In the following link, the National Weather Service Climate Prediction Center (CPC) provides a substantial amount of information, graphics, and data about the El Nino-Southern Oscillation (ENSO).
Under the heading of Historical is a link to data titled Equatorial Upper-Ocean Heat Content Anomalies(1979-present). There, they further qualify the data as “Equatorial Heat Content (average temperature in the upper 300 meters, deg C)”. With the longitudes listed (130E-80W, 160E-80W, 100-180W), the data are the average temperature anomalies of cross sections of the Equatorial Pacific. Refer to Figure 1 for a quick look at the longitudes of the data sets.
Figure 2 is a comparative graph of the three data sets of Average Temperature in the Upper 300 Meters of the Equatorial Pacific. The 100-180W data set, similar in longitude to NINO3.4 (120-170W) has the greatest variations of the three. Other than the differences in amplitude caused by coverage area, the data sets are remarkably similar.
ENSO has a great impact on the data. In A Different Way to Look at NINO3.4 Data I discussed the reasoning behind smoothing ENSO data with 2-year running-average filters: it captures a complete El Nino/La Nina cycle. Figure 3 contains the Average Temperature in the Upper 300 Meters of the Equatorial Pacific, but the three data sets have been smoothed with a 25-month filter. Note how from 1982 to 1991 there are two cycles that are fairly similar in amplitude and frequency. But then the cycles are interrupted in the mid-90s, prior to the increase to the 1997/98 El Nino. What caused the cycles to end?
Note also how, after the significant 97/98 El Nino and the subsequent multiyear La Nina, the 4+ year cycles have not reappeared. Curious. It’s a shame that longer-term data is not available.
COMPARISONS TO SST
Figure 4 is a graph of the SSTs for the same longitudes of Equatorial Pacific. As expected, ENSO dominates the data.
Figures 5, 6, and 7 are comparative graphs of the SSTs and subsurface temperatures for the three longitudes used by the CDC. The three graphs are very similar, so the following comment pertains to all three. Note how, prior to the significant 82/83 and 97/98 El Nino events, subsurface temperatures rose sharply before SST did, and how subsurface temperatures plummeted before SST did during the transition to La Nina.
One the problems of adding linear trends to short-term data are selecting start and end dates. (This really isn’t a problem for me in this post, since the start and end dates used in the following are those of the data sets. I didn't pick them.) If the start point is in a cycle trough while the end point is at a cycle peak, the trend would be influenced upwards. But the subsurface temperature data does not necessarily show that the start and end points are in either a trough or a peak. If anything, they’re close to being the same value, but more on that later. Regardless of this and other factors that could influence the trends in these data sets, I felt it important to show:
First, how flat the trends were in what the CDC called the Equatorial Upper-Ocean Heat Content Anomalies for the Pacific, and…
Second, that the linear trends for these data sets were actually negative.
Figures 8 and 9 are graphs (with linear trends) of Equatorial Pacific SST and subsurface temperature for the majority of the Pacific Ocean (130E-80W, Figure 8) and the data set (100-180W, Figure 9) that’s nearest to the coordinates of the NINO3.4 area. The SST anomalies for the majority of the Pacific, Figure 8, shows a slight increase, while indicating a slight decrease in subsurface temperature. Note that this decline occurred even though the temperature at the end of the subsurface data is slightly higher than the start temperature.
For the slice of the Pacific that nears the longitudes of the NINO3.4 area, Figure 9, both the subsurface temperature and SST show negative trends.
Figure 10 is the same comparative graph as Figure 1, showing the three data sets of Average Temperature in the Upper 300 Meters of the Equatorial Pacific, but in Figure 10, I've also shown their associated linear trends.
The magnitude of the ENSO events impacts the ability to determine the actual trends, so I’ve exploded the temperature scale in Figure 11.
The decadal trends in what the CDC calls the Equatorial Upper-Ocean Heat Content Anomalies for the Pacific are approximately:
130E-80W………………. -0.015 Deg C/Decade
160E-80W………………. -0.06 Deg C/Decade
180W-100W…….……… -0.08 Deg C/Decade
The source of the subsurface temperature data is included above in the opening paragraph.
Sea Surface Temperature Data is Smith and Reynolds Extended Reconstructed SST (ERSST.v2) available through the NOAA National Operational Model Archive & Distribution System (NOMADS).