As the Spring season moves along, signs begin to become more clear regarding the development of an El Nino. Although there has been relatively high confidence in the development of an El Nino for quite some time, there was still a small bit of doubt, as some expressed skepticism. However, over the last month or so, the atmosphere has undergone changes which should remove any seeds of doubt regarding whether an El Nino will be developing this Spring and Summer. The only doubts that exist now is the final strength of the event (El Nino events usually peak in the Autumn): high-end moderate/low-end strong, a strong event, or a super strong event. For more information regarding the formation of El Ninos and what they mean for our weather pattern, check out the article we published last month.
What has happened that makes us so confident? The easterly trade winds which usually keep warm water well to the west towards Indonesia have shifted to westerlies.
This means that the base state of the atmosphere which prevents El Nino events has completely changed to one that favors El Nino events. The typical climate pattern has easterly trade winds along the Equator, which blow warm water along the Equatorial Pacific to the west, towards Indonesia and Australia, leaving the rest of the Equatorial Pacific relatively chillier. Now that the surface currents have shifted to blowing from west to east, it becomes much easier for that same warm water from the west to move eastward throughout the rest of the Equatorial Pacific, leading to an El Nino.
To further put this into perspective, let’s compare our current state of the ocean currents with some previous El Nino events. Unfortunately, the ocean current data we’re using only goes back to 1992, but there are still several events post-1992 that we can use.
Before we begin comparisons, we’ll state a few thresholds for criteria (ranges indicate degrees above average):
Weak El Nino: Nino region 3.4 trimonthly peak of +0.5C to +1.0C.
Moderate El Nino: Nino region 3.4 trimonthly peak of +1.0C to +1.5C.
Strong El Nino: Nino region 3.4 trimonthly peak of +1.5C to +2.0C.
“Super” Strong El Nino: Nino region 3.4 trimonthly peak of +2.0C or greater (this definition is bit more subjective).
There is also language regarding the orientation of the El Nino. West-based means that the strongest warmth anomalies are on the central and/or western sides of the Nino regions. East-based means that the strongest warmth anomalies are on the eastern sides of the Nino regions. Basin-wide means that the anomalies are uniformly spread out. Our previous article discusses more about the implications of east vs west-based, but east-based El Ninos tend to amplify their effects further east, i.e. warmer temperatures spread throughout the entire country during the winter, and the increased southern jet stream is also stronger further east, reducing Atlantic Hurricane activity even more than central and west-based El Ninos.
Note how instead of what is typical (blue arrows, easterlies), the trade winds have reversed, becoming westerlies (red arrows).
May, 2006. This event eventually became a borderline moderate east-based El Nino.
Although there was a bit of a reversal in the eastern part of the Equatorial Pacific (perhaps partly explaining why this El Nino was east-based), we are certainly well ahead of this event in terms of shifting the entire Equatorial current.
May, 1994. This event eventually became a moderate basin-wide/slightly east-based El Nino.
Although there were some hints of the currents becoming westerlies, we are certainly ahead of the 1994 progression.
May, 2002. This event eventually became a moderate centrally/west-based El Nino.
The 2002-2003 El Nino event actually became a bit stronger than the El Nino events above, but you would never have thought that based on the ocean currents in May, 2002. They had barely begun to reverse at all.
May, 2009. This event started off as a borderline east-based/basin-wide El Nino, but shifted to west/centrally based during the Autumn and Winter. It was also a strong El Nino.
May, 2009 certainly had a reversal of the trade winds at this time, just as we do this year. However, even the strong El Nino event of 2009 seems to have lagged slightly behind this year with regard to the intensity and degree of the current shifting throughout the entire basin. The fact that we seem to even be ahead of 2009 bodes well for this upcoming El Nino to be a strong one.
May, 1997. This event eventually became the strongest El Nino on record, and it was east-based.
May, 1997 had the most impressive reversal of all the years listed. The only year it truly compares with from a basin-wide perspective is this May (2014). May, 1997 seems to be slightly ahead of this year, however, in that the entire basin had slightly stronger westerly winds as a whole. But it’s certainly hard to convince anyone that we are not very close with 1997. It seems prudent to say that although we may be slightly behind 1997 and ahead of 2009, that we are closer to 1997’s progression than we are with 2009.
Another key conclusion is that the El Ninos that have become strong in the past 20 years tended to reverse the state of the Pacific early (meaning by now). The moderate El Ninos did not make the progression that we made.
What about 2012? We had an El Nino watch that year, but that El Nino completely collapsed. How do we know this El Nino will not collapse again?
2012 was an interesting year. In a lot of aspects, that year had a lot of potential to produce a strong El Nino. The ocean currents certainly reversed in May, 2012.
Looking at the image above, it certainly makes sense that we had an El Nino watch. However, there were a couple of small minor caveats, even in the above impressive image. Note the persistent easterlies just west of 80 degrees W that were fighting with the strong westerly burst. This may have prevented that current reversal from truly altering the state of the ocean. The reversal also wasn’t as prominent near the Date Line (180 degrees).
Even with that said, the above image still shows a conducive environment for an El Nino of some kind to form. However, there is more to an El Nino than a reversal of the wind currents — that is only part of the equation. The combination of establishing westerly wind currents as well as a flattening thermocline (boundary between the warm, surface water and the cooler, deeper water) in the east Pacific, and warm water propagating to the surface in the east Pacific is key to the entire feedback process. Warm water further to the east is key to the circulation that favors the currents to shift, and the vice-versa is true as well. If both do not co-exist (i.e., if only one of them exists) then it is quite hard for the one that does exist to sustain itself, and the El Nino could break down. Here is a good animation of El Nino to illustrate this point.
Let’s take a look at the current state of the thermocline and the depth of warm water.
In a “normal” state, there is much more warm water in the western Pacific both at the surface and at large depths than in the eastern Pacific — the cool water is very close to the surface in the eastern Pacific. However, look at the flattening of the gradient as we head to the east, and the 28C isotherm has pushed well to the east. This is a strong indication that the thermocline is flattening.
In the anomaly image, look at how much significantly warmer than average water exists in the Central Pacific subsurface (160 Degrees W), and especially just below the surface and already hitting the surface in the Eastern Pacific. Also note how there are even cold anomalies in the western Pacific — further indicating a reversal from normal. Of course, that water is not cold, it is just colder than normal for that region. This means that any westerly currents and Kelvin Waves which push water eastward will still be moving warm water to the east, and by the time that water is pushed towards an area that typically sees colder water, it will be reflected with a warm anomaly. Furthermore, the aforementioned warm water at the subsurface near 160W should also push eastward due to the currents and Kelvin Waves, and upwell towards the surface.
Additional Note: The little x’s in the diagram mark where we are getting observational data from buoys. However, because of a lack of government funding, we are missing two buoys at 120W and 140W. This missing data most likely explains the apparent gap in the warm anomaly, and why the thermocline appears relatively sloped in that same area. If we had data there, the warm anomaly would probably look even more impressive, and the thermocline would appear flatter. The lack of funding started in June, 2013; any year before that has these currently missing buoys, including 1997. Yes, we had more observational data in this region almost 20 years ago than we do now.
Now, let’s take a look at the same image valid for May, 2012.
Notice how much steeper the thermocline was in the entire eastern Pacific in May, 2012 than it was this year. Additionally, although there are some warm anomalies near the surface in the eastern Pacific, they are not nearly as strong as the ones that we have this year — the subsurface does not have nearly as much heat. Furthermore, right below the warm anomaly in the eastern Pacific is a cold anomaly. This means that any upwelling cycle of a Kelvin Wave is likely to eventually push that cooler water towards the surface, weakening the warm anomaly. A weakening of the warm anomaly can facilitate a re-strengthening of the easterly trade winds, helping the El Nino to collapse. This May (2014) has no cold anomaly just below the warm area in the East Pacific.
Building further on that point, the warm anomaly is stronger but much better connected as well this May (2014), given the strong Central Pacific subsurface warmth, yet in May 2012, cold anomalies existed in those same locations (160W). That’s a further indicator of cold water moving eastward and potentially putting a dent into the warm anomaly. There was a decent amount of warmer water well to the west — but remember — warm water in that location does nothing to reverse the state of the thermocline at that very instant — another indicator that 2012 was well behind this year in terms of progression to an El Nino.
Now that we can put the 2012 collapsing scenario to rest, let’s further compare this year to 2009 and 1997.
The thermocline appears a bit flatter than it was back in May, 2012, but not as flat as it is in May, 2014. The 28C isotherm had progressed much further east in 2009 than it did in 2012, but not as far east as it has progressed in May, 2014.
The warmer anomalies are noticeably more significant in the 2009 image than they are in the 2012 image — particularly in the Central and Eastern Pacific, but not even close to as impressive as they are in May, 2014. The fact that May, 2014 is running well ahead of May, 2009, which turned out to be a Strong El Nino, evidences the potential for this El Nino to surpass that event. There is still a window where our upcoming El Nino event could be stronger than 2009, yet still weaker than 1997, so let’s move onto the May, 1997 comparison.
Not surprisingly, the above image is quite impressive with the progression. The thermocline had already shown significant signs of flattening, and the warm anomaly in the Central and Eastern Pacific subsurface is significantly positive. However, May, 2014 is actually very close to this progression: although 1997 had a slightly flatter thermocline to the east, May, 2014 had a further east progression of the 28C isotherm.
May, 2014 also had a very similar anomaly strength and orientation, as well. Remember the lack of buoys in 2014 that caused the apparent gap in the anomaly — when one compares the areas where we have the data, May, 2014 is a very close match.
Overall, it seems fair to say that May, 2014 is quite close to the progression of 1997 — certainly closer to 1997 than the 2009 progression — though 1997 probably gets the nod because the warm anomaly was slightly warmer, and the thermocline was slightly flatter. That being said, the above image is an entire monthly mean of May, 1997 — meaning it includes values from the end of the month, where the El Nino was already further along — whereas the May, 2014 image is simply a 5-day snapshot from May 3-8. Perhaps the very small differences are really only because of the time intervals in the comparisons — further indicating that May, 2014 and May, 1997 appears to be a great comparison.
However, in order for May, 2014 to have the same mean that May, 1997 did, we need to keep up with that extreme progression. While that is possible, and the ball is certainly rolling for a strong El Nino event, it may be unrealistic for the atmosphere to maintain 1997’s pace.
Considering it is the reversal of the easterly trade winds to westerly winds that begin an El Nino, we will further analyze the nature of westerly winds from the 1997 event. The above image shows wind anomalies along the Equatorial Pacific on the left, and SST anomalies on the right. As one moves further down, the months of the year progress.
1997 had an extremely potent burst of westerly winds in February, and another relatively potent one in April, which both played a major role in moving warm water eastward. Westerly surface currents will often help to enhance and favor these wind bursts, since there are thus no easterly currents to stop them. What really stands out, however, was an even stronger westerly wind burst that occurred in late May through mid June. This one occurred east of the first two, which keeps the “train” of warm water movement going, and allows it to move further east. As we can see in the SST anomaly chart, it was that westerly wind burst which really helped to get the El Nino going, as SSTs quickly skyrocketed; particularly in the eastern El Nino regions.
Let’s look at 2014, so far.
2014 had similarly strong westerly wind bursts earlier in the year and has even seen an eastward progression. No other Nino event that we know of has ever had the same intensity and location of these westerly wind bursts — except 1997, of course. This is an indication that the base state of the atmosphere certainly seems to be favoring westerly wind bursts to further generate a very strong El Nino event.
However, although more westerly wind burst events are likely down the road, that does not mean we should necessarily expect another one with the same strength as May, 1997.
1) The Southern Oscillation Index, which measures the surface pressure difference between Tahiti, and Darwin, Australia, has not been consistently negative, meaning the pressures at Darwin have been lower than at Tahiti. For an El Nino, you tend to want higher surface pressures to the west (Darwin), than to the east (Tahiti) because air tends to flow from high pressure to low pressure — meaning that a negative SOI would reinforce west-to-east currents. This is a pretty broad oscillation, and strong Cyclones near Australia in April made the SOI lower than it would have been, so this is not an overriding factor. That being said, 1997 already had a strong SOI drop at this point. The strong El Nino of 1982-1983 also had already seen a drop in the SOI.
2) 1997 was in the midst of a long-term warm phase of the Pacific Decadal Oscillation, while we are currently in a long-term cold phase. This Oscillation relates to the temperature of the water near the Gulf of Alaska — El Nino events and warmth in this area tend to go hand-in-hand because of the atmospheric circulation. Although we are currently in a short-term warm PDO regime, the long-term negative PDO phase we are in may yield nuances to the atmospheric circulation that prevent the PDO from skyrocketing as much as it usually does during El Ninos, which may limit the ceiling of the event.
This gives us the belief that this El Nino should finish stronger than the 2009-2010 event, but could fall short of the super El Nino of 1997-1998. However, the ball is already rolling enough for this El Nino to still have a solid chance to reach the +2.0C threshold for the “super” El Nino — but it just wouldn’t be as super as the 1997-1998 event. This also means that the El Nino may not stay east- based during its whole duration; another aspect that separated the 1997 El Nino event was the fact that although its +2.4C Nino region 3.4 trimonthly peak was impressive, it reached a peak >+4.0C in Nino region 1+2 (further east regions). If the next strong westerly wind burst is a bit weaker than the one in 1997, the atmosphere may not be able to successfully sustain an environment to maintain our extreme subsurface warmth further to the east, which may allow the El Nino to shift to the west a bit later in its life.
A very early forecast for this El Nino is for it to initially be an east-based high-end moderate to borderline strong El Nino during the summer, become even stronger during the fall and have a one-month peak of +2.3C in the Nino 1+2 region, then shift to a basin-wide event (as opposed to just being east-based), with a trimonthly peak around +2.0C in the Nino 3.4 region. Afterward, it should gradually weaken and shift to the west throughout the winter.
The potential is still there for this El Nino event to rival the 1997-1998 event considering how much the ball is already rolling and the extreme nature of the subsurface warmth that had only previously been seen in 1997 — but we need to see another strong westerly wind burst like in May, 1997 before we have confidence in going that direction. In about a month or so, we will post another El Nino update with a higher confidence forecast.
Thanks go to NOAA and the TAO Data Array for being the source of the majority of these images.