Back in a July 28, 2008 post, Polar Amplification and Arctic Warming, I used a comparative graph of Global, Northern Hemisphere, and Arctic lower troposphere temperature (TLT) anomalies to illustrate how Arctic TLT diverged from the other two after the 1997/98 El Nino. Refer to Figure 1.
In this post, I’ve segmented the Arctic (60N-90N) TLT anomaly data to further illustrate the impacts on the Arctic of not only El Nino events but of volcanic eruptions as well. Most if not all the Arctic warming since 1979 should be explainable with responses to natural events.
In the post The Lingering Effects of the 1997/98 El Nino, I employed a graph, Figure 2, of East Indian/West Pacific SST anomalies, comparing it to NINO3.4 SST anomalies. The Sato Index of stratospheric mean optical thickness, used to illustrate volcanic activity, was also on the graph, but it did not come into play in the time period of that post. Let’s take a closer look at that graph. During the aftereffect periods of the two volcanic eruptions, there was little visible response of the East Indian/West Pacific SST anomalies to the El Nino events, but the responses to the 1986/87/88 and 1997/98 El Ninos are remarkable. The delay of the rise of East Indian/West Pacific SST anomalies to the increase in NINO3.4 SST anomaly is a little over a year for the 86/87/88 event and about 6 months for the 97/98 event. Then notice that in both cases the East Indian/West Pacific SST anomalies do not respond to the decrease in NINO3.4 SST anomalies; they decrease gradually (It takes about 2 years for the East Indian/West Pacific SST anomalies to decrease 50% of the rise caused by the El Nino.) until swept upward by the next El Nino. Let me reword that: There are significant step changes in the East Indian/West Pacific SST. The East Indian/West Pacific SST anomalies increase approximately 0.2 deg C in response to a 1 deg C rise (trough to peak) in NINO3.4 SST anomaly, but there is little to no reaction to the subsequent La Nina events, just a gradual decrease. And since the East Indian/West Pacific SST anomalies had not returned to their pre-El Nino values before the next El Nino event struck, East Indian/West Pacific SST anomalies rose in steps.
The High Latitude North Atlantic Ocean (50N to 65N) had a similar but more extreme response. Refer to Figure 3. In the High Latitude North Atlantic, there was no response in 1982 to the El Nino, because the El Chichon eruption was the dominant natural event. That makes sense, considering El Chichon’s Central American location. The El Chichon eruption appears to have established a new “setpoint” from 1983 to mid-1986 for the High Latitude North Atlantic SST anomalies until they were swept upwards by the 86/87/88 El Nino, with a 1-year-plus lag. There they began a gradual decrease until they dropped again in response to the Mount Pinatubo eruption. Almost two years later in 1993, while the Mount Pinatubo aerosols are subsiding, the High Latitude North Atlantic SST anomalies increased sharply, most likely in response to the multiyear El Nino event taking place. The 1997/98 El Nino and the subsequent El Nino events in 2002/03, 2004/05, and 2006/07 drove High Latitude North Atlantic SST anomalies higher.
The curve of the High Latitude North Atlantic SST anomalies, Figure 3, will reappear in the following discussion. The smoothed versions of NINO3.4 SST anomalies and the Sato Index, Figure 4, are used in most of the comparative graphs of Arctic TLT anomalies. I’ve changed their scaling in each graph to better illustrate the responses of the Arctic areas being discussed. The scaling factors used for each of the graphs are in parentheses. The smoothing alters the appearance of the NINO3.4 SST anomaly and Sato Index curves, but it is necessary because I had to filter the TLT data to smooth the noise.
Figure 5 illustrates the coordinates of the Arctic areas discussed in the following. I apologize for the lack of clarity, but I reduced the size of the map to make it fit on the page. Unfortunately, that made things fuzzy.
In retrospect, I most likely could have combined the East Siberian and the Siberian subsets and combined the Alaskan and the Canadian subsets, but I was anticipating a relationship between North Pacific SST anomalies with the East Siberian and Alaskan TLT anomalies that did not pan out.
Let’s start with the Canadian dataset.
Figure 6 is a comparative graph of Canadian TLT, scaled NINO3.4 SST anomalies, and scaled volcanic aerosol mean optical thickness. Canadian TLT responded rapidly to the volcanic eruptions in 1982 and 1991. Without the volcanic reference it would appear that the Canadian TLT opposes the changes in NINO3.4, but looking at the responses to the 86/86/88 and 97/98 El Nino events, it’s clear they respond in a similar direction but lag by 12 to 18 months. There are two peaks in recent years, 2001 and 2006, that do not appear to be in response to ENSO events. They may be responses to spikes in atmospheric pressure or reversals of Arctic Ocean currents or another variation in the Arctic. (I could investigate those two spikes, but this post is about Arctic response to ENSO and volcanic eruptions.)
Greenland TLT, scaled NINO3.4 SST anomalies, and scaled volcanic aerosol mean optical thickness are shown in Figure 7. Greenland TLT has a quick and somewhat repeatable reaction to volcanic eruptions. It responds little to the El Nino event of 86/87/88, but does respond to the La Nina that follows in 88/89. Greenland TLT responds in a skewed fashion to ENSO events following the 97/98 El Nino and appears to be following the decline in NINO3.4 SST anomalies over the past few years. The two anomalous spikes in 84/85 and 95/96 appear to be linked to the volcanic eruptions and not the ENSO events, since a similar spike doesn’t occur after the 97/98 El Nino event.
A NOTE ABOUT A COUNTERINTUITIVE ARCTIC RESPONSE TO VOLCANIC ERUPTIONS
In “Volcanic Eruptions and Climate”, Alan Robock explains the expected impacts on global temperatures: “It has long been known that the global average temperature falls after a large explosive volcanic eruption. The direct radiative forcing of the surface, with a reduction of total downward radiation, cools the surface.” “Volcanic Eruptions and Climate” (6.6Mb) by Alan Robock can be found here: http://climate.envsci.rutgers.edu/pdf/ROG2000.pdf
Robock further explains the impact explosive volcanoes have on the stratosphere: “After the 1982 El Chichon and 1991 Pinatubo eruptions the tropical bands (30S–30N) warmed more than the 30N–90N band…producing an enhanced pole-to-equator temperature gradient. The resulting stronger polar vortex produces the tropospheric winter warming…”
In Chapter 2 of AR4, (page 195), the IPCC describes this winter warming further: “Anomalies in the volcanic-aerosol induced global radiative heating distribution can force significant changes in atmospheric circulation, for example, perturbing the equator-to-pole heating gradient…and forcing a positive phase of the Arctic Oscillation that in turn causes a counterintuitive boreal winter warming at middle and high latitudes over Eurasia and North America…”
Are the two anomalous spikes in 84/85 and 95/96 in Greenland TLT, Figure 7, indications of this?
Volcanic aerosols again are dominant over ENSO events in the graph of Alaskan TLT, scaled NINO3.4 SST anomalies, and scaled volcanic aerosol mean optical thickness, Figure 8. Alaskan TLT did not respond with a major rise to the 86/86/88 El Nino, but increased gradually from 1985 to 1989. It did, however, respond with a spike to the off-season rise in NINO3.4 SST anomalies in 1993 (a minor El Nino) and with lagged spikes to the 94/95 and 97/98 El Ninos. From 2002 to early 2008, Alaskan TLT shifted above the NINO3.4 SST anomalies, but in 2008, most of the divergence has disappeared.
SCANDINAVIA PLUS & SIBERIA PLUS
I’ve combined the Scandinavian and Siberian TLT data into one comparative graph with scaled NINO3.4 SST anomalies, and scaled volcanic aerosol mean optical thickness. Refer to Figure 9. Both Scandinavia and Siberian TLTs respond to the volcanic eruptions, but they, at first glance, appear not to react to ENSO events. The Siberian TLT response appears to counter NINO3.4 SST anomalies at times. It also has what appear to be post-volcanic-eruption spikes in 1983 and 1995. Scandinavian TLT appears to have slow saw-toothed variations in the 1980s. Scandinavian TLT then slowly declines until the 97/98 El Nino raises it and Siberian TLT. Note also how Scandinavian and Siberian TLTs oppose each other at times.
If I replace the NINO3.4 and Sato data with High Latitude North Atlantic (50N-65N) SST anomalies, Figure 10, the High Latitude North Atlantic SST anomalies appear to drive the underlying trends in Scandinavian and Siberian TLT.
While East Siberian TLT, Figure 11, reacts quite significantly to volcanic eruptions, it appears to oppose NINO3.4 SST anomalies during periods unaffected by volcanic eruptions. It’s as though East Siberian TLT was reacting to the Pacific Warm Pool, which feeds heat to ENSO events.
High Latitude North Atlantic SST anomalies also appear to drive the underlying trends of East Siberian TLT. Refer to Figure 12.
A LAST REFERENCE GRAPH
For those wondering, the curve of the High Latitude North Pacific SST (50-65N) anomalies, Figure 13, does not appear to be similar to any of the High Latitude (Arctic) TLT anomaly subsets.
The preceding should help show that recent Arctic warming could be considered natural responses to other natural occurrences, including ENSO events and volcanic eruptions. The underlying trends of the Scandinavian, Siberian, and East Siberian TLT subsets appear driven by High Latitude North Atlantic SST anomalies, which are also impacted by ENSO events and volcanic eruptions, in addition to Thermohaline Circulation (not discussed in this post).
The ERSST.v2 data and Arctic TLT Data (MSU-AHU) are available at the KNMI Climate Explorer website.