By Bob Henson

Between a record-strong El Niño and catastrophic floods, fires and drought, 2016 was a memorable year for weather and climate in North America as well as globally. What can we expect as we roll into 2017? A precise weather forecast is asking too much, but there is already a lot we can say about some key factors. Here are six developments to watch for in 2017. They're presented in rough order of increasing confidence, followed by details on each prediction.

1. Better Odds of El Niño Than La Niña, but a Neutral Pacific Still Favored

The biggest single driver of year-to-year atmospheric variations around the globe is the El Niño–Southern Oscillation (ENSO), including El Niño and its counterpart, La Niña. A year ago, it was virtually certain that the record-strong El Niño of 2015-16 would continue through at least the first few months of 2016, as it indeed did. This time around, the ENSO signal is far less clear-cut. Sea surface temperatures (SSTs) in the benchmark Niño 3.4 region of the eastern tropical Pacific have been inconclusive in recent months, hovering close to the La Niña threshold (at least -0.5 C below the seasonal average) since late July.

Figure 1. Departures from average sea surface temperature for this time of year as of mid-December show a diffuse, borderline La Niña signal across the central and eastern tropical Pacific. NOAA Climate Prediction Center

It's now become less likely that the ocean and atmosphere will commit to a well-defined La Niña event for early 2017. There's almost no telling what will happen later in the year, on the other side of the infamous "spring predictability barrier" that often separates one El Niño or La Niña event from another. One clue we do have is the unusual persistence this year of a belt of warmer-than-average SSTs from the central tropical Pacific to the west coast of North America. This warm phase of what's called the Pacific Meridional Mode may herald a new El Niño event in 2017-18, as niftily explained by Dan Vimont (University of Wisconsin Center for Climatic Research) in a recent post.

In their joint probabilistic outlook issued in early December, NOAA's Climate Prediction Center (CPC) and the International Research Institute for Climate and Society (IRI) called for decreasing odds of La Niña over the next few months, dropping to just 18 percent by late spring. Neutral conditions are deemed most likely by CPC/IRI, with 65 percent odds by spring and 53 percent by summer. And the odds of El Niño are expected to steadily rise throughout the first half of 2017, reaching 29 percent by summer. Strong El Niño events like the one we just had are usually followed by a significant La Niña event. If the atmosphere instead ends up cueing El Niño for 2017-18, it would reinforce the notion that we've entered a positive phase of the Pacific Decadal Oscillation—a sign that we might expect more El Niño than La Niña events for as long as a decade or two.

Figure 2. Probabilities of El Niño (red), La Niña (blue) and neutral conditions (green), for each overlapping three-month period from November-January 2016-17 (left) to July-September 2017 (right).NOAA / IRI

2. Wide Range of Possibilities for Atlantic Hurricane Action

The well-predicted demise of the 2015-16 El Niño boosted confidence in 2016's largely successful seasonal hurricane outlooks for the Atlantic, where wind shear was down from 2015 and sea-surface temperatures saw a spike atop their long-term warming trend. (See our roundup post from Dec. 27 on global tropical cyclones in 2016 and their connections to climate change.) Because ENSO is one of the biggest shapers of Atlantic hurricane seasons, our current uncertainty about next year's ENSO state means we can't say much yet about whether the 2017 Atlantic tropical season will be hectic, sedate or somewhere in between.

Forecasters at Colorado State University no longer issue formal seasonal hurricane outlooks as early as December, but CSU's Dr. Phil Klotzbach laid out his thoughts for us last week in a qualitative discussion. Along with monitoring ENSO, Klotzbach also keeps close tabs on the Atlantic Multidecadal Oscillation and Atlantic thermohaline circulation, which are cyclic natural variations in SST, surface air pressure, and oceanic flow across the North Atlantic. When the AMO is positive (warm) and the THC is strong, the Atlantic pumps out more hurricanes over periods that can range from 25 to 35 years. At other times, unusually cold waters prevail in the far North Atlantic, typically a sign of a slowdown of the THC and a ramp-down in Atlantic hurricane action.

Figure 3. Warmer-than-average waters covered most of the Atlantic Ocean from the equator northward for the period Dec. 1-18. NOAA / ESRL Physical Sciences Division, courtesy Phil Klotzbach

With cold waters widespread across the far north Atlantic in 2014 and 2015, Klotzbach hypothesized in a 2015 Science article that the active Atlantic period that began in 1995 may have already drawn to a close. Now he's not so sure. "I was generally thinking we had moved into a cold AMO, but we haven't yet seen the re-emergence of the cold anomalies in the far North Atlantic like we have the past couple of winters (at least not yet!)," Klotzbach told me in an email. For this analysis, Klotzbach typically uses SSTs across a box roughly bounded by 50 N-60 N latitude and 10 W-50 W longitude. Figure 3 shows that only part of this area currently has below-average SSTs. "We're just now moving into the height of winter, though, so we may still see some reemergence and anomalous cooling in the far North Atlantic this winter," said Klotzbach. "I decided to hedge with the outlook so far, and hopefully we'll have a better idea of what is coming up by the time the April forecast rolls around." Here are the five possibilities (with odds) put forth by Klotzbach in his December update:

40 percent chance: AMO/THC is above average and no El Niño occurs (resulting in a seasonal average Accumulated Cyclone Energy (ACE) activity of ~ 130).

20 percent chance: AMO/THC becomes very strong in 2017 and no El Niño occurs (ACE ~ 170).

20 percent chance: AMO/THC is below average and no El Niño occurs (ACE ~ 80).

10 percent chance: AMO/THC is above average and El Niño occurs (ACE ~ 80).

10 percent chance: AMO/THC is below average and El Niño develops (ACE ~ 50).

3. More Tornadoes and Tornado Deaths in 2017 Than 2016? Probably So

It's been a blessedly quiet year for U.S. tornadoes, climatologically speaking. According to Patrick Marsh (NOAA Storm Prediction Center), the year 2016 delivered a preliminary total of 1060 tornado reports through Dec. 28, with few or none expected through the rest of the year. This may sound like a very high total, but the number of final tornado reports typically drops from the preliminary total by about 15 percent after duplicate reports have been weeded out. The annual number drops even further relative to prior years when it's adjusted for "inflation" against earlier decades, when fewer people were watching and reporting every twister. Using a linear trend adjustment, Marsh estimates that the final, inflation-adjusted tornado total for 2016 will be around 888, which would be the lowest for any year going back to at least 1954 assuming that the database is normalized (inflation-adjusted) through 2015. "Four of the last five years—2016, 2014, 2013 and 2012—have been the quietest years on record when report inflation is accounted for," said Marsh.

This year did produce a few dramatic outbreaks during peak tornado season, but these played out mostly in open country, where few structures were damaged and few people were hurt. The deadliest events of 2016 were "off-season": seven people died in a Southeast and East Coast tornado outbreak on Feb. 23-24--the nation's second-largest February outbreak on record--and five deaths occurred across the South during an overnight outbreak on Nov. 29-30.

All told, tornadoes have killed only 17 people in the U.S. in 2016, well below the average toll of 46 per year over the three prior years. Assuming we make it to Dec. 31 without any additional tornado deaths, which looks almost certain, we'll have been graced with the least-deadly U.S. year for twisters since 1986, when only 15 people were killed. In data going back to 1875 provided by Harold Brooks (National Severe Storms Laboratory), the only other year with fewer than 20 deaths was 1910, with just 12 fatalities.

Figure 5. During El Niño events (top), the frequency of U.S. tornadoes typically drops. When a La Niña phase prevails (bottom), tornado frequency goes up (indicated by red areas). The effect is strongest in the boxed area.Nature Geoscience 2015, courtesy IRI

The strong El Niño of 2015-16 likely helped tamp down tornado activity this year, at least in the heart of Tornado Alley. Researchers at IRI/Columbia University have shown that the most active spring seasons for tornado and hail over the central U.S., especially the Southern Plains, are linked to strong La Niña events, while the very quietest seasons are related to strong El Niño events. In January 2015, the researchers, led by John Allen (now at Central Michigan University), called for better-than-even odds (54 percent) of a below-average number of tornadoes this year, as opposed to the 33/33/33 percent split (below, above, and near average) one would otherwise expect. (See more details at this conference presentation).

As with Atlantic hurricanes, even a mostly quiet season can still produce deadly mayhem if one destructive event, such as a major landfalling hurricane or a family of violent tornadoes, happens to hit the wrong place at the wrong time. "It's an ongoing challenge to think about how to convey this information," Allen told me. "I think it's also worth noting that we still don't have a lot of other climate signals for improving our forecasts when we don't have ENSO-driven predictability." It's thus hard to tell how tornado counts will evolve in 2017, since the ENSO signal is so weak. However, given the very low activity this year, there's a good chance that we will see more twisters prowling the nation in 2017 than we did in 2016.

Figure 6. The U.K. Met Office predicts that the 2017 global temperature (forecast range shown in green at right) will likely fall below the record value expected to be set in 2016. The dark line shows global temperature since 1850. UK Met Office

Next Page