Why you shouldn’t get excited about long range forecasts …

I ran across a weather blog yesterday (sadly, by someone with an AMS seal) playing up a possible hurricane hitting New Orleans next week, based on the GFS model forecast.  Here’s the forecast map in question, the 225 hour forecast from the 18z GFS/FV3 run on Tuesday, August 13th (the Tuesday Afternoon run).  Looks bad for the sweltering, drunken masses on Bourbon Street …

But, as a reader of this blog, I know that while the new models are better, global weather models still have a tendency to spin up vortexes, so I waited to see what the next run on Tuesday night showed before freaking out …

Oh, now it’s just a tropical depression or storm, and not a problem for the weekly NOLA Friday Night Bacchanalia.   But, what does that mean for my oil and cattle futures?  Lets look at the early morning (06z, Wed. 14 August) for the same time (03z Friday, August 23rd, now “only” 213 hours away) …

Hey, where’s my storm?  Now it’s just a blob of rain on the Alabama/Louisiana border?

Forecast tracks from the global models have improved a lot in recent years, once a storm spins up, and are better at forecasting formation in the short term (0-5 days).  But this example is why these long range models are not that useful for long range hurricane forecasts, and why the National Hurricane Center only does outlooks out five days.  These models just aren’t designed for it, and this far away, the uncertainty is so great you can get exhausted tracking every little vortex that, over the life of the model, will spin up, but don’t even exist in the “real” world (not to mention the systems that spin up in the real world).  So anybody talking about storms forming a week away is probably just phishing for clickbait.  Now, talking about those who talk about them, that’s a public service and perfectly ok …

NOAA’s Atlantic Hurricane Season Update

Last week the US National Oceanic and Atmospheric Administration (NOAA) released an updated seasonal forecast, and due to the waning El Nino has increased their forecast for the number of storms expected this year, now saying there is an increased chance for an “above average” season.  What does that mean to you, the huddled masses cowering in fear along the shoreline, waiting for your inevitable doom?

Exactly nothing (assuming you have a hurricane plan already, which you should no matter what the seasonal forecast says).

First, even if you knew *exactly* how many storms were going to form in a year, it tells you nothing about how bad the season will be.  There have been above average years in raw numbers with no hurricane landfalls.  1992 was a below average year – well, except for Hurricane Andrew.  So unless you know where they are going to go, even one hurricane can ruin your day, and 20 can be no big deal if they are all fish storms.

Second, the numbers used to compute the averages are becoming more and more suspect.  This year’s “hurricane” Barry more than likely would not have been classified as a hurricane in past years for a number of reasons (before anyone yelps, no, this isn’t part of the Vast Global Warming Conspiracy(tm), it’s because of better observation systems that can see small patches of possible hurricane force winds, and different classification criteria).

I really don’t like the hype around seasonal forecasts and their updates.  Dr. Mark Johnson of UCF and I used to do them (including something NOAA doesn’t do, landfall probabilities), but the media circus and subsequent fear mongering were just a bit too much.  We still generate them, and they have decent enough skill, but they aren’t really “actionable” except for narrow applications.  About the only thing they are good from a public safety standpoint is “awareness,” but there are other ways of doing that than shoveling out the statistical stables …

So if you haven’t put together a plan yet, slap yourself and go to visit the FEMA web site and get some checklists to think about, consult your local EMA for risk maps for your risk of flooding (which is by far the major threat to life; the golden rule is shelter from wind, evacuate from water), and put together a plan.  Then don’t worry about it.

Where does the time go?

Well, it’s a monotonic nondecreasing function, so the physicist in me says that’s a stupid question … but I am a bit chagrined to realize my last post was in February. My only excuse is I’ve been working on a very complex (and interesting) climate modeling project that I hope to report on here over the next few months as stuff gets written up and published.  Meanwhile, it is already the northern hemisphere hurricane season so here is a reminder: my automated system that runs the Haetta/TC model puts KML (google earth) outputs showing storm impacts (not just wind speeds, but “plain english” impacts like “trees down, some roof damage”) at this web site:

https://s3.amazonaws.com/enki.public.rtproducts/haetta/status.html

Hope to resume more frequent posts towards the end of this month …

Atlantic low: the saga of invest areas

Here we have yet another example as to why you shouldn’t get excited much about invest areas.  At one point yesterday NHC gave the system a 90% chance of becoming a named storm in the next 5 days.  This morning?  Only 40%.  Here is the forecast model map as of 6am Tuesday:

And if you haven’t learned by now reading this blog, it’s also why you have to be careful with track models and not cherry-pick the scariest line.  Notice the lack of nice colored tracks?  That’s because the major models don’t spin it up and can’t track it very far.  Many of those gray curved longer lines are alternative scenarios where an artificial vortex was inserted into the model.  So the most likely scenario for this thing is that it will spread rain across the northern Islands, Puerto Rico, and Haiti, but not become an organized system.

Atlantic low, California Fires

The low pressure system NHC is watching continues to develop, and the odds of something spinning up later in the week (Tuesday -Thursday) are increasing.  Here is what it looks like as the sun rises on Monday morning (12 November) …

NHC now give it a 50% chance of formation in the next 48 hours (through Tuesday), and a 90% chance in the next 5 days, so they are increasingly sure it will do something later in the week.  The GFS lowtracker loses it, but the model has a low east of the Bahamas by Thursday, other models (such as the UK models) have it going into the Caribbean.  As usual, with a weak, non-system, just identifying a center, much less predicting where it will go, is difficult. We may have more later today, but check back late tomorrow or Wednesday for a better picture of what it will do.

The situation in California is much clearer, and pretty bad.  Weather the next few days is expected to be favorable for the fires to spread.  Using satellite data, it seems that impacts are on the order of $20 Billion dollars so far – in other words, comparable to the hurricanes this year (Florence and Michael).  Worse, there is over $100 Billion(!) of infrastructure within five  miles of the active fires, which are not as of yet contained (one big fire is only 15% contained).  So it may well get worse …

The Thing In the Atlantic …

A few folks have asked about the press reports talking up a potential late storm in the Atlantic.  NHC has an invest area flagged, with a zero percent (yes, zero, nada, nothing, нуль) chance of formation in the next two days.  Beyond that it might spin up.  Here’s the Haetta/TC impact map based on pretty much sheer speculation and fairy dust …

There isn’t much for the models to go on at this stage, so don’t take this seriously.  Check back in maybe Tuesday to see if this is a real thing or not.  In the meanwhile today (Sunday, 11 November) is Veterans Day in the US, and Armistice Day elsewhere.  It is worth pondering the events of 100 years ago, with the end of the First World War.  In many ways that war is still being fought in places like Israel/Palestine, Kurdistan, across the Middle East, and even Ukraine.  Wars rarely end – they just tend to produce sequels – so please take a moment to think about those who serve in them.

Michael now tropical storm, post-landfall estimate $21 Billion

Unlike most storms, Michael completed a round of intensification just before landfall and may have reached 135 kts (155mph) before landfall just east of Panama City, Florida.  After landfall it rapidly tracked into Georgia, and as of 6am the center (now a tropical storm) is just entering South Carolina north of Augusta. Here is the radar from just before 6am, with active warnings (mostly for flood) …

 

Michael will continue to weaken throughout most of today as it crosses SC, but should the winds will begin to increase again as it interacts with the system that is accelerating the storm to the northeast, and it transitions into a extratropical storm.  Here’s the latest impact swath map …

So how bad?  Unlike several recent bad storms like Harvey, Florence, or even back to Sandy, Michael is a traditional hurricane event where the most intense damage is in a narrow swath along the coast and track of the storm cause by either wind, waves, or storm surge, with inland damage mostly caused by wind.  Economic impacts are likely in the $25 Billion dollar range.    The models are estimating that of that, the FEMA Flood Insurance Program will probably take something like a $3 Billion hit, private insurance $9 to $10 Billion, so that insurance will cover just about 50% of the impacts, which again is much more of a “traditional” storm than the recent events with so much uncovered (by insurance) damage.

Michael at landfall: Damage and Categories, impact now estimated at $16 Billion

Michael will make landfall early this afternoon, and where it makes landfall damage will be pretty epic.  Here is the 6am radar from Eglin Air Force Base …

As always with a hurricane, the two questions are where and how bad.  The where part has always been pretty straightforward with this storm.  Here’s the major models (along with the watch and warning areas) – that’s a really tight grouping, and the meteorology on it is very straightforward that Michael will continue north, then start to turn northeast as it gets caught up in a frontal system that is presently over the southeast (the rain over Georgia yesterday and today is from that, not the Hurricane).

The how bad part has changed from yesterday, maybe not as much as the hype of going from category 2 to category 4 might indicate although the dollar value did take a big jump.  Yesterday morning the intensification had paused, and there was a descent chance it would stay at the high 2/low 3 range (that’s what I thought would happen), but outflow increased and shear decreased and the intensification picked up again.  But it’s important to keep perspective, especially away from landfall.  Right at landfall, yes, that 20 or 30mph wind increase makes a big difference.  But that should be in an area where people have evacuated (anyone who stayed on the Florida coast in the warning area was unbelievably foolish). What you do for a Cat 3 and a Cat 4 are the same: get out.

Although landfall will be near Panama City, Storm surge will actually be worst in places like Port St. Joe and Apalachicola, both because surge is worst to the right of landfall (where the winds stay onshore and the water piles up) and due to the unusual shape of the Big Bend of Florida, that kink in the shore line caused by Apalachee Bay just south of Crawfordville to Perry.  Surges could be as much as 18 feet in Apalachicola bay where the water piles up. Here’s one storm surge estimate:

The economic impact estimate doubled from approximately $7 Billion yesterday to nearly $16 Billion based on this morning’s simulation.  Will take a closer look at that in the next post.  Here is my Haetta/TC impact swath estimate, based on the 5am National Hurricane Center forecast:

So what about inland?  This is a potentially a bit worse for far south Georgia, but the storm should rapidly decay once inland and as it interacts with the front.  Trees down, power outages, those in any mobile home in the southwest corner of Georgia (the area SW of a line from Columbus-Albany-Valdosta, such as the Moultrie and Thomasville or Bainbridge areas), and anyone in the orange band in this map in a mobile home that is not securely strapped down, or in an older mobile home – 25 years or older – or RV type vehicle should move to a secure shelter.  Tornadoes are always a concern with this kind of storm so keep your weather alert radio nearby.  As for the coast of GA/SC/NC, tides will run 2 or 3 feet above normal, and right now “normal” is already  about a foot above average due to the fall equinox (“King Tides”).  Here local knowledge is key, so check with the products from your local National Weather Service office (www.weather.gov).  In the Savannah area, for example, that means places like US80 (the road to Tybee Island) will flood around high tide, other low-lying areas near the marsh or on Tybee itself will get some shallow flooding.  Given the speed the storm is moving it shouldn’t be nearly as bad as storms the last few years like Irma and Mathew, but if you are near the water beware.  Otherwise, gusty winds, rains (and street flooding or inland flash floods on creeks in the usual vulnerable areas), and again the occasional tornado must be watched for.  Power is going to go out.  Sigh.  Glad I have a 15kw solar backup system …

Dramatic Michael Update! (ok, that’s clickbait, nothing changed significantly since this morning but here’s some pretty pictures).

The forecast from NHC at 11am really didn’t change significantly other than some tweaks in the track and intensity.  Watch and warning areas didn’t change, Damage swath, damage estimate from my Haetta/TC model didn’t change much at all.  Timing a little slower inland.  So, if you’re getting ready for a Cat 3 hurricane (and if you’re in the warning area, especially around Panama City, you’re an idiot if you aren’t), or in inland Georgia getting ready for some high winds and rain, keep doing that.  See the earlier post for details.

Here’s an animation of the GFS model. Click to get the full loop.  I hate these things.  Sure, they are dramatic and pretty, but they don’t tell you much of anything about potential damage or who needs to be afraid and who needs to just chill …

And here’s a visual band image of the storm.  Got an eye, and may strengthen some more before landfall.

Hurricane Michael Update, Tuesday Morning 9 October: pre-landfall estimate $7 to $10 Billion

The potential impacts of Michael on Florida have become somewhat clearer now that the storm has entered the warm putrid waters of the Gulf of Mexico (sorry folks who live on the Gulf, as someone who lives on the Atlantic coast, it just smells funny to me).  The short version is that while Michael is still “only” a Category 1 with 90mph winds, this is a potentially serious storm that could reach 120 mph or Category 3 strength before landfall.  If you live in the Florida panhandle in the warning areas take immediate action to protect life and property!

Not much to fuss over about the forecast track – the major model families agree on the big picture, with landfall near Panama City.  Here’s the major models, along with the official track (in Red) and the watch and warning areas noted:

Intensity forecast is tricky.  While there are conditions that are favorable for further strengthening, there are also some factors working against it.  I’ve heard speculation saying Michael could reach Category 4 intensity (130mph), but I think that’s increasingly unlikely (in fact I doubt it will be much over 100mph at landfall).  But for planning purposes people in the warning areas should plan on a Category 3 – if it’s not that strong, super, if more, it won’t likely be much more because NHC is pretty careful about underestimating landfall intensity.

As for impacts, there are three areas to talk about: the 80 mile wide (30mi left, 50mi right) swath right around where the eyewall hits land, the Gulf Coast (especially the Big Bend area), and the inland and Atlantic coast.  The landfall area will get hit pretty hard by “traditional” hurricane impacts – extreme winds, rain, waves, and storm surge.  If Panama City takes a direct hit it will be ugly, current estimate is $7 to $10 Billion in damage. Of course, wobbles and things breaking that shouldn’t (or stuff not breaking that in theory should) matter for those kinds of estimates. As the storm decays inland trees down, flash flooding, power outages, that sort of thing begins to dominate.  Another factor for storms like this are tornadoes – these tend to spawn more than a few.  Michael won’t be a Florence/Harvey kind of storm where most of the damage is due to rain flooding after landfall, this is more of a “traditional” 24 hour hurricane event.  But that still means flash flooding along the track.  The second area to watch are areas south of Perry FL, in the “Big Bend” areal  The geometry of the shoreline and strong onshore winds could create some pretty impressive storm surges even away from the usual place you worry about such things to the 80 miles or so to the right of the location of landfall.  If you live in coastal areas prone to flooding, pay attention and take action.

The third area to watch are places right on the coast of Georgia, SC, and NC.  While they will see gusty winds and rain, places right on the shoreline are at risk of some flooding around high tide Wednesday Evening, and the two tides on Thursday.  On the Georgia coast, high tides are already running a foot or so above normal because we are near the Fall Equinox, when the sun and moon are lined up and we have higher than average tides.  Add a foot or two of water on top of that and it can cause problems in low lying areas.  For example, in the Savannah area, using the Fort Pulaski tide gauge for reference, the average high tide is about 7.5 feet.  The “no wind” forecast for the Thursday morning high tide (10:23am) is for about 8.4 feet.  US 80 starts to flood around 9.4 feet, and parts of Tybee (14th street) flood at about 10 feet.  So if Michael adds one to two feet to that tide, which seems reasonable at this point, US 80 will flood, and parts of Tybee may see some *shallow* flooding at the time of high tide.  It won’t be dangerous unless you do something dumb, but will be hazardous and inconvenient.

So you see local knowledge is key here – beware of getting tide gauge references and topographic map datums mixed up!  Tide gauges are referenced to something called “Mean Lower Low Water” or MLLW because boats generally care more about not enough water (eg how low) than how high over fears of running aground.  Topographic maps used to be referenced to mean sea level (MSL) – which on the GA coast is 3.5 feet *higher* than MLLW.  But today topography is usually referenced to the WGS 84 datum, which is close to but not exactly MSL.  I often see local TV weathercasters screw this up and scare the crap out of people because they mixed up MSL and MLLW.  Your local NWS office (www.weather.gov, then click on your location) usually has information to sort all that out in their “Hurricane Local Statements” and other products.

Here’s the Haetta/TC (my new impact model) estimated impacts, using the official NHC Forecast for the track and intensity of the storm.