A guest post by Bob Tisdale
UPDATE 1 (January 12, 2009):
In my extremely brief description of an El Nino event, I wrote, “…and a subsurface oceanic temperature boundary layer called the thermocline pushes the warm subsurface water to the surface.” My oversimplification may be misleading, and while it does not undermine the intent of this post, a better explanation is available in the following video from NASA Scientific Visualization Studio video titled “Visualizing El Nino”: http://svs.gsfc.nasa.gov/vis/a000000/a000200/a000287/a000287.mpg
If I rewrite that sentence in the future, it would read something to the effect, “During El Nino events, natural changes in atmospheric and oceanic conditions cause the warm water that was ‘contained’ by the Pacific Warm Pool to shift east along the equator. The warm subsurface water rises to the surface.”
h/t Gary for noting the poor wording.
NOTE: For those who are new to the subjects of El Nino events and sea surface temperatures, I’ve tried to make the following discussion as non-technical as possible without overlooking too many aspects critical to the discussion. It includes detailed descriptions of many of the processes that take place before, during, and after El Nino events. The period after an El Nino event is often neglected, but it holds the oceanic responses that are the most significant over multiyear periods.
Two things have always stood out for me in a graph of Global Sea Surface Temperature (SST). The first was the Dip and Rebound in the ERSST.v2 version of the Extended Reconstructed SST data from the 1800s to the 1940s. The link above discussed it in detail.
In Figure 1, I’ve boxed SST anomaly data for the period from 1854 to 1976 to indicate that, other than the dip and rebound and the temporary rise in the early 1940s caused by a multiyear El Nino, there really wasn’t a rise of any note in SST between the late 1800s and the period from the mid-1940s to mid-1970s. The ERSST.v2 data used in this post illustrates little to no change in SST anomalies from the one period (late 1800s) to the other (mid-1940s to mid-1970s).
Second: After 1976, Global SST anomalies appear to rise in three steps. It’s very visible if monthly SST anomaly data has been smoothed with a 37-month filter, Figure 2, or if annual data has been smoothed with a 3-year filter. Many people try to correlate those steps with variations in TSI, because they seem to coincide with solar cycles. They don’t, so those trying to make the correlation fail in their efforts.
Zooming in on the period from January 1976 to present, Figure 3, and changing the filtering from 37-months to 12-months do not eliminate the appearance of steps. Why did Global SST rise in steps after 1976?
Based on the title of this post, the rising step changes were caused by El Nino events, three in particular. The NINO3.4 SST anomalies from January 1976 to November 2008 are shown in Figure 4. Most people familiar with the recent El Nino-Southern Oscillation (ENSO) record could guess correctly that the 1997/98 El Nino event was one of the El Ninos that caused a step change. If the magnitude of El Ninos was the only factor, the second logical choice would be the 1982/83 El Nino, since it ranks a close second in terms of peak NINO3.4 SST anomaly. Yet that El Nino event did not create a rising step change in global SST anomalies, because another natural event had a greater impact on global climate.
A volcanic eruption. The El Chichon eruption of 1982 interrupted the normal heat distribution processes of the 1982/83 El Nino. Many persons understand and cite this on blogs. Few realize, though, that the 1991 eruption of Mount Pinatubo also interrupted a significant series of El Nino events. The Mount Pinatubo eruption didn’t occur at the same time as a singular El Nino event with monstrously high SST anomalies, but the string of El Ninos it influenced was significant in its length. “Full-fledged” El Nino events occurred in 1991/92 and 1994/95, with a minor El Nino occurring during 1993. At minimum, two of the early-to-mid 1990s El Ninos had their heat distribution processes altered.
Figure 5 is a comparative graph of East Indian-West Pacific SST anomalies, scaled NINO3.4 SST anomalies, and inverted Sato Index of Stratospheric Mean Optical Thickness data (used as a reference of volcanic eruption timing and intensity). The data in Figure 5 have been smoothed with a 12-month running-average filter. The step changes in the East Indian-West Pacific SST anomalies are quite obvious. The graphs included in the following discussions are edited versions of Figure 5. In the latter graphs, I have simply limited the years in view to the periods being discussed. The three periods (January 1976 to December 1981, January 1981 to December 1995, and January 1996 to November 2008) are also shown in Figure 5. The periods were divided in this way because, working backwards in time, the first period discussed (1996 to 2008) has been covered in an earlier post and is, therefore, easiest to explain, the second period (1981 to 1995) includes the two volcanic eruptions, and the third period (1976 to 1981) is what was left over. Note that the NINO3.4 and Sato Index data are provided to illustrate timing and timing only; they have not been scaled to suggest magnitude of cause and effect. I did not want to get into a debate about scaling.
In Figure 6, I’ve blocked off the area of the East Indian and West Pacific Oceans illustrated by the black curve in Figure 5 and in illustrations that follow. The coordinates are 60S to 65N, 80E to 180. It represents a significant portion of the world oceans, in the range of 25 to 30% of global sea surface from 60S to 65N.
Figure 7 is a comparative graph of the NINO3.4 SST anomalies, inverted Sato Index, and the SST anomalies for the oceans segments not included in the East Indian-West Pacific SST anomaly dataset above. These include the East Pacific, the Atlantic, and the West Indian Oceans contained by the coordinates 60S-65N, 180-80E. The East Pacific-Atlantic-West Indian Ocean data (red curve) is overlaid onto the East Indian-West Pacific data (the black curve in Figure 5) during the discussions that follow to show the interactions between datasets.
A final preliminary note: The filtering is used to reduce the visual impact of the noise within the datasets. It also affects (smoothes) the abruptness of the change in the Sato Index data when the volcanoes erupted. It has a minor visual impact, but it is something to consider when viewing the graphs that include the volcanic eruptions (Part 2). The impacts of the smoothing are shown in Figure 8.
I illustrated the cause of the step change AFTER the 1997/98 El Nino in a video posted on the thread titled The Lingering Effects of the 1997/98 El Nino. The YouTube link is here: http://www.youtube.com/watch?v=4uv4Xc4D0Dk
Take five minutes and watch the video. It will help to illustrate the phenomena taking place and the causes.
Note: In the graphs for the video, I used the Optimally Interpolated SST anomaly data (OI.v2). The monthly time-series data for it starts in November 1981, and since I wanted to cover the period starting in 1976 in this post, I had to switch datasets. The SST anomaly data used in the following discussion is from the Extended Reconstructed Sea Surface Temperature, Version 2 (ERSST.v2), available from the National Climatic Data Center (NCDC). It runs from January 1854 to present.
THE STEP CHANGE FROM 1996 TO PRESENT – A RECAP AND EXPANSION OF DISCUSSION
The SST anomalies for the West Indian-East Pacific Oceans from January 1996 to November 2008 are shown in Figure 9, along with scaled NINO3.4 SST anomalies and the final few years of the inverted Sato Index data. The Sato Index ends in 1999, but because there has not been an explosive volcanic eruption capable of lowering global temperatures significantly since 1991, its end in 1999 has no affect on the discussion.
Note: You may wish to click on the TinyPic link (While holding the “Control” key) to open Figure 9 in a separate window. That would eliminate the need to scroll back and forth. This discussion goes on for a full page of single-spaced text in MSWord form.
The Pacific Warm Pool, also known as the Indo-Pacific Warm Pool, is an area in the western equatorial Pacific and eastern Indian Ocean where huge volumes of warm water collect due to a number of natural processes (normally attributed to ocean currents and trade winds). The Pacific Warm Pool is visible in SST data and in subsurface ocean temperature data; the warm pool reaches down to depths of 300 meters. Figure 10 illustrates its location. Over decadal periods of time, it expands and contracts in area and increases and decreases in volume. http://i42.tinypic.com/2hdqydy.jpg
During El Nino events, natural changes in atmospheric conditions cause the warm water that was “contained” by the Pacific Warm Pool to shift east along the equator, and a subsurface oceanic temperature boundary layer called the thermocline pushes the warm subsurface water to the surface. The high SST anomalies in the eastern equatorial Pacific are known as an El Nino. It is a natural process that occurs at irregular intervals and magnitudes. The eastern equatorial Pacific SST anomaly data is divided into areas for monitoring purposes. Refer to Figure 11. These areas are known as NINO1, 2, 3 and 4. Global temperature responses to El Nino events correlate best with the SST anomalies of an area that overlaps NINO3&4 areas. That area is called NINO3.4. That’s the data set used in the following discussions.
Back to the discussion of Figure 9: The purple curve in Figure 9 shows the SST anomalies for the NINO3.4 area [5S-5N, 170W-120W] in the eastern Pacific. The data has been reduced in scale by a factor of 0.2 so that it doesn’t overwhelm the graph. During the 1997/98 El Nino event, NINO3.4 SST anomalies rose to their highest levels during the 20th century. Its impact is visible in the long-term and short-term Global SST anomaly data shown in Figures 2 and 3. It affected global and regional temperature and precipitation patterns in the short term afterwards.
That’s usually about the end of a discussion of the 1997/98 El Nino. The video showed, however, that other processes continue long after an El Nino event. Much of the heat that rises to the surface during the El Nino is then transported west by the equatorial ocean currents, recharging the Pacific Warm Pool for the next El Nino and heating the surface of the East Indian-West Pacific Oceans. It’s important to keep in mind that before the El Nino most of the warm water was below the surface, contained by the Pacific Warm Pool. Since it’s below the surface to depths of 300 meters, it is not a part of the calculation of global SST, or global temperature, for that matter. Then, after the El Nino, much of it is on the surface and included in the SST data. The resulting rise in the SST anomalies of the East Indian-West Pacific Oceans (the black curve in Figure 9) lags the change in NINO3.4 SST anomaly by a few months. As shown, East Indian-West Pacific Ocean SST anomalies reached their peak in 1998, but by that time, NINO3.4 SST anomalies had already dropped back to “normal” levels. Then the NINO3.4 SST anomalies dropped further, into the subsequent La Nina (Negative) levels, but the East Indian-West Pacific Ocean SST anomalies only dropped a portion of the amount they had risen, about one-half of it. And before the East Indian-West Pacific SST anomalies can slowly decrease fully to the levels they were at before the 1997/98 El Nino, NINO3.4 SST anomalies increase in 2000 and cause the East Indian-West Pacific SST anomalies to rise again. That’s the step change.
In summary, a large volume of warm water that was once below the surface of the Pacific Warm Pool was raised to the surface by the El Nino and distributed across the surface of the East Indian and West Pacific Oceans, causing SST anomalies to rise in that region. East Indian-West Pacific Ocean SST anomalies began to drop but had not had enough time to return to “normal” before the start of the next El Nino event, which swept them upwards again.
They are slowly returning to the levels they were at before the 1997/98 El Nino, but because they were “pushed” higher again and again by the El Nino events of 2002/03, 2004/05, and 2006/07, the return has taken more than a decade.
In Figure 12, I’ve added the SST anomalies for the East Pacific, Atlantic, and West Indian Oceans to the comparative graph. (It’s another graph you may want to open in a separate window.) The East Pacific-Atlantic-West Indian Ocean SST anomalies mimic the rise and fall of the NINO3.4 SST anomalies during the 1997/98 El Nino—to a point. Note how, during the La Nina that followed it, the NINO3.4 SST anomalies have dropped well below the levels they had been at before the start of the 1997/98 El Nino (highlighted with the blue line and arrows), yet the East Pacific-Atlantic-West Indian Ocean SST anomalies don’t follow the NINO3.4 SST anomalies below the level they had been at before the 1997/98 El Nino to any great extent; that’s another (but smaller) cause of the step change in Global SST anomalies after the 1997/98 El Nino. Then the East Pacific-Atlantic-West Indian Ocean SST anomalies follow the rise in NINO3.4 SST anomalies from 2000 to late 2002, the peak of the next El Nino. And, from 2003 to present, the SST anomalies for both of the major portions of the global oceans (red and black curves) “normalized” to levels near to one another, modulating back and forth as each area, at different time lags, responds to variations in NINO3.4 SST anomalies. These include the additional El Nino events of 2004/05 and 2006/07, and finally a substantial La Nina in 2007/08. Because of that La Nina, the East Pacific-Atlantic-West Indian Ocean SST anomalies (red curve) have dropped down close to the levels they had been at prior to the 1997/98 El Nino, but it has taken more than 10 years.
In Figure 13, the Global SST anomaly curve from January 1976 to November 2008 (same graph as Figure 3) has been annotated to indicate the causes of the step change. As illustrated and discussed in the preceding, the temperature rise resulted from the significant step response of the East Indian-West Pacific SST anomalies to the 1997/98 El Nino event–that was compounded by a similar response (but of lesser magnitude) to the 2002/03 El Nino—that was then “maintained” by the El Nino events of 2004/05 and 2006/07.
CLOSING TO PART 1
That’s enough for one post. In the second part, I’ll cover the two earlier periods. For a preview, simply scroll back up to Figure 5 and note the step changes during those two periods and the effects of the two volcanic eruptions. (Remember that the Sato Index data is only there to illustrate the timing of the volcanic eruptions.) I’ll also add another phenomenon that confirms the step changes caused by the El Nino events are drivers of global temperature anomalies.
Smith and Reynolds Extended Reconstructed SST Sea Surface Temperature Data (ERSST.v2) and the Optimally Interpolated Sea Surface Temperature Data (OI.v2) are available through the NOAA National Operational Model Archive & Distribution System (NOMADS).
The Sato Index Data is available from GISS at: