Yellowknife (Aug 13/03) - Just because Environment Canada predicts normal or above normal temperatures in Yellowknife this winter does not mean one can shed her or his warm woollies.

Larry Sonley, a meteorologist with Environment Canada's Meteorological Service in Edmonton, Alta., says in reality, such predictions are based on a comparison of temperatures from the past 30 years.

"For the person on the street, on a day-to-day basis, they won't notice much of a difference," he notes of the air temperature.

"In fact, it could be quite cold."

Yellowknife is also predicted to experience above normal precipitation for the next six months, again which may go unnoticed.

"These (predictions) are not quantitative decisions. It could just be a couple of centimetres."

Sonley says every place has its peculiarities and Yellowknife is no exception when it comes to weather patterns or anomalies.

For example, air temperatures may be cooler in summer due to north or northeasterly winds, while autumn may experience slightly higher temperatures because of the warmth of the waters of Great Slave Lake.

Cloudy days could result in air masses flowing south from Great Bear Lake, or less commonly, from the Beaufort Sea.

South summer winds in the city could be a result of moisture or thunderstorms from the Prairies.

Also, winds will be lighter in Yellowknife than further north because it is within the tree line, he says.

"That's why Yellowknife will seldom get a classic blizzard of more than six hours."

In areas such as the diamond mine locations near Lac de Gras, 385 kilometres, it is not uncommon to have 24-hour severe blizzard weather, he adds.

The national weather centre located in Montreal has one of the fastest and biggest computers in Canada for monitoring weather patterns, he says.

Still, temperatures have a higher rate of accuracy than the amount of precipitation when making predictions -- as moisture is discontinuous, he notes.

Winds at heights of 1,800 feet are also easier to determine as precipitation will interfere with the results.

Often, the rate of accurate weather predictions will be the result of both how well initialized the computer is with data and what assumptions have been made with the data provided, he says.