Normally, I’ve been segmenting data according to ocean and hemisphere or some portion thereof. I needed a break from the series I’ve been doing on the tropics, so I decided to divide the Globe into quadrants, Figure 1, and see what oddities came to light when I plotted the SST anomalies for the four areas.
As you’ll note, the simplistic approach did a reasonable job of isolating ocean data sets. There’s some overlap–Pacific with Atlantic, Atlantic with Indian, etc.–but it’s a starting point to see if there would any reason to carry an investigation further. The quadrants also include their respective portions of the Arctic and Southern Oceans, so they’re not solely those major oceans. The findings do justify that I put time into creating full ocean data sets and present them in a future post.
Note: All graphs in this post are of monthly long-term data sets, 1854 to 2008. The comparison graphs have been smoothed with 37-month filters, while the graphs of single data sets use 12-month filters. The NINO3.4 data in Figure 3 has not been smoothed.
SST ANOMALIES BY GLOBAL QUADRANTS
Figure 2 shows the SST anomalies for four segments of the globe:
– Far West – 90 to 180W (Purple)
– Near West – 0 to 90W (Red)
– Near East – 0 to 90E (Green)
– Far East – 90-180E (Blue)
The big eye catcher is the 6- to 7-year lag of the Far West SST anomalies during the late 1800s/early 1900s drop in SST. The Eastern Pacific makes up the majority of this area.
There were 4 moderate El Nino events in 1896/97, 1899/1900, 1902/03, and 1904/05, Figure 3, with only 2 mild, slightly counteracting La Ninas in 1898/99 and 1903/04.
The relative strengths of those El Ninos versus the La Ninas at that time are better illustrated by the NINO3.4 data that’s been smoothed with a 37-month filter. Refer to Figure 4. Could these El Ninos have delivered sufficient heat to the East Pacific to keep its SSTs from dropping at the same time as the other data sets?
Figures 5 through 8 are graphs of the individual data sets for the four quadrants smoothed with 12-month filters. They’re being provided as reference. All four quadrants display the typical drop in SST anomaly from the late 1800s to the early 1900s. SSTs rise to the 1940s in the Near West, Figure 6, and Near East, Figure 7, quadrants, then decrease to the mid-1960s, when then they rise again. The spike around 1940 in the Near East data (Figure 7) is an extreme example of the abnormal rise in SST anomaly in the Indian Ocean and in parts of the Southern Hemisphere. Illustrating good portions of both hemispheres of the Pacific Ocean, the Far West, Figure 5, and Far East, Figure 8, data sets rise more gradually from 1910s to present, with a comparatively minimal rise in the 1940s.
Figure 9 illustrates the SST anomalies for the Far East and Far West quadrants. The delayed decrease in the Far West SST anomaly around 1900 is very clear in this illustration. The two data sets, while they do share common overall trends, appear at times to be out of synch, as if heat is shifting back and forth across the Pacific. Which it does.
During El Nino events, the Pacific Warm Pool (contained within the Far East data set) delivers heat to the NINO3.4 area (part of the Far West data set), where it is upwelled. Through oceanic Rossby waves, the warm surface water is sent east, where it rebounds off the Central and South American Coasts. The oceanic Rossby waves then travel North and South along the coasts and westerly in multiple directions. In addition, clockwise ocean currents in the North Pacific and counterclockwise ocean currents in the South Pacific return the heat to the West Pacific (Far East data set). Refer to the following video from NASA, especially the 97/98 El Nino that starts in January 1997 with the first of two Kelvin waves travelling west to east. Note: It’s a large video, 17mb.
Figure 10 illustrates the SST anomalies for the Near East (primarily the Indian Ocean) and the Near West (primarily the Atlantic). While they follow the same overall trends, there are substantial periodic differences that do not appear to be caused by ENSO.
Subtracting the Near East (Indian) from the Near West (Atlantic), creates a curve, Figure 11, that bears what appears, at first, to be a function of the Atlantic Multidecadal Oscillation (AMO).
But when the curve created by subtracting the Near East from the Near West data set is compared to the AMO, the differences become obvious. See Figure 12.
A comparative graph of the North Pacific Residual and “Near West Minus Near East” data shows a better correlation, but it’s not perfect. Also, it would be tough to explain.
What’s that cycle induced by? Maybe that will become obvious when I divide Global SST into ocean data sets, including both hemispheres.
Sea Surface Temperature Data is Smith and Reynolds Extended Reconstructed SST (ERSST.v2) available through the NOAA National Operational Model Archive & Distribution System (NOMADS).