August 28, 2008
mad mariner your daily boating magazine
  Home| About| Contact| Advertise | Free Registration
 
Search Forums
Google
 Search Forums on MadMariner.com  Search the Web on Google

Forecast Accuracy

Tom_Tripp
Posts: 39
Joined: 2007-08-07

Hi Doug - Welcome aboard! 

 I've noticed that the Tropical Prediction Center still gives us the chart based on the Old Mariner's Rule of 1-2-3 -- graphically presenting the naturally increasing margin of error over time in the future for a current forecast.

The question is:  How accurate, generally speaking, has NWS forecasting become in recent years?  Is there an experience-based rule-of-thumb that says, "you can rely on a 24-hour forecast to about 70%; a 48-hour forecast to about 55%, etc." ??

 The old 1-2-3 rule is great when you're tracking a rather well-localized phenomenon like a tropical storm or hurricane.



TheWeatherMan
Posts: 23
Joined: 2007-12-05

Seems you already know the basics about the Tropical Prediction Center (TPC) "cone of uncertainty"...but let me give more background to others.  The National Hurricane Center (a part of the Tropical Prediction Center) provides the forecast track of tropical cyclones (hurricanes and tropical storms) with a spread of possible locations of the cyclones (this is the cone of uncertainty) which is based on the historical accuracy of track forecasts.  As you might suspect...the farther you go out in the future...the wider this cone gets. 

 Forecasting in general (not just the NWS) has gotten much more accurate over the past 20 years due to advances in numerical models, an increase in number and sophistication of weather observations, and of course improvements made by weather forecasters.  Today's 5-day forecast is as accurate as the 3 day forecast 20 years ago (more or less). 

Just as there is uncertainty in sports predictions and economic forecasts...weather uncertainty goes with the territory.  As a user...you need to recognize the fact that uncertainty decreases from the long term to the immediate short term.  For example...if it is foggy outside right now...I can look out the window and determine that in 5 minutes...it will still be foggy.  However...I become less confident that it will still be foggy outside 2 hours from now (as we all are familiar with fog "burning off" throughout the day).  Point is...use longer range forecasts (5-days in advance) as guidance...not as gospel.   

 

 



Contact Us
[FLASH MOVIE GOES HERE]
Home| About| Contact| Advertise| Press| Link To Us| News Boxes| Free registration| Masthead| Privacy | Editorial Policy
© 2008 Mad Mariner LLC P.O. Box 15282, Washington, DC 20003, (888) 256-5011, information@madmariner.com