No Weather Prediction Yet

Requests for a prediction for the upcoming winter are starting already, but the developments through the remainder of October and November are significant to the chances of getting it right. Comments on my predictions and weather/climate commentaries fall into two categories: Those from people who have recognized that the majority of my predictions for both temperature and snowfall have been correct during the past 12 winters, and from those who good naturedly chide me about how accurate I am about weather that has already occurred. With this in mind, I thought I'd give a bit of a summary of some of the research and past conditions that help to improve the chances for predictions to be correct.

By way of explanation, meteorologists need the very best and latest of data and images for a correct short-term forecast. For long-term forecasts, current conditions, or those during the past month, or even the past year do not shed sufficient light on what may happen over the next four months. I use seasonal, 12 month, and five-year trends along with long-term climate change to supplement other information that is occurring globally.

Readers might be interested in a recent study that I conducted to try to determine the onset and effect of climate change in this part of the world. My data for temperature, snowfall and total precipitation goes back to 1966. I used the 25-year period from 1966 to 1990 as a reference period and compared that to the data since that time, which covers 18 years to the end of 2008. It happened that the year 1990 was the warmest year to that point, signifying that the impact of climate change may have started for coastal Nova Scotia. It was significant that no apparent warming trend had occurred in this area over the previous 50 years. The warmest year on record for this area was established in 1999. Over the past 18 years, total precipitation has averaged only slightly higher, but with greater extremes than during the first 25 years. Eight of the 12 monthly precipitation record highs have occurred since 1991. There have been several flood or near flood occurrences during this period, a rate much higher than during the previous 25 years. Surprisingly average snowfall has been essentially unchanged. The temperature tells a different story. The average temperature has been a half degree higher since 1991 than the previous 25 years. That doesn't seem like much but when one considers that global warming is represented by about one degree over the past 100 years, a half degree in less than 20 years is very significant. What does all this have to do with winter? Both winter and late summer/fall have averaged warmer since 1991 than the 25-year reference. Only the May to July period has been relatively the same. The fact that winters have averaged warmer is a bit misleading. Normal winters have been rare, another illustration of a tendency to extremes that has been anticipated with global warming. During 19 winters from 1991 to 2009, six were relatively severe, 10 were very mild and only two were normal. The winter of 2008 was very unusual in that it was relatively mild but with a lot of snow. This certainly illustrates that predicting a normal winter is not the safe way out. How does this factor into a prediction for this winter? That will be the subject of the next article, which will include observations of trends and the indicators from summer and fall conditions. Don't expect a prediction until the end of November.

Fred Giffin