From our expert: How we make the Fall Forecast

Visit this Fall Forecast Guide to the Season for the Fall Forecast, Winter Weather Preview and more.

Monday, September 12, 2016, 9:00 PM - Sweltering temperatures in the Great Lakes. Full throttle heat in the B.C. interior. Drenching rains in parts of the prairies and 'most average summer' on The Rock. The story of summer 2016 is in its final pages, prompting the question: What's next with fall, and moreso, winter?

The Weather Network has released its 2016 Fall Forecast Forecast (and winter preview). Here's an explanation of the process and the key considerations for our drawing up fall and winter forecasts.

What can summer 2016 tell us about the upcoming fall and winter?

Unfortunately, one cannot make generalizations about the next season simply based on the current season. History shows us that for every hot summer that was followed by a mild winter, there was also a hot summer that was followed by a cold winter.

However, the hot summer in the Great Lakes has left us with lake water temperatures that are well above what we typically see this time of year. While, that did not significantly influence our forecast for the season as a whole, the warm water will influence individual weather events (such as the timing of the first frost, and early season lake effect snow events). For more discussion on this topic, please see the following article on the Great Lakes water temperatures and what it means for winter.

Which season is more difficult to forecast – fall or winter?

The seasons of transition (fall and spring) are more difficult to generalize than the winter and summer seasons. Weather patterns are more likely to lock into place during a winter or summer season and there are often clues that point to a dominant pattern for those seasons.

However, the fall (and spring) are seasons of transition in which we should expect frequent fluctuations back and forth between above seasonal and below seasonal temperatures and it can be more difficult to determine which way the balance will sway. Warm falls will still have periods of more winter-like weather and cold falls will still have periods of warm weather. The goal of a fall forecast is to determine which pattern will dominate.

The other challenge is that while most people have a good sense of what is normal for their local area during the winter or summer, it is easy to lose a sense of what is normal for a given day during the fall. Temperatures that are considered to be "normal" steadily drop by 1 to 2 degrees C per week through the season. Weather that would be colder than normal in early October would be warmer than normal by the end of October. Therefore, a forecast for a warm fall could verify against the official data, but perception might be that the forecast was off because temperatures still trended colder faster than what most wanted to see.

What are the keys to the Fall 2016 and Winter 2016-17 forecasts?

One of the key considerations in developing our seasonal forecasts are ocean weather temperature patterns around the world. One region of particular interest is the region near the Equator in the Pacific Ocean. When ocean water temperatures in this region are significantly warmer or colder than normal, this can become the dominant driver of the weather patterns across North America.

This time last year, ocean water temperatures were steadily becoming warmer than normal as we progressed towards one of the strongest El Niño events on record (shown on the left map below).

This was a key to the mild winter last year for most of Canada. However, the pattern this year is rather different as ocean water temperatures in this region are now running cooler than normal (right below).

The current pattern is not expected to progress all the way to a strong or even moderate La Niña event during the upcoming fall and winter. While weak La Niña or even neural conditions do impact global weather patterns, this pattern is not as dominant in its influence as is a strong El Niño or strong La Niña event.

Therefore, temperature patterns in other regions (such as the North Pacific and North Atlantic) will be more influential during the upcoming seasons, especially during the winter. A striking feature in the above map is the large region of very warm water (relative to normal) off the west coast of Canada and south of Alaska. This feature as well as the relatively warm water in the Atlantic Ocean to the east of North America were keys in our search for analogue years (more on analogue years below) which can offer clues on the dominant weather patterns for the upcoming fall and winter seasons.

What is the process for developing a seasonal forecast?

There are two primary elements that we use to develop our seasonal forecasts. First, we analyze global weather patterns and sea surface temperature patterns and then we seek to identify years in the past that had similar patterns and study the weather patterns that were associated with those years across North America.

Years in the past that had similar patterns to current and forecast patterns are called analogue years. Identifying the correct analogue years and researching their associated weather patterns is usually more helpful than the computer models when it comes to forecasting the weather patterns for upcoming seasons.

However, we do analyze computer models which are produced by weather services around the world.

These models have some success in predicting how large scale patterns around the globe will evolve during the next few months. However, the models often disagree with each other or come up with forecasts that does not make sense meteorologically and we need to be able to recognize those errors.

To see how this has influenced our fall forecast and winter preview, please check back on Monday, September 12 at 9:00 pm EDT when we release our forecast for the fall of 2016 with a preview to winter 2016-2017.