A common question visitors to our site ask is 'Does TheWeatherOutlook change its long range forecasts if conditions are not developing as expected?' The simple answer is 'no'. Each week our 14 day discussion forecast is updated twice, but our 45 day and seasonal forecasts are only updated at the beginning of each month, and at the turn of the current season.

My opinion is this is a better approach than updating a long range forecast every week to fit the current pattern development and the output from computer based medium range forecasting models. Using the latter approach could well lead to wild fluctuations in the forecast from week to week. How is that then judged? Of course it would give the benefit of being able to selectively claim that the seasonal forecasts were 100% spot on! Because the chances are that if they were updated every week, one of the updated would be pretty much spot on.