As I write this, winter 2016–2017 has been mild across nearly all of the country, with above-normal temperatures and below-normal snowfall being the rule in most locations.
This is generally in agreement with The Old Farmer’s Almanac’s long range forecast for this winter with the most significant exception being the heavy precipitation that fell across all of California. While we did forecast rainfall to be above normal across northern California, we forecast below-normal rain in the central and southern portions of the state and we did not forecast the above-normal snow that fell in the mountains.
This anomaly required a closer look at the historical correlations on which we base our predictions.
We correctly forecast that a weak La Niña would occur throughout most of this past winter and, based on the historic record, this usually means that rainfall will be below normal across central and southern California.
You may recall that last winter, 2015–16, we correctly forecast that most of California would have below-normal precipitation, despite the fact that we were incorrect in our forecast that there would not be a strong El Niño.
The historical correlations between the phase of ENSO (El Niño Southern Oscillation) and California weather have been very strong, although not perfect. But since this is the second consecutive winter in which they have not followed this historical correlation, I searched recent research studies on this phenomenon for an answer.
While I found nothing definite, there is speculation that there are distinct types of El Niños and La Niñas that affect the weather differently. There is also concern among some scientists and/or meteorologists that global climate change is altering the relationship between ENSO and weather.
Predicting the Unpredictable
This would seem to be another example of our struggle in recent years to incorporate the effects of global climate change into the Almanac’s forecasting methodology, which has traditionally been based upon the precepts that the Sun controls Earth’s weather and that the use of solar cycles as a predictive tool enables us to forecast Earth’s weather a year or more in advance.
Once we have determined what the solar activity is likely to be, we find analogues where solar cycles were similar and forecast the future weather to also be similar.
However, two factors have made our forecasts especially difficult to make in recent years: (1) the last time activity in the solar cycle was this low, we did not yet have widespread reliable weather records, so we are forced to extrapolate and speculate on the past weather activity in seeking our analogues; and (2) with the climate changing, the effects that the Sun has on Earth’s weather patterns seem to be changing from past patterns.
We will continue to study the changing weather patterns and relationships and what they mean for future weather. Our goal is to bring you the most accurate long range forecasts possible and to maintain, or even improve upon, our traditional 80 percent accuracy rate.