Before I get to the ECMWF model weekly update, I have been studying the El Nino Southern Oscillation more closely over the past few months as it appeared that we were headed for at least a weak El Nino starting this fall and into the winter.

Over the past month, however, the trend towards El Nino has stopped and we basically remain in a near-neutral phase.

El Nino is the abnormal warming of the sea surface temperatures in the equatorial Pacific. The waters in this region were clearly warming relative to normal this summer and appeared to be headed toward El Nino. However, over the past several weeks we have seen a larger area of cooling relative to normal in the central equatorial Pacific. I certainly do not think that we are headed back toward La Nina, but I am starting to believe that the ENSO will have minimal impact on North America's weather this coming winter as it hangs close to neutral.

One note, we are are in the cold phase of the PDO and typically it is hard to get moderate to strong El Nino's under that phase. This may partially explain the reluctance of El Nino to show its face.

The SOI index has also come back up again, which is in the wrong direction if you are looking for an El Nino to form.

Long-range computer models have also been trending away towards the idea of a solid El Nino over the past couple of months and appear to be headed towards near-neutral. El Nino is basically an average of 0.5 or higher for several months. La Nina is an average of -0.5 or lower for several months. See below.

This certainly does not make our winter forecast any easier, especially when you lose a key signal such as ENSO. Once again, this winter will be highly dependent on the PNA, AO and NAO oscillations which have much greater short-term variability than ENSO. The skill in predicting these phases is quite low beyond 30 days.

A couple of things that I will certainly consider when I finalize the Canadian winter forecast will be water temperatures surrounding Canada and the lack of sea ice, at least for the start of the season.

I suspect that the warm SST anomalies surrounding northern and Atlantic Canada will certainly have an impact on the pattern in those areas.

I have also been studying analog years for this forecast and just when you think you may have a fairly close matching year something else comes up and I have to throw it out. One of the seasons that surprisingly came up recently was the winter of 2008-09. It was not an El Nino winter, but near neutral and eventually trending downward in the direction of La Nina. What I found was an upper-level pattern with similarities to 2008, and with this year and the latest ECMWF model also having some similarities.

Also, I will be surprised if this upcoming winter is similar to last winter in terms of temperatures and snowfall across southern Canada. The far north of Canada still looks warm compared to normal. Hard to go against that nowadays.

BTW, I am personally not a fan of using analog years before 1990 as the overall climate pattern has changed (warmed) significantly in the Northern Hemisphere, especially above 50 degrees north since then.

The AccuWeather Canadian winter forecast will be released on Oct. 17.

Weekly ECMWF forecast model Outlook...

Report a Typo