Month: June 2019

It’s Just a Matter of Time

Posted on Updated on

The hurrier I go, the behinder I get.

— White Rabbit, Alice in Wonderland (Lewis Carroll)


How did it get so late so soon?

— Dr. Seuss

 

 

 

Pressed for Time

Have you ever procrastinated on something for so long that the stress you feel from hurrying at the last minute makes you feel like you’re more and more behind?  We have to factor time into a lot of decisions we make in life, but like the White Rabbit and Dr. Seuss, we often feel as though we didn’t leave ourselves enough time to get everything done before a deadline.

Traffic on Interstate 5 on a rainy day in Seattle, Washington (Wonderlane, flickr)

Imagine this scenario:  you wake up at 7 AM on a Monday morning, and you have a critically important meeting you MUST attend at the office at 9 AM.  If you miss the meeting, it could hurt your career opportunities.  You open your window blinds, and you see it’s raining cats and dogs outside.  On a good day, it takes you 30 minutes to drive from home to work.  “Shoot,” you think to yourself.  “I probably should leave a little early to give myself enough time in case traffic is bad.”  You hurriedly shower, get dressed, eat some breakfast, and arrange for your significant other to get the kids to school on time.  Rushing out the door with your coffee at around 8:15 AM, you pull up your favorite navigation app on your smartphone, select your office as your destination, and dread consumes you.  Estimated arrival:  9:15 AM.  The rain has caused so many accidents and traffic jams on your route to work that it’s going to take at least an hour to get there.  There’s no way you’re going to make it there before the meeting starts.

Now imagine another scenario:  you wake up at 8 AM on a beautiful Saturday morning, and you and your family decide that you’re going to spend the day at the beach.  In no real rush, you get up, eat some breakfast, pack some food and drinks, and gather your towels, chairs, and beach umbrella.  The kids are in even less of a hurry, but you finally get them all in the car.  It normally takes you 30 minutes to drive from home to the beach, although you figure that the nice weather will probably mean a lot of other people will have the same idea today, and traffic could be a little heavy.  You pull up your navigation app on your smartphone, select your favorite beach, and you’re suddenly a bit annoyed.  It’s going to take almost 45 minutes to get there.  “Oh well,” you think.  “We’re in no particular hurry, and the beach will still be there when we get there.”  Forty-five minutes later, you’ve arrived at the beach, you plop yourself on the sand, and time melts away.

As you can imagine, you’d probably approach these two scenarios very differently, and you’d probably have vastly different emotional reactions to the things that make travel time uncertain.  The question is how risky are you willing to be while planning when to leave.  How bad will it be if you don’t get to your destination at the time you want to get there?  Fortunately, navigation programs and apps allow us to account for time uncertainty depending on our tolerance for risk.

Let’s say I will be driving from Miami to Orlando, and my goal is to arrive at 3 PM.  When I get directions from my navigation app, the program allows me to select an “Arrive by” time, telling me it will take between 3 hours 10 minutes and 4 hours to get there by 3 PM.  The program is trying to account for the typical drive time and the uncertainties (like traffic or road construction) that could make that time longer.  So now I have to decide how much risk I want to take on.  If I have a high risk tolerance (it’s not the end of the world if I don’t arrive exactly by 3 PM), then I’ll probably decide to go with the low end of the time range (3 hours 10 minutes) and leave at 11:50 AM.  If, on the other hand, I have a low tolerance of risk and must be in Orlando by 3 PM, then I’ll probably give myself the full 4 hours and leave Miami at 11 AM.  If traffic on Florida’s Turnpike turns out to be light (ha!) and I get there early, no harm, no foul.  I’ve avoided undue stress and may have even left myself some time to grab a coffee before my 3 PM appointment.

Driving directions and estimated driving times from Miami to Orlando, Florida, according to a popular online navigation program.

When Will the Winds Start?

Things aren’t much different when it comes to the arrival or onset of winds associated with a hurricane or tropical storm.  When we make a forecast for a hurricane’s future track and size, we can derive a time at which tropical-storm-force winds would begin in a city, based on that specific forecast.  We call that a deterministic approach because it in no way accounts for uncertainty in the hurricane’s future track or size.  (We attacked the issue of deterministic forecasts in a previous blog post about storm surge forecasting).  It’s like assuming we won’t hit any extra traffic that will slow us down when driving from Point A to Point B.  But what if the storm moves faster than we’re forecasting?  Then the winds will arrive in the city sooner.  What if the storm gets bigger than we’re forecasting?  That, too, will cause the winds to begin in the city earlier than forecast.

During the 2018 hurricane season, we here at the National Hurricane Center in Miami, and fellow forecasters at the Central Pacific Hurricane Center in Honolulu, began producing new products called the “Arrival of Tropical-Storm-Force Winds” graphics for every tropical cyclone within our areas of responsibility.  These graphics serve as a sort of “navigation app,” giving you the times that tropical-storm-force winds are most likely to begin at different locations based on the latest official forecast, as well as “earliest reasonable” times that the winds could begin if the storm speeds up or grows in size.  The times provided by these graphics can help you decide when your preparations for a storm should be complete according to how much risk you’re willing to take that you won’t have them quite done in time.  If you have no tolerance for risk and must be completely prepared before the winds start, then you’d go with the “earliest reasonable” time.  If you have some wiggle room and can afford not to have everything done before the winds start, then maybe you’d be OK going with the “most likely” time.

Let’s look at an example from Hurricane Michael from 2018 to show how these graphics can be useful.  Here’s the first forecast issued by NHC for Potential Tropical Cyclone Fourteen at 4 PM CDT October 6, when the pre-Michael disturbance was located over the northwestern Caribbean Sea.

NHC Forecast Cone for Potential Tropical Cyclone Fourteen (Pre-Michael) Advisory 1 at 4 PM CDT, Saturday, October 6, 2018.

This first official forecast showed the center of eventual-Michael reaching the Florida Panhandle around 1 PM CDT on Wednesday.  But obviously the outer wind field from the storm would reach the coast before that time—you just can’t deduce when that will occur from this particular graphic.  Here’s what each of the “Arrival of Tropical-Storm-Force Winds” graphics showed for that particular forecast:

Most Likely Arrival Time graphic for Advisory 1 of Potential Tropical Cyclone Fourteen in 2018.
Earliest Reasonable Arrival Time graphic of Advisory 1 for Potential Tropical Cyclone Fourteen in 2018.

 

On the left, the Most Likely graphic shows that tropical-storm-force winds were most likely to have begun at locations along the Florida Panhandle between 8 pm Tuesday and 8 am CDT Wednesday, which would have given people about 3 to 3 ½ days to get ready.  On the other hand, the Earliest Reasonable graphic on the right shows that tropical-storm-force winds could have begun at locations along the Florida Panhandle coast as early as 8 am CDT Tuesday, lessening the preparation time to about 2 ½ days.  (Editor’s note:  You’ll note that I’ve used bold red and black coloring of the text in these scenarios to match the bold red and black titles of the two versions of the graphics above).  Not only would these times help people decide when to have their preparations done, but they also help emergency managers decide when to call evacuations, based on how much time it would take to get people out of areas vulnerable to storm surge before tropical-storm-force winds begin.

So when did sustained tropical-storm-force winds actually arrive on the coast of the Florida Panhandle?  According to the Surface Wind Field graphic, they began roughly around 4 am CDT Wednesday, which falls within the “Most Likely” range discussed above.  In the case of Michael, the track forecast turned out to be very good, and the Most Likely Arrival Time product provided an accurate onset time of tropical-storm-force winds.

Surface wind field for Hurricane Michael, Advisory 15, at 4 AM CDT, Wednesday, October 10, 2018.

Not all track forecasts are this accurate, however.  Consider Hurricane Nate, which made landfall along the Gulf Coast about a year earlier in 2017.  The first official forecast issued by NHC for Tropical Depression Sixteen at 11 am EDT Wednesday, October 4 showed the center of eventual-Nate reaching the Gulf Coast Sunday morning (see below).  The corresponding arrival time graphics showed tropical-storm-force winds most likely to begin overnight Saturday, but they could have begun as early as during the day Saturday.

NHC Forecast Cone for Tropical Depression Sixteen (Pre-Nate) Advisory 1 at 11 AM EDT, Wednesday, October 4, 2017.
Most Likely Arrival Time graphic for Advisory 1 of Tropical Depression Sixteen in 2017.
Earliest Reasonable Arrival Time graphic of Advisory 1 for Tropical Depression Sixteen in 2017.

Nate moved faster across the Gulf of Mexico and a little farther west than was originally forecast, and its tropical-storm-force winds first reached the coast during the day on Saturday.  For this particular storm, the times indicated on the Earliest Reasonable graphic (right) ended up being closer to the times when tropical-storm-force winds began in southeastern Louisiana.

Surface wind field for Hurricane Nate, Advisory 14, at 4 PM CDT, Saturday, October 7, 2017.

The problem is that we can never nail arrival times exactly because we can’t know beforehand if a storm will follow the official forecast or deviate in some way that affects when winds will first reach the coast.  That’s why it’s probably prudent to consult both versions of the product and consider what types of decisions you must make before a storm arrives.  But if you want to be sure that you’ll be prepared before the winds start, it’s advisable to go with the “earliest reasonable” version of the graphic.

There’s one caveat to think about:  just because a location is covered by times in the graphics, it doesn’t mean that tropical-storm-force winds will definitely occur at that site.  NHC also provides versions of the graphics that show the arrival times overlaid on top of the overall probability of a location receiving sustained tropical-storm-force winds during the next 5 days.  So, in reality, the arrival times should be thought of as conditional.  They are the possible times that tropical-storm-force winds could begin, assuming that tropical-storm-force winds occur at all.  As an example, look at the Most Likely Arrival Time graphic issued for Hurricane Florence, Advisory 50, at 5 pm Atlantic Standard Time (AST), Tuesday, September 11.  This graphic shows that locations along the southern coast of North Carolina have a near certainty (>90% chance as indicated by the purple shading) of receiving sustained tropical-storm-force winds, which would most likely begin Thursday morning.  Farther north, locations along the coast of Delaware only had a 20-30% chance (as indicated by green shading) of sustained tropical-storm-force winds, but if they happened to occur, they would most likely begin Friday morning.

Most Likely Arrival Time graphic for Advisory 50 of Hurricane Florence issued at 5 PM AST, Tuesday, September 11, 2018. This version of the graphic also includes the cumulative 5-day probability of locations receiving sustained tropical-storm-force winds (colors).

With that, the time has probably arrived to end this particular blog post.  Some may have wanted it to end earlier, which is reasonable, but most likely you are craving more information.  In a second blog post, we’ll cover how the arrival times are derived from the official forecast, how the earliest reasonable and most likely times are calculated, and some of the social science research that went into developing the graphics.  Stay tuned!

— Robbie Berg

 

 

Advertisements