Curves vs. lines between NHC fcast dots: ACTUAL IMAGES

[LarryWx]

It is said that a picture is worth a thousand words. Thanks to Skeetobite's website, I now have a Jeanne map with NHC forecast coordinates connected by Skeetobite's CURVED connections that I can compare directly with the NHC site's LINE connections. These maps are both based on the 5 AM 9/24/04 Jeanne advisory. Note that Skeetobite's curved connection map shows Jeanne's eye crossing the coast and going well inland near Vero Beach vs. the NHC straight-line map showing her eye not going inland until it barely does so ~160 miles further up the coast in the St Augustine area. Real life examples like this tell me this is a significant issue.

First, here is the NHC site link to the Jeanne movie: freeze on frame #52 and then zoom in to see the 5 AM 9/24 map with STRAIGHT lines

http://www.nhc.noaa.gov/archive/2004/JE ... hics.shtml


Next, notice the big difference from Vero to St. Augustine on the Skeetobite map for 5 AM 9/24 using the same dots but with CURVES:
http://www.skeetobite.com/weather/maps/ ... eets43.gif

[caneflyer]

If people pay attention to NHC's advice to focus on the cone, not the specific forecast points, then it shouldn't make any difference how the points are connected. The actual difference in track between the straight line segments and the curves, in terms of miles, is very small, compared to the size of the cone and the expected error in the forecast. The answer here is not to pretend that the track can be forecast with such accuracy that we need to worry about such fine scale detail, but with better education about the overall risk of being NEAR the forecast path of a tropical cyclone.

Another problem with curves is that there is no unique way to draw a curve through a set of points. There are many, many possible solutions. At least with lines there is only one way to do it.

[LarryWx]

If people pay attention to NHC's advice to focus on the cone, not the specific forecast points, then it shouldn't make any difference how the points are connected.

Caneflyer,
Thanks for your reply. Following your logic, wouldn't this be an argument to either keep the dots but get rid of any connecting lines or to get rid of the dots themselves? Why even display the dots at that site if they want the reader to focus on the cone? It almost seems like the NHC is contradicting itself.

My preference for curves is only if there must be dots with connections displayed.

[caneflyer]

Yes, it is a very good argument for dropping the explicit track from the graphic. However, NHC announced in New Orleans last week the results of a survey that showed that an overwhelming majority of users wanted to see the line. For better or worse, NWS headquarters decided, as a result of that survey, to keep the graphic as is.

I agree that it sends a mixed message.

[Derek Ortt]

the idea of a curved line does make some sense as points do tend to depict sudden shifts in track.

Now, the general public should not be trying to decipher forecast discussions anyways since they are intended for other mets and are quite technical. The public should be focusing on the Hurricane Warning and that should override anything.

That being said, maybe on nwhhc, there will be a curved path utilizing cubic spline interpolation (I say maybe because I will have to write a MATLAB program to do this, and it does not like to interpolate if the X-axis is not monotonic)

[vacanechaser]

Good comparison there Larry. Nice job on finding the maps...

However, I still don't see the real problem here. Like waht was said above, it is within the cone of probablity. And yes, for the most part the average Joe out there may not know whether or not that is a sudden shift in track, or, aturn after making landfall. The NHC has said for years that the track forecast is just a point on a map. That point really means nothing. Either way, the folks looking down the barrel of a major hurricane should know if it goes in in Daytona or Jacksonville they are still very likely to see the maximum winds due to the close proximity to land as the eye approached. I just think that is the point that needs to be driven home. WAVY did a great job of explaining this very point from what I have seen and herd after the fact in 2003 with Isabel. Thats where most of the public is going to get their info. Local TV and radio may need to do a better job in explaining it.

If you live in the warning area, "BE PREPARED"

Jesse V. Bass III
http://www.vastormphoto.com
Hurricane Intercept Research Team

[vacanechaser]

Another thing. In a prior post on the board, someone said something about "stair-stepping". That is true. I have been to the Hurricane Conference and taken a "Hurricane Planning Course" but on by FEMA.. In the course I was in Stacey Stewart and Jack Beven were there to walk us through development, tracking and other aspects of their jobs. In the course Jack had us track an old used up hurricane, Georges I think. With the info given to us from the hunter aircraft and satellite estimates we had turns, curves, twists and loops. After all was said and done, the forecast track was right on the money. How is that?? After Georges made landfall and you go back look at the forecasts, where they may have had a straight line to the coast he may have jogged north or whatever off course. In the long run seeing the forecasts with the straight lines on the map and the actual track, as Jack said, get rid of all the loops and jogs and sair-steps, you have a straight path to the coast. And he was right. It was right on the money.

I know that this may not go for all hurricanes, and the Jeanne map you showed is a great example of where I may agree with you on the curved lines.

This little rant here is just to explain where the "straight lines" come into play as the forecasters discribed it in the class. I hope it all made sense. If it didn't tuff!! LOL... J/J

Jesse V. Bass III
http://www.vastormphoto.com
Hurricane Intercept Research Team

[Ixolib]

If people pay attention to NHC's advice to focus on the cone, not the specific forecast points, then it shouldn't make any difference how the points are connected.

Caneflyer,
Thanks for your reply. Following your logic, wouldn't this be an argument to either keep the dots but get rid of any connecting lines or to get rid of the dots themselves? Why even display the dots at that site if they want the reader to focus on the cone? It almost seems like the NHC is contradicting itself.

My preference for curves is only if there must be dots with connections displayed.

Yep, Larry, I agree with your assessment. Posted almost the same thing in this post http://www.storm2k.org/phpbb2/viewtopic.php?t=56472 when this issue came up in January.

Ixolib wrote:

Their own editorial seems to contradict their bottom line. If they expect the unexpected - "no matter where the forecast line leads" - then why do they WANT to keep the line??? Hmmmm...