UPDATE (August 7, 2011): I’ve added a supplement to this post to cover a topic I excluded from the original. Refer to Supplement To “ENSO Indices Do Not Represent The Process Of ENSO Or Its Impact On Global Temperature”.
I’ve attempted to write this post at an introductory level for people who are not familiar with the El Niño-Southern Oscillation. For further introductory discussions, refer to An Introduction To ENSO, AMO, and PDO – Part 1. There are additional posts linked toward the end of that post that will also help those who are interested in learning about ENSO.
Because ENSO is a complex subject and because the post is written for those who aren’t familiar with it, the post is long, over 6500 words. It contains twenty-three illustrations, four gif animations and a video, so it will take time to run through it. It may also take a while to load due to the size of the animations.
Those who are familiar with my posts on the multiyear aftereffects of significant ENSO events will likely find this to be a rehashing of points made in the earlier posts, but for them, there are a few new discussions. Those readers should be able to find those new discussions by scrolling through the boldface headings and graphs.
A simple explanation of the El Niño-Southern Oscillation can be found at the NASA Earth Science Enterprise “For Kids Only” El Nino – An Introduction webpage. They write:
“El Niño, an abnormal warming of surface ocean waters in the eastern tropical Pacific, is one part of what’s called the Southern Oscillation. The Southern Oscillation is the see-saw pattern of reversing surface air pressure between the eastern and western tropical Pacific; when the surface pressure is high in the eastern tropical Pacific it is low in the western tropical Pacific, and vice-versa. Because the ocean warming and pressure reversals are, for the most part, simultaneous, scientists call this phenomenon the El Niño/Southern Oscillation or ENSO for short.”
Since El Niño and La Niña events can have drastic effects on regional temperature and rainfall around the globe, scientists monitor the strength and timing of El Niño and La Niña events using a number of indices. The Southern Oscillation Index monitors the pressure difference between the eastern and western tropical Pacific, and there are a number of indices based on the Sea Surface Temperature (SST) anomalies of regions along the equatorial Pacific. These ENSO indices are important weather forecasting tools, but they are often incorrectly thought to represent all of the effects of the ENSO Process.
More specifically, while the ENSO Indices allow us to monitor the strength and timing of El Niño and La Niña events, they only represent the effects of the ENSO process on the variables being sampled. That is, the Southern Oscillation Index is only an expression of the measured pressure difference between two points, and a Sea Surface Temperature-based index only represents the measured surface temperature of a specific region in the equatorial Pacific.
This post presents the ENSO process. It also presents the effects of the ENSO process on global sea surface temperatures during the satellite era that are captured by the ENSO indices and, more importantly, the effects that are not captured.
El Niño-Southern Oscillation (ENSO) indices are used to show the frequency and magnitude of El Niño and La Niña events. The Southern Oscillation Index represents the difference in Sea Level Pressure between Tahiti and Darwin, Australia. And there are a number of ENSO indices based on eastern equatorial Pacific Sea Surface Temperature (SST) anomalies. These include the Cold Tongue Index, the SST anomalies of the NINO3.4 region, and the Oceanic NINO Index, which is the 3-month average SST anomalies for the NINO3.4 region. Then there is the Multivariate ENSO Index that uses multiple factors to determine the strength and timing of El Niño and La Niña events. Global surface temperatures are noticeably impacted by El Niño and La Niña events. It is, therefore, common practice by some researchers to attempt to account for the effects of El Niño and La Niña events by determining the linear relationship between an El Niño-Southern Oscillation (ENSO) index and global surface temperatures. Researchers will then subtract the linear impacts of the ENSO index from the global surface temperatures and claim the remainder represents the effects of Anthropogenic Greenhouse Gases on global surface temperatures. This practice wrongly assumes that the ENSO index represents all of the effects of the ENSO process.
THE ENSO INDICES
As noted above, the Southern Oscillation Index represents the difference in Sea Level Pressure between Tahiti and Darwin, Australia. Because ENSO is a coupled ocean-atmosphere process, the Southern Oscillation Index can be used to illustrate the timing and strength of El Niño and La Niña events. It is in a form, however, that is not easily associated with variations in global surface temperatures, so we’ll exclude it from the rest of the discussions in this post. So let’s look at the Sea Surface Temperature-based ENSO indices.
Figure 1 shows the locations of the regions for commonly used Sea Surface Temperature-based ENSO indices. They both cover portions of the eastern equatorial Pacific. The NINO3.4 SST anomalies (and the three-month average of those temperature anomalies known as the Oceanic NINO Index or ONI) are the measured average SST anomalies for the area bordered by the coordinates of 5S-5N, 170W-120W. The Cold Tongue Index (CTI) represents the average measured SST anomalies for the area bordered by the coordinates of 6S-6N, 180-90W. During an El Niño event, warm surface and subsurface waters from the Pacific Warm Pool slosh east and the Sea Surface Temperatures rise along the eastern equatorial Pacific. During a La Niña event, tropical Pacific trade winds are stronger than normal and this exposes the cooler subsurface waters in the eastern equatorial Pacific, so the Sea Surface Temperatures drop in those regions.
Sea Surface Temperature anomalies of the eastern equatorial Pacific are therefore impacted by those El Niño and La Niña events. The variations in Sea Surface Temperature anomalies caused by ENSO events can be seen in the comparison graph of NINO3.4, Cold Tongue Index (CTI) and Oceanic Niño Index (ONI) SST anomalies, Figure 2. (The SST dataset used in this post are the satellite-based Reynolds OI.v2 SST data.) As illustrated, the variations between the three datasets are very similar in magnitude and timing. When the Sea Surface Temperatures of the NINO index regions rise above a threshold for a period of time, it qualifies as an official El Niño and when they drop below a threshold and remain there long enough, the La Niña is considered an official event. According to the NOAA definition, an “official” El Niño event occurs when the Oceanic Niño Index SST anomalies exceed +0.5 for 5 straight months, and, likewise, for the drop in SST anomalies in the eastern equatorial Pacific to be considered an “official” La Niña, the Oceanic Niño Index SST anomalies must remain lower than -0.5 deg C for 5 straight months. (Since the Oceanic Niño Index is a 3-month average of NINO3.4 SST anomalies, NOAA normally describes the threshold as 5 straight “seasons”, where, for example, Nov-Dec-Jan and Dec-Jan-Feb are considered two consecutive “seasons”. Refer to NOAA’s Oceanic Niño Index “Warm and Cold Events By Season”webpage.)
There is another ENSO index called the Multivariate ENSO Index (MEI). It is based in part on the Sea Surface Temperature anomalies of the NINO3 region of the eastern equatorial Pacific. The NINO3 region is bordered by the coordinates of 5S-5N, 150W-90W, so it’s east of the NINO3.4 region shown in Figure 1. But there are a number of other observations from the tropical Pacific that contribute to the Multivariate ENSO Index. These additional variables include sea-level pressure, zonal and meridional components of the surface wind, surface air temperature, and total cloudiness fraction of the sky. For more detail refer to the Multivariate ENSO Index homepage. The MEI is presented in standardized form; that is, the data is divided by its standard deviation. Even with all of these additional modifications, the variations in the Multivariate ENSO Index data are very similar to those of the NINO3.4 SST anomalies as illustrated in Figure 3.
Since there is little difference between the Multivariate ENSO Index and the Sea Surface Temperature-based indices during the satellite era, and to simplify the rest of this discussion, we’ll use the Sea Surface Temperature-based index NINO3.4 SST anomalies.
AN OVERVIEW OF THE ENSO PROCESS
1. A La Niña Recharges Ocean Heat– The warm waters that provide the fuel for an El Niño event are “stored” in an area of the Western Tropical Pacific called the Pacific Warm Pool. During La Niña and ENSO-neutral periods, warm surface waters are pushed to the west by the trade winds and accumulate in the Pacific Warm Pool. In effect, the trade winds “stack up” the warm water against the land mass of Indonesia. Since this warm water reaches depths of 300 meters, we cannot use Sea Surface Temperature to illustrate its variability during ENSO events. So we’ll use the NODC Ocean Heat Content (OHC) data, which represents in part the ocean temperature to depths of 700 meters. Figure 4 contains four maps that show the OHC anomalies in January 1995 (the peak of the 1994/95 El Niño), January 1996 (the peak of the 1995/96 La Niña), January 1997 (just before the start of the 1997/98 El Niño), and January 1998 (the peak of the 1997/98 El Niño). As discussed earlier, warm water from the surface and below the surface of the Pacific Warm Pool sloshes east during an El Niño. This causes the Ocean Heat Content in the western Tropical Pacific to fall and results in negative OHC anomalies in the western Tropical Pacific. And during the transition from El Niño to La Niña (and during the La Niña itself) some of the “leftover” warm water is pushed back to the west by the trade winds, which resume their normal east-to-west direction and strengthen during the La Niña. The strengthening of the trade winds during the La Niña also reduces tropical Pacific cloud cover. The reduced cloud cover allows more Downward Shortwave Radiation (visible sunlight) to warm the tropical Pacific to depth. The trade wind-driven equatorial ocean currents carry this warm water to the west where it “accumulates” in the Pacific Warm Pool. In this way, the warm waters in the Pacific Warm Pool in the Western Tropical Pacific are recharged by the La Niña for the next El Niño event.
The variations in the Ocean Heat Content anomalies of the Pacific Warm Pool are closely coupled to variations in an SST-based ENSO index. But it’s an inverse relationship, meaning when one falls the other rises and vice versa. Figure 5 illustrates how closely they are coupled. It’s a time-series graph that starts in 1995 to capture the events leading up to the very strong 1997/98 El Niño. It compares the Western Tropical Pacific Ocean Heat Content (in Gigajoules per square meter) to inverted NINO3.4 SST anomalies (in degrees C). The NINO3.4 data was inverted very simply by multiplying it by -1.0. When the NINO3.4 SST anomalies rise in response to an El Niño, the Ocean Heat Content anomalies of the Western Tropical Pacific fall because it is supplying the warm water that fuels the El Niño. And when NINO3.4 SST anomalies fall in response to a La Niña, the Western Tropical Pacific Ocean Heat Content rises, because some of the leftover warm water is returned there and because an increase in Downward Shortwave Radiation warms the tropical Pacific to depth and that water is also carried west to the Pacific Warm Pool.
2. An El Niño Discharges Ocean Heat– Figure 6 shows central Pacific Sea Surface Temperatures (not anomalies) during December 1996 (an ENSO-neutral month immediately before the 1997/98 El Niño), December 1997 (at the peak of the 1997/98 El Niño), and December 1998 (at the first peak of the 1998/99/00/01 La Niña). During the ENSO-neutral and the La Niña phases, the trade winds have pushed the warmer surface waters to the west. During the El Niño phase, an initial weakening of the trade winds allows the warm water in the Pacific Warm Pool to migrate east. The warm water (in excess of 28 deg C in Figure 6) covers a larger surface area during the El Niño. The larger surface area allows more evaporation to occur, and evaporation is the primary way that the oceans discharge heat.
Note: As illustrated and discussed, the El Niño phase is the truly anomalous phase. Trade winds reverse direction, warm water is transported from west to east, etc., during the El Niño. It could be argued that there really are two phases of ENSO, El Niño and other, since the La Niña phase is simply an exaggeration of “normal” conditions.
During an El Niño, cloud cover and precipitation accompany the warm waters eastward from the West Pacific Warm Pool. We can illustrate this by comparing our ENSO index (NINO3.4 SST anomalies) to the satellite-based Eastern Tropical Pacific (20S-20N, 180-80W) precipitation anomalies based on Climate Anomaly Monitoring System (CAMS) – OLR Precipitation Index (OPI)data. Refer to Figure 7. Note that the NINO3.4 SST anomalies have been scaled (multiplied by a factor of 0.4) for this illustration, and that both datasets were smoothed with a 13-month running average filter to reduce the noise in the precipitation data. As shown, Eastern Tropical Pacific precipitation rises and falls with the ENSO index. And during the 1997/98 El Niño, the Eastern Tropical Pacific precipitation increased significantly.
The Eastern Tropical Pacific Ocean releases heat during an El Niño, and this occurs primarily through evaporation. As the moisture-laden air rises, it cools, and as it cools, it can hold less moisture. The result is rain. By returning to a liquid state, the moisture in the air releases the heat that was used to evaporate it. This warms the atmosphere. We can illustrate the aftereffect of this by comparing the ENSO index (scaled by a factor of 0.5) and the Lower Troposphere Temperature (TLT) anomalies for the eastern Tropical Pacific, Figure 8. Lower Troposphere Temperature (TLT) anomalies are measured about 3,000 meters above sea level. This is yet another dataset that shows the closely linked relationship between ENSO and Eastern Tropical Pacific Temperatures.
3. Notes About The Response of Global Surface Temperatures to ENSO – There are studies that discuss in great detail how and why Global Surface Temperatures vary outside of the tropical Pacific during an El Niño event. Surface temperatures in some parts of the globe warm, others cool. This can be seen in the sequence of lagged correlations of surface temperatures with NINO3.4 SST Anomalies from Trenberth et al (2001) Evolution of El Nino–Southern Oscillation and global atmospheric surface temperatures”. Refer to their Figure 8, which is presented here as my Figure 9. They are a series of correlation maps for two periods. (Trenberth et al were illustrating the differences in the evolution of El Nino events between the two periods.) Areas shaded in reds indicate temperatures in those areas are positively correlated with the ENSO Index, meaning the temperatures in those areas rise when the NINO3.4 SST anomalies rise and drop when the NINO3.4 SST anomalies drop. And areas shaded in blues indicate the surface temperatures in those areas are negatively correlated with the NINO3.4 SST anomalies, meaning they drop when NINO3.4 SST anomalies rise and vice versa. The intensity of the shading indicates how well the variations in surface temperatures match the variations in the NINO3.4 SST anomalies on a statistical basis. The darker shades represent a closer correlation. The left-hand maps illustrate the lag correlations for the period of 1950 to 1978 and the right-hand column depicts the same for 1979 to 1998. I’ve highlighted the correlations at zero lag. Obviously, for global surface temperatures to rise during an El Niño event, the total warming of the areas that rise in temperature must exceed the total cooling of those areas that cool during the El Niño. The changes in surface temperature outside of the tropical Pacific during an El Niño event that are shown in Figure 9 are caused by changes in atmospheric circulation, not due to a transfer of heat. The processes that cause these changes are discussed in Trenberth et al (2001) and will not be repeated in this post. Refer also to Wang (2005) ENSO, Atlantic Climate Variability, And The Walker And Hadley Circulation.
NOTE: The Conclusions section of Trenberth et al (2001) includes the following statement (my boldface):
The main tool used in this study is correlation and regression analysis that, through least squares fitting, tends to emphasize the larger events. This seems appropriate as it is in those events that the signal is clearly larger than the noise. Moreover, the method properly weights each event (unlike many composite analyses). Although it is possible to use regression to eliminate the linear portion of the global mean temperature signal associated with ENSO, the processes that contribute regionally to the global mean differ considerably, and the linear approach likely leaves an ENSO residual.
In simpler words, the analyses performed in Trenberth et al only captured the global and regional variations in surface temperature (and other variables) that aligned with the major rises and falls in the ENSO proxy (NINO3.4 SST anomalies). And since there are regional responses to ENSO processes that do not follow those rise and falls, the authors are acknowledging that their analyses do not account for them. For example, suppose there were parts of the globe that warmed in response to both El Niño and La Niña events. They do exist, and they are not discussed by Trenberth et al, yet they play a major role in the rise in global Sea Surface Temperatures during the satellite era.
Back to the discussion of the response of Global Surface Temperatures to ENSO:
Animation 1 is taken from the videos in the post La Niña Is Not The Opposite Of El Niño – The Videos. I’ve used it in a number of posts since then. It presents the 1997/98 El Niño followed by the 1998 through 2001 La Niña. Each map represents the average Sea Surface Temperature anomalies for a 12-month period and is followed by the next 12-month period in sequence. Using 12-month averages eliminates the seasonal and weather noise. There are a number of things to note in Animation 1, including: The El Niño and La Niña events cause changes in the sea surface temperatures in the central and eastern tropical Pacific. We’ve illustrated and discussed the causes of these changes. Also, as the El Niño is taking place in the eastern tropical Pacific, note how the sea surface temperatures warm first in the Atlantic, then in the Indian Ocean, and then in the western Pacific. The warming is caused by changes in atmospheric circulation. [Refer again to Trenberth et al (2001) and Wang (2005).] And by the time these changes in atmospheric circulation make their way east around the globe to warm the western Pacific, the El Niño is transitioning to La Niña.
Figure 10 compares the ENSO index NINO3.4 SST anomalies (scaled by a factor of 0.08) to Global Sea Surface Temperature anomalies. The impacts of El Niño and La Niña events are obvious. During an El Niño event, global sea surface temperatures rise, and during a La Niña event, global sea surface temperatures drop. But notice that the global sea surface temperatures do not drop as far as the ENSO index during the La Niña.
The fact that the Global Sea Surface Temperature anomalies do not respond fully to the 1998/99/00/01 La Niña is easier to see if we smooth the data and align the two datasets, Figure 11. That difference between the scaled NINO3.4 SST anomalies and the global sea surface temperature anomalies is often attributed to anthropogenic global warming. In reality, the difference is caused by the portions of the ENSO process that do not correlate with the rises and falls of the ENSO index (NINO3.4 SST anomalies) and are, therefore, missed by studies like Trenberth et al (2001).
4. ENSO Redistributes Warm Water – We’ve discussed that during an El Niño event a significant volume of warm water migrates east from the Pacific Warm Pool and spreads across the central and eastern equatorial Pacific. We’ve also discussed that during the La Niña the trade winds resume their normal east to west direction and strengthen. The Pacific Equatorial Currents increase in strength and carry the warm water back to the Western Tropical Pacific. This is illustrated in Figure 12, which compares cross sections of equatorial Pacific temperatures at depth (not anomalies) for the December 1996 (an ENSO-neutral month immediately before the 1997/98 El Niño), December 1997 (at the peak of the 1997/98 El Niño) and December 1998 (at the first seasonal peak of the 1998/99/00/01 La Niña).
But there is another part of the ENSO process that returns warm water from the eastern equatorial Pacific to the western tropical Pacific, and it’s called a Rossby Wave. It can be seen in a video of Sea Level residuals from the Jet Propulsion Laboratory. I’ve highlighted the Rossby wave in screen captures from the JPL video in Figure 13. The upper right-hand cell shows the formation of the Rossby wave and the lower left-hand cell captures the Rossby wave travelling from east to west at approximately 10N, carrying leftover warm water back to the Pacific Warm Pool.
The Rossby wave can be seen in the first 10 to 15 seconds of Video 1.
THE ENSO INDICES ONLY REPRESENT SPECIFIC EFFECTS OF THE ENSO PROCESS
As discussed, the Sea Surface Temperature-based ENSO Indices only represent the Sea Surface Temperature anomalies of specific regions of the eastern equatorial Pacific. Nothing more, nothing less. When warm water from the Western Tropical Pacific migrates east during an El Niño, the SST-based ENSO indices only measure the effect the warmer water has on the sea surface temperatures of that specific region. Likewise, when the trade winds increase in strength and push the warmer water back to the Western Tropical Pacific, they expose the cooler subsurface waters in the eastern equatorial Pacific, and the SST-based ENSO Index measures the decrease in the surface temperature in the specific region.
It is understood that global temperatures rise and fall in response to El Niño and La Niña events. We illustrated that in Figures 10 and 11. But the rises and falls in Global Surface Temperatures in response to ENSO occur primarily in the Eastern Pacific Ocean. Refer again to the correlation maps from Trenberth et al (2001), Figure 9. To illustrate this, we’ll compare scaled (0.22) NINO3.4 SST anomalies and the East Pacific Ocean SST anomalies from pole to pole (90S-90N, 180-80W). If we account for the impacts of the volcanic eruptions on the Sea Surface Temperature anomalies of the East Pacific Ocean, Figure 14, we can see that Eastern Pacific SST anomalies rise and fall with the ENSO index.
A Note About The Adjustments For Volcanic Aerosols: The method I used to account for the impacts of the major volcanic eruptions of El Chichon in 1982 and Mount Pinatubo in 1991 was discussed in the post Sea Surface Temperature Anomalies – East Pacific Versus The Rest Of The World, under the heading of Accounting For The Impacts Of Volcanic Eruptions.
SUMMARY SO FAR
We’ve illustrated that many variables in the Tropical Pacific are closely coupled through the process of ENSO and, therefore, the timing and magnitude of their variations agree with the changes in the ENSO Index. We’ve also discussed the ENSO processes that cause these variables to mimic the ENSO index. We’ve plotted and compared Western Tropical Pacific Ocean Heat Content, Eastern Tropical Pacific Precipitation, Lower Troposphere Temperature (TLT) anomalies in the Eastern Tropical Pacific, and the volcano-adjusted Sea Surface Temperature anomalies for the East Pacific from pole to pole. There are other ocean and atmosphere metrics in the tropical Pacific that are closely coupled, including trade wind strength and direction, total cloud amount, eastern equatorial Pacific Ocean Heat Content and thermocline depth, and Sea Air Temperature. But outside of the tropical Pacific (and outside of the Eastern Pacific from pole to pole), the relationships exist but the variations may not rise and fall with the ENSO Index.
Let’s add the volcano-adjusted SST anomalies for the Rest Of The World (90S-90N, 80W-180) to the graph. The Rest Of The World data includes the Atlantic, Indian and West Pacific SST anomalies from pole to pole. Refer to Figure 15. The variations in the East Pacific dwarf the variations for the Rest-Of-The-World (Atlantic, Indian and West Pacific) data. Also note how the volcano-adjusted SST anomalies for the Rest of the World respond to El Niño and La Niña events. After a delay of a few months, they rise proportionately to the major El Niño events of 1986/87/88 and 1997/98, but then they only drop back part way during the La Niña events that follow.
And between the El Niño events of 1986/87/88 and 1997/98 and between the events of 1997/98 and 2009/10, the SST anomalies for the Rest Of The World hardly rise, if they rise at all. This can be seen in Figure 16, which is a graph borrowed from the post Does The Sea Surface Temperature Record Support The Hypothesis Of Anthropogenic Global Warming?The linear trend of the Rest-Of-The-World data between the 1986/87/88 and 1997/98 El Niño events is slightly negative, and between the 1997/98 and the 2009/10 El Niño events, the trend for the Rest-Of-The-World SST anomalies is slightly positive at only 0.001 deg C per decade.
The East Pacific dataset represents approximately 33% of the surface area of the Global Oceans and its Sea Surface Temperatures have not risen over the satellite era. So the East Pacific SST anomalies do not contribute to rise in Global Sea Surface Temperatures since November 1981, Figure 17.
This means that the ENSO-caused upward steps in the Rest-Of-The-World (Atlantic, Indian and West Pacific) data are responsible for the rise in Global Sea Surface Temperatures, Figure 18.
It’s obvious that that an ENSO index like NINO3.4 SST anomalies does not represent part or parts of the process of ENSO after the significant El Niño events of 1986/87/88 and 1997/98.
ENSO INDICES DO NOT ACCOUNT FOR THE WARM WATER THAT WAS RELEASED FROM THE PACIFIC WARM POOL DURING THE EL NIÑO
Let’s examine the 1997/98 El Niño again to see if we can determine why the SST anomalies for the Atlantic-Indian-West Pacific Oceans shift upwards after significant El Niño events. In the next animation, we’ll be comparing Sea Surface Temperature Anomalies and Sea Level Anomalies for the Tropical Pacific. There is a temperature component to the Sea Level anomalies, so they, in part, represent the temperature from the sea surface to the ocean floor. The Sea Level anomalies will, therefore, capture subsurface processes that are missed by the Sea Surface Temperature data. Also keep in mind that the ENSO index (NINO3.4 SST anomalies) only represents the Sea Surface Temperature of a specific area of the equatorial Pacific. I’ve highlighted it again in Figure 19.
Animation 2 presents a series of Tropical Pacific Sea Level and Sea Surface Temperature anomaly maps from January 1997 to December 2001. It captures the impact of 1997/98 El Niño and the 1998/99/00/01 La Niña on those two variables. There are a number of things to note. The West Pacific Sea Level anomalies drop significantly during the El Niño, while the Sea Surface Temperatures do not. This implies that much of the warm water that supplied the El Niño came from below the surface of the Western Tropical Pacific, from the area known as the Pacific Warm Pool. Also, as discussed earlier, during the transition from El Niño to La Niña, a Rossby wave forms in the northeastern tropical Pacific, at about 10N. The Rossby carries leftover warm back to the Western Tropical Pacific. And, as long as the La Niña conditions remain in the East-Central equatorial Pacific, the Sea Level and Sea Surface Temperature anomalies remain elevated in the Western Tropical Pacific. Keep in mind that the trade winds strengthen during the La Niña, which results in a decrease in tropical Pacific cloud cover, which results in an increase in Downward Shortwave Radiation. The increased Downward Shortwave Radiation (visible sunlight) helps to maintain the elevated Sea Level and Sea Surface Temperature anomalies in the Western Tropical Pacific. And all of that warm water has to go somewhere.
Let’s expand the view for the next animation, and present the 1997/98 El Niño and 1998/99/00/01 La Niña events a different way. We’ll see how those ENSO events combine to create the upward step in the East Indian and West Pacific portion of the Global Oceans. Animation 3 presents a series of global sea surface temperature anomaly maps. The maps and time period are the same as those presented in Animation 1. To the right of the maps is a comparison graph of NINO3.4 SST anomalies and the SST anomalies of the East Indian and West Pacific Oceans (60S-65N, 80E-180). The data in the graph are smoothed with a 12-month running-average filter that corresponds to the maps and it fills in with time.
As illustrated, the East Indian-West Pacific SST anomalies drop slightly as the El Niño is evolving. Then the East Indian-West Pacific SST anomalies start to rise as the changes in atmospheric circulation work their way east around the globe. As the conditions in the Eastern Tropical Pacific transition from El Niño to La Niña, the leftover warm waters are returned to the Western Tropical Pacific. Ocean currents there (western boundary currents) carry the warm water poleward, especially to an area east of Japan called the Kuroshio-Oyashio Extension or KOE. The East Indian-West Pacific SST anomalies peak mid-year in 1998 but do not return to the level they were at before the El Niño. The SST anomalies in the East Indian-West Pacific Oceans have been kept at an elevated level by the La Niña. In short, the upward shift in the East Indian-West Pacific SST anomalies is a function of ENSO. It is just as much of a function of ENSO as the rise and falls in the Sea Surface Temperature anomalies of the NINO3.4 region. The only difference: the NINO3.4 SST anomalies respond to each El Niño and La Niña event, while the upward shifts in the East Indian and West Pacific SST anomalies occur when a strong La Nina event follows a strong El Niño event. In that respect, the East Indian-West Pacific SST anomalies would be an imperfect ENSO index.
A Note About Subsurface And Surface Waters– This quick discussion may help explain what you’ve just witnessed. Before the El Niño, warm water has accumulated to depths of 300 meters in the Pacific Warm Pool in the Western Tropical Pacific. Refer back to Figure 6. The warm water that is below the surface is not included in measured Sea Surface Temperature, obviously, because it is not on the surface. As the warm water sloshes east during the El Niño, the warm water that had been below the surface is now on the surface, covering a greater surface area, and Tropical Pacific sea surface temperatures rise. During the transition to the La Niña and during the La Niña itself, the warm water that had been on the surface in the Eastern Tropical Pacific is returned to the west. Some of the warm water is stored again at depth in the Pacific Warm Pool, but some of it remains on the surface, creating what some have called a ratcheting effect on the surface temperatures of the East Indian and West Pacific Oceans.
Can That Portion Of The ENSO Process Be Captured By ANY ENSO Index?
Figure 20 compares the East Indian-West Pacific SST anomalies to the SST anomalies of the Atlantic-Indian-West Pacific (Rest-Of-The-World data) discussed earlier. The same upward shift occurred in response to the 1986/87/88 El Niño and 1988/89 La Niña. As shown, much of the variability of the Atlantic-Indian-West Pacific (Rest-Of-The-World data) data, including the upward shifts, is explained by the ENSO-caused variations in the East Indian-West Pacific data.
There is an important difference, however. In Figure 16, I illustrated that the linear trend for the Rest-Of-The-World (Atlantic-Indian-West Pacific) SST anomalies were flat between significant El Niño events. Let’s remove the El Niño-related months from volcano-adjusted East Indian-West Pacific SST anomalies, and for consistency between the two graphs, we’ll use the same method used in Figure 16, which was discussed in Does The Sea Surface Temperature Record Support The Hypothesis Of Anthropogenic Global Warming?As illustrated in Figure 21, the linear trends between the 1986/87/88 and 1997/98 El Niño events and between the 1997/98 and 2009/10 El Niño events are negative. It means the ENSO-induced rises in the East Indian-West Pacific SST data are decaying with time as one would expect.
That means that there has to be another ocean basin with positive linear trends between those El Niño events in order to create the flat trends of the Rest-Of-The-World (Atlantic-Indian-West Pacific) shown in Figure 16. And of course that ocean basin is the North Atlantic, with its well-known additional mode of variability called the Atlantic Multidecadal Oscillation or AMO, Figure 22.
Figure 23 illustrates the North Atlantic SST anomalies and the linear trends between the significant El Niño events. And again, for consistency between the graphs, I’ve isolated the El Niño-related months using the same method as Figures 16 and 22. The linear trends between the significant El Niño events are positive as anticipated.
A QUICK NOTE ABOUT THE NORTH ATLANTIC
The North Atlantic SST anomalies also show process-driven upward steps in response to ENSO. Animation 4 is similar to Animations 1 and 3. The series of maps are the same, but in Animation 4, the graph compares scaled NINO3.4 SST anomalies to North Atlantic SST anomalies during the 1997/98 El Niño and 1998/99/00/01 La Niña. Note in the maps how the tropical North Atlantic warms in a delayed response to the El Niño. As discussed in Wang (2005) ENSO, Atlantic Climate Variability, And The Walker And Hadley Circulation,the Northern Tropical Atlantic SST anomalies warm during an El Niño because of the changes in atmospheric circulation caused by the El Niño. During the El Niño, trade winds weaken in the tropical North Atlantic. This reduces evaporation and causes the Sea Surface Temperatures to rise. The weakening of the trade winds in the Tropical North Atlantic also causes less upwelling of cool waters from below the surface and this enhances the warming. What I found interesting in the animation: Notice how the warm water created by the El Niño appears to get swept into the Gulf Stream during the La Niña. This would mean that the trade winds in the Tropical North Atlantic grow stronger during the La Niña, and the warm water that had been created during the El Niño was pushed into the Gulf Stream where it gets carried north. And as illustrated in Animation 4, the North Atlantic SST anomalies do not drop proportionately during the La Niña.
A BEGINNING TO A MORE REALISTIC VIEW OF ENSO?
In the 2010 paper “Removing ENSO-Related Variations from the Climate Record”, Compo and Sardeshmukh open their abstract with (my boldface):
An important question in assessing twentieth-century climate is to what extent have ENSO-related variations contributed to the observed trends. Isolating such contributions is challenging for several reasons, including ambiguities arising from how ENSO is defined. In particular, defining ENSO in terms of a single index and ENSO-related variations in terms of regressions on that index, as done in many previous studies, can lead to wrong conclusions. This paper argues that ENSO is best viewed not as a number but as an evolving dynamical process for this purpose.
It appears that Compo and Sardeshmukh have made a good start toward treating ENSO as a process and not noise, but, from what I can gather, they missed the significance of the “leftover” warm water from the El Niño and therefore failed to account for its contribution to the rise in Global SST anomalies.
WHY IS THIS IMPORTANT?
We’ve presented and illustrated that there are many interrelated ocean-atmosphere metrics in the tropical Pacific that vary in response to ENSO events and that the variations in those metrics agree in timing and magnitude with an ENSO index. We’ve presented and illustrated that global surface temperatures do NOT agree with the ENSO index because there are additional ENSO processes that are not represented by an ENSO index.
Yet there continue to be climate studies and blog posts that portray ENSO solely as noise in the global surface temperature record. As Compo and Sardeshmukh put it, those other studies treat ENSO as a number and not a dynamical process. An example of such a paper is Fyfe et al (2010) “Comparing Variability and Trends in Observed and Modelled Global–Mean Surface Temperature”.Those studies scale and shift an ENSO index so that the month-to-month wiggles in the ENSO index align with the wiggles in the global surface temperature data. They use statistical models to find the best fit between the wiggles of the ENSO index and the global surface temperature data. Then they subtract the scaled and shifted ENSO index from the global surface temperature data and incorrectly claim the difference is caused by anthropogenic greenhouse gases and other anthropogenic forcings. We know the claim is incorrect because we’ve see that the East Pacific Ocean is the only ocean basin where the SST anomalies actually follow the variations in ENSO index during the satellite era, and we know it’s incorrect because we’ve watched what caused the upward shifts in the Rest-Of-The-World data.
Some might think that the fact that some climate studies misrepresent ENSO indicates that climate models do not represent ENSO properly. The climate models require anthropogenic forcings to cause sea surface temperatures to rise, when in reality, much of the rises in SST are driven by the process of ENSO. For those interested, I’ve illustrated how poorly the IPCC models hindcast and project Sea Surface Temperature anomalies in the following two posts:
A short summary of those two posts is, the IPCC multi-model mean data have no bases in reality.
The Sea Surface Temperature-based ENSO indices, the Sea Level Pressure-based ENSO index, and the Multivariate ENSO Index represent the impact of ENSO events on the measured variables, nothing more, nothing less. They are useful for determining the frequency and magnitude of ENSO events and for forecasting the short-term impacts of ENSO events on global weather. But ENSO Indices cannot be used to determine the impact of ENSO events on global surface temperatures, because ENSO indices do not represent the ENSO process or the impact of ENSO on the coupled ocean-atmospheric processes.
The Reynolds OI.v2 SST data is available through the NOAA NOMADS website:
The NODC Ocean Heat Content data, CAMS-OPI precipitation data, and UAH MSU TLT data are available through the KNMI Climate Explorer, which also served as the source for the maps used in Animations 1, 3 and 4:
The maps used in Animation 2 are available through the AVISO website: