Could a changing climate mean more Sandy tracks?

The effects of a changing climate on our weather have been debated for years. With each extreme weather event, more people are left wondering whether it can be attributed to climate change or not. After all, many scientists have stated that a changing climate would mean more extreme weather events. However, it is irresponsible to say that every extreme weather event can be attributed to climate change, because that would imply that there was no extreme weather before climate change began.  Remember, the weather is inherently “extreme”, and our daily averages and means are derived from extremes on both sides of the spectrum: hot and cold, wet and dry, and windy and calm. We do not hover around our average highs and lows every single day.

That being said, there may be a link to a changing climate and more extreme weather phenomena, such as the extreme track that “Sandy” took in late October, 2012 — which devastated the lives of many people, and with whom many shoreline communities are still recovering from.

The theory behind this relates to a phenomenon called Arctic Amplification. This relates to the fact that in a changing climate, the arctic regions will warm more rapidly than latitudes further south. This is because of climate feedbacks that we discussed in our recent article, regarding Arctic Sea Ice and how we are not seeing nearly as much loss in ice extent as we saw last year. Essentially, the arctic warms faster than latitudes further south, because a loss of ice leads to a chain reaction that creates additional warming to the initial warming effect. The chain reaction is the feedback, called the “snow/ice albedo” feedback. Snow and ice are much better at reflecting heat and solar radiation back into space than water is — water will tend to absorb more heat. The ability to reflect the heat back into space is called albedo. In warming the Arctic, the ice and snow melts, creating more water. This means that there is more area of low albedo and less area of high albedo. Thus, more heat will be absorbed into the Arctic regions, further yielding warming. Because the feedback allows for more warmth without actually adding heat to the system, the warming of the Arctic is “accelerated” and amplified. Thus, the Arctic will heat faster than the lower latitudes. (In case you are wondering, the Antarctic will not warm to this same extent, because the change in ocean currents, among other things in a warmer climate, will actually yield some cooling to “cancel out” some of the warming).

A quick illustration of how Arctic Amplification affects the temperature gradient between the Equator and Arctic. The temperature gradient was a lot stronger before climate change really took off.

Figure 1a: A quick illustration of how Arctic Amplification affects the temperature gradient between the Equator and Arctic. The temperature gradient was a lot stronger before climate change really took off. Image credit goes to

Since we know that the Arctic warms faster than other latitudes, scientists are now trying to arrive at how this relates to our weather and makes it more extreme. Dr. Jennifer Francis of Rutgers University published a paper last year, along with Dr. Stephen Vavrus, detailing the possible connections. They first confirmed the fact that the Arctic is warming faster than other latitudes and then used that fact to theorize, and then test, that winds would slow down at the 500mb level. The research suggests that because the Arctic is warming faster than other latitudes,  the temperature gradient between the Arctic and the Equator decreases. It is this gradient that helps to form winds from 500mb and higher in the atmosphere — the jet stream being included. Reducing this gradient would thus weaken these winds, and the study confirmed a weakening in the winds — with the weakening being most prevalent in October through March (Sandy occurred in late October). Visuals of the decreasing temperature gradient between the Arctic and Equator can be seen in Figures 1a and 1b.



Now that the Arctic has warmed more than the Equator, the temperature gradient decreases, thus leading to a slower jet stream.

Figure 1b: Now that the Arctic has warmed more than the Equator, the temperature gradient decreases, thus leading to a slower jet stream. Image credit goes to

This may seem a bit counter-intuitive at first — a weak jet stream leading to strong storms taking unusual paths does not seem to make sense. But diving down a bit deeper, it all fits. What Francis also studied in her paper was how a slower jet stream means a more amplified jet stream — one that has deeper ridges and troughs. A good analogy is this: if you have a river flowing very rapidly, it is much more likely to stay in that straight line than it is to diverge in any direction. However, if you slow down the flow of the river, it is much more likely to veer from that straight line, and perhaps dip or rise. In this analogy, the wind is the river. Thus, the slower jet stream and winds at the 500mb level are likely to dip and rise more than usual. When you have larger ridges and larger troughs with slower winds, weather systems slow down, since storms will follow the path of the ridges and troughs. So instead of a fast, straighter track, they will take slower, more wavy tracks. Since storms would meander in an area for a longer period of time, this would increase weather extremes. And this does not just go for storms — it goes for high pressure systems as well, and it is a slow, large area of high pressure due to a very amplified ridge that helped to steer “Sandy” into New Jersey.

The regions that seem to be most affected by these changes are the Arctic, Greenland, and NW Atlantic areas. An increase in warmth, combined with the slower, more amplified jet streams lead to warmer, “blocking” patterns in these areas, which can prevent strong storms from escaping out to sea. The strong ridging being sustained is what acts as the “block”. What’s also important to note from Francis’ findings is that ridge amplification in the Arctic is further enhanced. This is because if the Arctic is being warmed rapidly, it makes it easier for warmer 500mb heights to “stretch” into those regions, resulting in stronger ridges. This effect is in addition to the already increased amplification of the longwave pattern.

A 500mb composite anomaly generated from the ESRL, centered around the weather pattern during "Sandy". Note the strong ridge in the NW Atlantic. This helped to steer "Sandy" into NJ.

Figure 2a: A 500mb composite anomaly generated from the ESRL, centered around the weather pattern during “Sandy”. Note the strong ridge in the NW Atlantic. This helped to steer “Sandy” into NJ.

The weather pattern that occurred right around the time that Sandy hit illustrates these points well. Figure 2a shows the 500mb height anomalies from October 27-29, 2012. There are negative anomalies in the Central and SE US — which was a trough that helped turn “Sandy” northward to begin with (“Sandy” itself in this image is the “L” just off the SE coast). But what really stands out is the extreme, large positive anomaly covering the entire NW Atlantic and southern Greenland. The 500mb heights are nearly 400 meters above normal, which is absolutely unprecedented; and it covers such a large area. Remember, Francis’ research indicated stronger, more amplified ridging. This essentially acted as a brick wall that prevented “Sandy” from tracking any further north and east, which is what most hurricanes do when they make the turn north. For physical reasons that we will not get into too much, storms cannot just plow through ridges. With strong ridges, there are strong high pressure systems at the center of them, and winds flow clockwise around them. Thus, at the southern end of the ridge, there was an easterly flow (winds blowing from east to west) that “pushed” Sandy westward.

What also stands out is the strong negative anomaly to the south of the ridge. The negative height anomalies are generally associated with low pressure, in which winds flow counter-clockwise. Thus, on the northern side of the low pressure, we also have an easterly flow. There was a second area of easterly winds that pushed “Sandy” westward, in addition to the fact that the ridge to the north is a huge block. It got to the point where it was inevitable that “Sandy” was going to turn west.

500mb vector wind anomaly composite from the ESRL, right around the time of Hurricane Sandy.

Figure 2b: 500mb vector wind anomaly composite generated from the ESRL, right around the time of “Sandy”.

Figure 2b shows the 500mb vector wind anomalies, further illustrating the above points. The clockwise flow around the high pressure ridge and the counterclockwise flow around the low pressure trough combined to create a significant positive easterly flow anomaly, which helped to push Sandy to New Jersey. You can even visualize the ridges and troughs from this map — the upstream side of the ridge in eastern Canada and Labrador, peaking all the way in Central Greenland, then the downstream side of the ridge near Iceland and Western Europe.

It is this type of anomalous pattern that could result from climate change, based on Francis’ findings. There was also a huge blocking pattern in the extremely snowy winters in the eastern US during the 2009-2010 and 2010-2011 seasons, amongst other extreme weather events that have occurred recently. As Francis said in her paper, it is difficult to attribute any individual extreme weather event to such potential climate change phenomena, but these weather patterns and associated extreme storms with freaky paths are consistent with the findings in her work. As snow and ice melt in the Arctic continues, it is increasingly likely that more of our weather will be influenced by Arctic Amplification. That is not to say that every year is going to have extreme Sandy-like weather, since climate change occurs over such a large time scale. But what is perceived as abnormal may become a tad less abnormal in the coming years.

However, there is opposition to this theory. A paper recently published by Dr. Elizabeth Barnes attempts to refute the claims made by Francis. Her main points are:

1) Using a different method to calculate meridional (degree of north-south flow as opposed to east-west flow), the increases are not statistically significant. This different method also shows that ridge elongation is also not statistically significant.

2) With regards to the ridging into the Arctic, the 500mb height contour that is at the crest of the ridge is higher, but the ridge itself is not more amplified.

3) The wave speeds are not significantly lowered, especially if you also choose to analyze the 250mb level of the atmosphere.

4) Thus, it is insufficient to say that blocking events have increased in the past 30 years.

Francis has already strongly defended her claims in response to Barnes’ paper, and her comments can be read in this article from the Capital Weather Gang.

Although Barnes is obviously a very reputable scientist, I strongly take Francis’ side on this debate. Her comments in the article were enough to convince me; but to be objective, I took a look at the Barnes paper myself, to get a better context as to the points that Francis refutes in her comments.

As far as her initial point is concerned it is hard to say, without deeper knowledge and understanding of the subject, which method to calculate meridional flow is more accurate.  But even with the Barnes method, there was still a positive trend in meridional flow, which does not totally contradict Francis’ findings. Although the increases were not significant, the data are somewhat skewed by the fact that Barnes included years going all the way back to 1980 in her findings. Francis notes that most of the changes have occurred in the last fifteen years or so. When including a lot more years, a given change will appear to be less significant within that larger time period.

The NAO index since 1830. Note the peaks and valleys. Also note that the valleys have gotten lower and lower. This supports the fact that blocking has increased, potentially due to Arctic Amplification. Image credit goes to

Figure 3a: The NAO index since 1830. Note the peaks and valleys. Also note that the valleys have gotten lower and lower. This supports the fact that blocking has increased, potentially due to Arctic Amplification. Image credit goes to

This point can also be used to refute the fact that blocking has not increased. Francis’ research was not meant to say that Arctic Amplification was the only factor to increase blocking, but that it would help. Taking this fact into account, it is important to note that the North Atlantic Oscillation (NAO) — often a big indicator of blocking — was in a positive decadal phase throughout all of the 80s and 90s. During that time, there was not enough Arctic Amplification for effects to really take shape with regards to the phase of the NAO. Additionally, it is a bit inconsistent to compare “blocking” throughout two different decadal phases. It would have been much more prudent to compare the blocking we have seen now, compared to the last negative decadal phase. That way, you can better isolate the true impacts of Arctic Amplification. Once you do so, you will see that the NAO’s negative decadal phases have grown more negative in the recent past compared to earlier decades, which is supported by Figure 3a.

NAO values from the CPC since 1950. Note that most of the recent negatives have only occurred recently, and do appear a tad more negative than the previous negative decadal phase.

Figure 3b: NAO values from the CPC since 1950. Note that most of the negatives post 1980 have only occurred recently, and do appear a tad more anomalous than the previous negative decadal phase.

A time series of the NAO from the CPC (Figure 3b) confirms the fact that there was a strong positive phase of the NAO throughout the 80s and 90s, which would thus skew data in an elongated time series. This is the logic behind yearly means lagging the actual observations, and I think that is the problem with Barnes’ work. It also shows how much the NAO has dipped since Figure 3a was created.

Now, one could make the valid argument that the recent sample size is too short to make any definitive conclusions. But that is where the logic behind the physical science and meteorology comes in — when you take the recent findings in data, combined with what physically makes sense regarding jet-stream speed reduction, the jet-stream becoming more meridional, and Arctic Amplification, it is not unreasonable to speculate what the future may hold. You cannot say that what has currently occurred and what may occur in the future is insignificant, just because of the past, when the physical science supports the aforementioned claims. By using data going all the way back to 1980, you’re essentially just stating that what we have currently seen is only within a short sample size, not that it has not occurred.

With point #2, Francis makes a good, quick point that increasing the heights in the northern latitudes is the exact phenomenon that will help the ridges become more amplified to begin with. All of that said, I do not think Barnes disproved anything that Francis said, since Barnes did not contradict the fact that heights would increase in the northern latitudes.

With point #3, most of the figures in Barnes’ paper did indicate slower wave speeds, which supports Francis. Some time series were not significant, but there was still a noticeable decrease in most of them. The only figure that truly contracted Francis was when Barnes switched to 250mb. But 250mb can be highly variable in that it is way too close to the stratosphere, which experiences its own set of variables. 500mb is thus a better level to look at; you cannot just cherry-pick the data to support your hypothesis.

It seems to me that the Barnes paper had a pre-determined outcome, and Francis’ claims are better supported meteorologically. It may be fair to expect more extremes in weather in the coming years, along with more blocking episodes as Arctic Amplification is still occurring. What’s even more concerning is the fact that the feedback associated with Arctic Amplification strongly lag the initial warming. If we stopped all fossil fuel burning instantly, it would take several decades for the warming of the Arctic to stop, since all of the additional fresh water that was generated from the melting ice will still be absorbing more sunlight.

That being said, and I will emphasize this again: not every year is going to have record-breaking blocking and extremes because of this. There is still a lot that needs to be learned about Arctic Amplification and how it affects our weather. Not every individual extreme weather event can be attributed to climate change and Arctic Amplification. But on average, over a large time scale (perhaps decades), Francis’ findings do support more extreme weather — and it is fair to say that after further research, we generally agree with her findings.