Frozen in Time
Jun 18, 2008
A Different Kind of Analog System for Ultra Long Range Forecasting

By Joseph D’Aleo, CCM, AMS Fellow

I have known Jim Witt for decades. He was Science Director for the Dutchess County schools in New York and co-founder of Fleetweather, Inc. In 1962 he initiated a very unique high school weather station at the school in which he taught. The United States Weather Bureau, extremely interested in the program, deemed it the most advanced high school weather program, not only in the entire United States but also in the entire world. Even in those very early years, he possessed a state-of-the-art Teletype machine, which brought current weather data directly into the school. He knew where it was snowing and when it would reach our area. This generated great excitement. He also had our own Weather Radar. This was donated, installed and repaired by Air Weather Service located in Lakehurst, New Jersey. In addition, he had access to computers; yes, even that far back he had use of computers, and the students used these constantly. Graduates of this weather program went on to get their doctorates in Meteorology and Computer Sciences and are now making their mark in prestigious positions such as: Director of Research of the National Hurricane Center, The Lead Forecaster for the NASA manned flight center, The Executive Vice President of Accu-weather, The Head Scientist and Director of the National Center for Atmospheric Research. Another student has written the computer program (GFS Model) which all meteorologists in the world use each day to make their weather forecasts. Jim and his students also explored the solar system and its connection to weather. Jim has continued his research for many decades and has done a celendar each year with daily forecasts with every penny raised going directly to children who are seriously or terminally ill or who are physically or mentally challenged. To date, he has distributed over $1,700,000. He is well known in the New York area (see New York Times story here) for his long range forecasts and is a frequent guest on WOR radio. See some of the testimonials here for his calendar forecasts. 

Jim notes that lunar and solar cycles have a great influence, not only on our weather, but also on our climate. Years of pain staking research suggest they do. The sun is the main driver of our weather and climate. Much of his energies have centered on what really influences the sun and how these changes affect the jet stream, storm development and storm movement on earth. To produce a long-range weather forecast, all the information is fed into a series of computer programs. The programs then calculate an analog date. An analog date is a date, from the past, which has conditions (Sun & Earth, Sun & other Planets, Moon & Earth) similar to the date for which we are making the weather forecast. Unfortunately, weather observations at best, only “look back” one hundred years or so. This is just a “blink of an eye” when compared to the history of weather on earth. If the analog date happens to be a date from the 14th century, it is completely useless since no weather records exist that far back. Because of a lack of weather data during the “early years” it is very possible the analog date he will finally use could be the 4th, 5th or even 6th choice. Naturally, this will not nearly be as accurate as the first choice, the closest match.

Using this approach and gathering complete global historical data, Jim has with one of his former students, Neal Townsend, a specialist in weather computing, developed a very intriguing site here with forecasts for locations throughout the United States and abroad out two years.  Check it out. The first ten days are official government forecasts. Long range forecasts start day 11. His verification statistics are impressive. I have been often surprised how well the forecasts have verified especially for major events. I believe you may also.  I myself have used analog approaches for years using ENSO, QBO, PDO, solar and many other factors with some success. Jim’s approach takes a step back and looks at the factors that might actually control some of these events or indices and the intraseaonal variations we experience.  Again you can see the site here.

Page 1 of 1 pages