This winter is going to be warmer, and it has nothing to do with global warming. The National Weather Service is changing the formula used to calculate wind chill.

Wind chill, sometimes called the wind chill index or factor, is a theoretical calculation of what a particular combination of temperature and wind would feel like on exposed skin.

However, the WCI, originally developed by antarctic explorers Paul A. Siple and Charles Passel in 1939, has been the subject of considerable criticism as being too simplistic and overstating the chilling effects of the wind.

The index also has been criticized because it's expressed in degrees, the same as temperature. Temperature is a measure of the amount of heat in the air, while wind chill has to do with heat loss and not actual temperature.

That's why the dog's water dish outside didn't freeze, even though the local weather forecaster said the wind chill was 18 degrees.

A Joint Action Group for Temperature Indices was formed by the National Weather Service, a number of other U.S. federal agencies and the Meteorological Service of Canada. Chaired by National Weather Service meteorologist Mark Tew, a San Jose State University department of meteorology graduate, this group has reached agreement on a new wind chill index formula that will be implemented Nov. 1.

The new index is based on wind speeds at face level -- about five feet above the ground instead of 33 feet, where official readings are usually taken. This reduces wind speed by about two-thirds.

The new formula also uses modern heat transfer theory for the human face. Finally, the upper threshold for light winds is reduced from 4 mph to 3 mph.

A further adjustment for the amount of sunshine -- clear, partly cloudy or mostly cloudy -- will be incorporated into the model in 2002.

Copies of the old and new wind chill formulas as well as wind chill charts can be found on the Web at http://ggweather.com/windchill.htm.

Q.  Is there a thermometer with a device for determining the wind chill factor? I know television weather forecasters have them, but can the average person find one? Mary Buzzell - Campbell

A.  A weather monitoring system needs both the temperature and the wind speed to compute wind chill. Most complete weather stations and some hand-held thermometer-anemometer combinations either compute the wind chill factor automatically or provide a table to calculate the values. Hold off getting a new instrument until the changes in wind chill calculation can be implemented.

Q.  Why is Flagstaff, Ariz., so much cooler than Phoenix, Tucson, Yuma and Lake Havasu? It seems like Flagstaff even has fewer 90-degree days than San Jose. Emanuel Lee - Fremont

A.  Like real estate and many things meteorological, the main thing to consider is location, location, location. Or, at least in Flagstaff's case, the elevation. Flagstaff's elevation is approximately 7,000 feet, while Lake Havasu, Phoenix and Tucson are at 500 feet, 1,100 feet and 2,200 feet above sea level, respectively.

Because air is less dense at high elevations, it is considerably cooler. On average, the difference is about five degrees per 1,000 feet of elevation.

And, for the record, San Jose averages 18 days of at least 90 degrees, while Flagstaff gets only three.

Jan Null, founder of Golden Gate Weather Services, is a retired lead forecaster with the National Weather Service. Send questions to him c/o Weather Corner, San Jose Mercury News, 750 Ridder Park Drive, San Jose, Calif. 95190. You also can telephone questions at (510) 657-2246, fax them to (510) 315-3015 or e-mail them to weathercorner@ggweather.com. Please indicate in your e-mail where you live.