AboutDave Dombek Expertise A weather enthusiast from an early age, Dave Dombek is now an AccuWeather meteorology expert.
Experience Dave received his B.S. degree in meteorology from the Pennsylvania State University in May of 1980.
Dave joined AccuWeather in July of 1980, beginning in the Snow Warning Service late that fall after his few month period of training. Within a short period of time, Dave got involved in radio broadcasting. He spent over 25 years broadcasting forecasts on radio stations throughout the Mid-Atlantic and Northeast states, and has also done newspaper forecasts and television briefings. Currently, Dave is the lead television briefer for all of AccuWeather's television clients in the northeatern quarter of the nation. In his daily forecasting routine, Dave helps coordinate the forecasts in the northeastern part of the country. He is also the Director of Forecaster Hiring and is in charge of hiring new meteorologists. Finally, Dave is the supervisor of the climatology department at AccuWeather.
An avid outdoorsman, Dave enjoys hunting and trout fishing. He also enjoys watching football and is a student of history and current events.
If Dave doesn't already have enough on his plate, he is married with two teenage daughters. He also leads a small Messianic congregation in the State College, Pa. area.
Expert: Dave Dombek Date: 4/1/2008 Subject: Two questions
Question Hi--
A couple of questions about weather prediction (a subject I know nothing about).
[1] I can imagine two ways of predicting the weather:
One is historical: You find the day in the past that most resembled todays readinsgs (temperature, moisture, air pressure, etc.) and then predict that tomorrow will be like the day after the past day.
Or you can work more from theory and make a prediction based on an understanding of the interactions of the various factors.
Which(or both) of these methods does the Weather Service use? (Do meterorologists use one method and climatologists the other?
)
[2] Stats ar kept on baseball players and stock brokers. Does anyone check the accuracy of weather prediciton? (For example: does anyone check to see if it actually rained on more or less half of the days in the past five years when a fifty percent chance of rain was predicted?)
Many thanks,
Steve
Answer Steve,
I can't answer for the NWS, since I work for AccuWeather. However, for short to medium term forecasting, it is definitely more the second method that we use. When forecasting more long term, then we do use some historical data.
All kinds of stats are kept on forecast accuracy, ranging from high and low temperatures, mean temperatures, degree days, cloud cover, precipitation totals...including rain and snow, etc., etc.
We do a lot of that research in-house here, and I'm sure other private companies plus the NWS also keep records on accuracy.
