The tornado will be in Dunwoody at this interesection at 5:43pm: not true

I am a big fan of modern technology progress and innovation. I love my iPod and iPhone and lap top and my portable lightning detector etc. I also love computer graphics and animation and colorful action packed movies and special effects.  But there is a danger when these gee-whiz advances are put in the hands of Madison Avenue. Any fan of the hit cable show Mad Men knows facts truth and sales promotion and marketing are not always in sync. The carny folk at the carnival and circus are always entertaining, and magic is fun, but its important to remember there is no magic, just illusion and slight of hand.

This is true of your broadcast media, and don't you forget it. Radio and TV are after all, entertainment business. Nothing more. They once were more when it came to the news, but not any more.

So if some station tries to tell you they can show you more, just remember, it is show biz. Another scientific paper published in another scientific journal and presented to a scientific conference offers more evidence that when I say beware of TV exaggeration it is not out of jealousy or competition. I use these radars too, and give the time of arrival info just like TV, so this advice applies to my reports, National Weather Service reports, any weather reports.

The study found approximately half the observed pathcast distance errors were greater than 1.9 miles at the initial warming time and AT LEAST 7 miles at 45 minutes lead warning time. About 10% of pathcast errors were AT LEAST 4.3 miles at initial warning time and AT LEAST 14 miles at 45 minutes lead time.

The relatively large errors at longer forecast time ranges illustrate the difficulties encountered when treating a storm threat as a single point moving across the map. Ideally thunderstorm hazards should be treated as areal threats with ranges of time of arrival and departure that incorporate motion uncertainty.  So remember to best account for motion uncertainty as well as any bias along the motion vector (slow or fast) speed error you should think of it as plus or minus 8.5 miles on average and plus or minus at least 20 minutes. But realize the erratic movement of tornadoes and the variation in forward speed of tornadoes make these numbers just the roughest of guidelines.  The main point to take away is to treat any thunderstorm over or near you as a potential tornado if you are under a tornado watch or even near a tornado warning, regardless of the ETA being given by radio or TV forecasters and radar. Be smart be safe.

This research was recently published in the Bulletin of the American Meteorological Society. Here is the link to the study presented at the Conference on Weather Analysis and Forecasting: http://ams.confex.com/ams/23WAF19NWP/techprogram/paper_154236.htm