I see people with zero tropical cyclone experience talking about this or that model.  It’s interesting and exciting to discuss the models, especially beyond the 5 day limit of the NHC forecasts.  But NHC only forecasts to five days for a reason, and if someone can’t answer basic questions like what are the components of the TVCN model and what is its current 96 hour cross-track error as of the latest forecast package ( 103 nautical miles if you’re curious), laugh condescendingly and move on.

It is important to realize the roles various parties play in this process, and what their agendas are.  Of course everyone wants you to be safe – but they also want other things and it’s important to realize what those things are, and how it influences their point of view. NHC’s purpose is forecasting with an eye towards issuing watches and warnings.  They want to be “right” – but they also really really don’t want to be “wrong.”  This means they tend to be “conservative.”  What does that mean exactly?  It means they would rather, on average, overestimate things like storm wind speeds and impacts, because they don’t want people to be surprised by a worse storm than predicted (which potentially leads to more casualties and congressional hearings).  So you will often see NHC use the phrase “forecast of least regret”.  However, they are of course aware of that, and strive for a balance.  It’s a tough job.  The bottom line here is that NHC has the best mix of experience and resources to make those watch and warning calls.

Unfortunately, the people involved in the actual evacuation decisions are a) less experienced and b) often have political, legal, and experiential biases that lead to problems.  Key among these are two trends: politicians (especially Governors) making evacuation decisions, and local emergency managers coming from a law enforcement rather than a civil defense background.  I ranted a bit about this in a Facebook post this spring if you’re curious.

The News Media also has some biases and agendas that make information from them sometimes problematic. The first is they want to have a local focus.  That’s great, but in a situation like Florence the stations in each market want to talk about what would happen if the storm hit their viewing area.  So if you review the weathercasts from outlets from Jacksonville FL to Wilmington NC, with a few exceptions you’d think the storm was headed to each one, without a lot of perspective on what the probability of any given area being hit.  The time limitations of newscasts just doesn’t allow for that kind of nuance, even if the on air meteorologist had that kind of background and experience.  Weather is a key “grab” for local markets in the competition for market share.  So they want to make their weathercasts flashy, interesting, and make you want to come back for more.  It’s not that they want to intentionally mislead you, but in their efforts to play on your emotions to make you stay tuned, download their app (which are designed as much as marketing tools as information sources – be very careful with them as they often siphon off your contact lists, browsing history, and other privacy related info!), combined with the time limitations, the net effect is that they are often quite misleading as to the actual likely impacts of a storm on you personally.

The net effect of all this is that NHC is conservative (which is reasonable), emergency managers exaggerate a bit to try to get you to take action and follow their directions (I have a problem with this), Politicians often want to use disasters to make themselves look authoritative (this is disgusting), your local news media is basing their reports towards the dramatic (this makes me want to beat up grass), so the information you get is not just worst case, it’s often blown completely out of proportion across the board to the point where the message that needs to get to people who are often at real, life threatening risk is diluted, and the risk to everyone else distorted.

So what about Enki Research?  As a scientist, I’m looking for some kind of objective truth with respect to the economic and humanitarian impacts of hurricanes.  In my analyses, I don’t care so much if I’m high or low on any given storm, only that on average over many events (and note I often analyze 50 or 60 storms and dozens of earthquakes a year), that the errors in the estimates of economic and humanitarian impacts are minimized.  This is very different from virtually every other analyst you are likely to encounter outside the academic world, especially available to the general public.  As noted above, most other sources have a built in bias towards the high end on impacts – they would rather overestimate the impacts than underestimate them.  But that carries with it some risks.  Overwarning (“boy who cried wolf” effect) is one, increased economic impacts are another.  In one sense I don’t care if you read this or not (in fact, I sort of wish you wouldn’t so I could just do research and not have to spend an hour typing up this crap!).  But I’m also glad folks find it helpful, and as long as you, the unwashed masses 😛 want me to continue posting my analyses, I will …

#Florence Track Notes for Saturday 8 Sept 2018

First, a reminder of a great “one stop shop” for the official word on what a given storm is up to, the National Hurricane Center’s “Key Messages” graphic. It’s pretty straightforward, to the point, and part of the official package on a storm.  Worth watching.

Anybody remember the days when hurricane forecasts only went out three days, and the forecast track models were labeled “for intergovernmental use only”?  It’s great that more data is easily available for both researchers, planners, and the public alike, but it’s also fostered a cherry picking attitude among some who are at least as concerned with hits, viewers, and visuals as they are with accurate forecasts, and for the vast majority of people (including most TV meteorologists and government officials) these things can be scary if you don’t know how to interpret them.  Here’s the forecast track model map as of this morning; some important models are highlighted, as is the official NHC forecast:

Interpreting “spaghetti maps” takes years of training and experience.  It’s one reason why you should be watching that red line (the official NHC track) and not worrying so much about all the other clutter.  It’s like going to the doctor – you have lab work with all kinds of scary numbers, MRI scans, etc.  What matters, and what doesn’t?  It takes an expert.

With hurricanes there are always two related questions: where is it going, and how bad will it be when it gets there.  We’ve gotten somewhat better at the “where is it going” (although better is a relative term; average day 5 forecast errors for the official forecast are around 200 miles). For Florence, there is a big question about the steering Tuesday and later.  There will be high pressure to the north and east of the storm.  Exactly where and how strong is uncertain, and that is causing the reasonable day 6 (Thursday)  forecast location to be within a 800 mile wide swath ranging from the GA/SC  border to out between NC and Bermuda!  The “how bad” part is even harder.  Florence is currently a tropical storm.  It should strengthen, and could be as strong as a category 3 storm at landfall.  However, intensity is tied closely to track, so with so much uncertainty in the track, the intensity forecast is also uncertain.

So what’s the short version as of Saturday AM?  Florence has the potential to be a bad storm for somebody on the US East Coast, Bermuda, and/or Canada later in the upcoming week (Wed/Thu/Fri).  Most likely estimate at this point for hazardous impacts is the North Carolina coast, with points northward (DelMarVa, etc) also likely to see impacts, but South Carolina is not out of the question, and even the Georgia Coast in theory is within the “cone of uncertainty.”  HOWEVER, the storm is still beyond the range of the primary hurricane models.  Forecasts beyond 4-5 days are fraught with error.  For technical reasons noted above there is a lot of uncertainty as to when and where the storm will make its northward turn.

What should you do?  Well, if you have a hurricane plan just watch the theatrics (and if you live on the coast you’re crazy not to have thought about what you would do, but if you procrastinated again this year look at the FEMA site for some checklists).  Still plenty of time to wait and see, then act if it turns out to really be headed your way.  It will be tomorrow before we have some clarity as to the future path (and even then there will be some fuzziness – there always is – but at least it will be actionable fuzziness).  It is possible some state and local officials will “jump the gun” and start actions this weekend, but if I were them I would be watching, getting things ready in the background, reassuring (not scaring) the public, and getting ready to kick off actions first thing Monday morning.

Hurricane Florence – ok smart guys, where’s it gonna go?

Here’s the forecast track model plot as is available when the NHC forecasters are preparing the 5pm advisory, with a few models highlighted.  Their last forecast (at 11am) is in red …

Well?  Clock’s ticking … 😛

In all seriousness, tropical cyclone track forecasting is both a science and an art.  You can’t just “cherry pick” a model, especially just pluck one out because it happens to show the storm passing near your viewing area.  If NHC, who do this every day, and are arguably the best and most experienced at this, why not pay attention to what they are saying?  Yes, I sometimes have issues with their forecasts (especially intensity) on technical grounds, but the biggest problem I see with the current process is how raw NOAA and ECMRF data are (mis) used by the media, and how state and local politicians and some EMA’s screw up the process.  Five days is plenty of time to prepare for a storm if you’ve done your pre-season homework.  If you haven’t, check out the FEMA hurricane checklists.  It’s too late for insurance, but otherwise time to think about what you are going to do.  No need for drama, or hanging on every forecast, unless you just need the exercise.

Tropical Trouble! Well, for #Japan. #Florence, blobofcloudsincaribbean no real threats

As the sun rises over the Atlantic, TV weathercasters and weather bloggers alike dither in glee as a storm off of Africa (its rains suitably blessed) forms to attracts viewers that would otherwise be enjoying their Labor Day weekend, such that said viewers may be convinced they really need a lawyer to get money to spend on guns … hmmm.  Maybe that’s what Warren Zevon meant.  And,  yes, that’s a two-fer of 1980’s hit song references!

The system passing over the Cape Verde islands didn’t form as rapidly as the models indicated, but did pull together enough for NHC to name it overnight as Tropical Storm Florence.  It’s a minimum tropical storm, although it should strengthen a bit the current forecast does not show it becoming a hurricane as it heads out into the middle of the Atlantic …

I’ve seen a few sources getting excited about long range models showing more storms spinning up over the next few weeks.  Don’t pay any attention to that.  Yes, conditions are forecast to be more favorable over the next month than previously.  BUT: that’s climatology at work – September is the peak month for storms.  These models have very little demonstrated skill at this sort of thing.  Take a look at Florence – just two days ago it was forecast to become a major hurricane by multiple models.  This morning?  Not so much.  Tropical Cyclone formation and intensity changes are still a big area of research; scaring the public with these iffy forecasts isn’t responsible.

There is a tropical system (tropical meaning hot, sticky air) moving towards Florida this weekend, then into the Gulf.  NHC has it as a low (20%) chance of forming a storm if and when it makes it that far.  Here’s a really cool visual band image from this morning, taken by the NPP VIIRS sensor, showing the system, lit up by the light of the moon (which is just past full).  Nice to see some lights on in the still recovering Puerto Rico and the Virgin Islands.  This system might bring a lot of rain to Florida on Monday or Tuesday.

Of much more concern than either of these two systems is Typhoon Jebi.  Currently a “Super Typhoon”, it should begin a turn to the north and weakening today. It is forecast to hit Japan early next week as a Category 2 storm, although as JTWC says, confidence is low as to the exact track and intensity.  Even if weaker, it likely will dump more rain across central Japan, and area that has seen devastating floods this year.  This is potentially a bad situation …


Florida declares “state of emergency” over Alberto

There was a time when a “state of emergency” (SOE) meant something.  Nuclear war.  Major natural disaster.  Riots.  Devastation on a grand scale.  Dogs and cats sleeping together, that sort of thing.  But not any more, at least in modern America.  Election years seem especially prone to SOE’s and sober pronouncements by Those In Power that while the situation is dire,  under their beneficent leadership, It Will All Be Ok 😛 .  Politicians running for higher office love them – they get to “show leadership”.  But snark aside, what are some of the factors here?

Declaring SOE in advance of a potential disaster has become a normal way of doing business. It supposedly allows preparations for the disaster.  If the event actually happens, then a “disaster declaration” is made triggering further actions.  In one sense I get it, because the normal workings of bureaucracy have become so slow, convoluted, and constrained, that getting things done in a timely manner (or even at all, given the hyper-partisanship rampant in America these days) is difficult, and a streamlined process for preparing and ordering preparations is perhaps needed.  But there are also down sides.  For one thing, “emergencies” become the normal way of doing business.  In most states, an SOE allows the Governor to suspend the normal budgeting and contracting process, and this encourages abuses. It allows order preventing Price Gouging (which is a good thing, but I guess is OK the rest of the year, given Disney prices!).  It also often has civil rights implications with respect to Police powers and private property, something the US already has problems with.

Another issue is that over time it causes disaster fatigue – most of these “disasters” are  ultimately localized events, or events where there are widely scattered impacts but, in the great scheme of things, while they do hurt those immediately impacted, the events are really not that bad from a wider perspective. I worry that people don’t react with the urgency they should in cases where there really is a threat, when every inconvenience is treated as an “emergency”.

The economic impacts of SOE, watches, warnings, and disaster declarations go far beyond government operations. We are likely at a point where we spend more on preparing and anticipating disaster than if we did nothing except protect lives or when something truly catastrophic was imminent, and cleaned up later.   Many private firms are forced (either for the avoidance of liability, or other reasons) to follow the status of Government, which can be highly disruptive.  A major factor here is the insurance industry.  Now that Alberto is a “named storm,” many onerous insurance restrictions kick in such as restrictions in writing new policies, much higher deductibles, and so forth.  While linking insurance provisions to disaster response may seem to make sense at first cut, in fact it has had a significantly negative impact on consumers.

On the surface, it may seem that declaring an SOE for an incoming storm makes sense, but as with so much in the area of disaster planning, response, and mitigation, it’s a lot more complicated than it seems.