Wednesday, June 19, 2019

Ambiguous SSTAs, Wind Patterns Could Render PDO Ineffective for Fall

A combination of ambiguous sea surface temperature anomalies (SSTAs) and wind patterns over the Pacific mean that the Pacific Decadal Oscillation (PDO) could be rendered for a little while as too ambiguous to use in seasonal forecasting as we move into the fall. Click on any image to enlarge it.

Graphic showing SST anomalies (shaded) and surface wind patterns (arrows) during the positive phase of the PDO (left image) and the negative phase (right image).
Source: University of Washington
The Pacific Decadal Oscillation comes in two phases: a “warm” phase (positive phase) and a “cool” phase (negative phase). The state of the PDO is identified primarily by the alignment of sea surface temperature anomalies over the Pacific basin. When SSTAs are notably below normal east of Japan into the waters south of Alaska, the PDO is said to be in the positive phase. In contrast, when anomalies are above normal in those same areas, the PDO is negative. This seems upside-down, so it’s helpful to also look at the anomalies immediately offshore the western coast of North America. Indeed, in a positive PDO those coastal waters exhibit positive SSTAs, while a negative PDO typically brings colder waters.

Let’s see how the PDO looks currently.
SST anomalies over the globe, as of June 17, 2019.
Source: NOAA

A look at the Pacific basin as of the last graphic of SST anomalies doesn’t provide much of a clear-cut direction of where the PDO currently stands. While there are cooler than normal sea surface temperatures wrapping into the waters offshore of the western United States – typically indicative of a negative PDO – there is a mass of colder than normal waters south of the Aleutian Islands and extending west towards Japan – typically indicative of a positive PDO. Additionally, there’s a swath of above-normal SST anomalies extending from Hawaii into the Gulf of Alaska, which doesn’t fit comfortably into either composite of the two PDO phases, making the picture quite muddy.

However, also shown on that two-panel image above are surface wind patterns exhibited during the different PDO phases. In a positive PDO, surface winds over the north Pacific flow from west to east, while in a negative PDO these surface winds flow from east to west. Let’s take a look at recent surface wind patterns in the Pacific to try and decipher the PDO.
Surface winds over the Pacific basin, from April 1st through June 16th.
Source: ESRL
Over the months of April, May, and the first half of June, surface winds were seen flowing from west to east over the waters south of the Aleutian Islands, a prominent mark of the PDO being in the positive phase. Contrasting with this, however, is a channel of northerly winds just offshore western North America, which the composite graphic at the top of this post indicates is associated with a negative PDO. Much like the SSTA comparison, surface wind patterns seem to be giving us conflicting signals that make it difficult to draw out which phase the PDO is actually in.
There’s one more method I want to look at to determine the state of the PDO.
Seasonal correlation of 500-millibar geopotential heights with the PDO in the months of April through June.
Source: ESRL
Shown above is an image that provides valuable insight on to how the PDO affects the atmospheric pattern. It is a seasonal correlation image, and although that sounds daunting, its interpretation is rather straightforward. Suppose, for a moment, we assume there is a positive PDO in place. The graphic above takes that information and asserts, based on history, that the positive PDO will result generally in positive 500-millibar height anomalies over the western swath of North America – in other words, a ridge. Why? Because for all the warm colors in this graph, 500-millibar heights are positively correlated to the PDO’s state. This means that in a negative PDO, all areas under warmer colors would see lower 500-millibar heights. Similarly, colder colors on this chart mean that if the PDO is in a positive (negative) phase, 500-millibar heights will be lower (higher) in areas with colder-color shading, indicative of troughs (ridges).

Let’s see if we can use this image to determine where the PDO is now.
500-millibar geopotential height anomalies from April 1st through mid-June. 
Source: ESRL 
Unfortunately, even this method doesn’t provide much more clarity on the state of the PDO. During the April-through-mid-June time period, we have seen below-normal geopotential heights (stormy weather) south of the Aleutian Islands into Japan. Using the image immediately prior to the one directly above, colder colors are draped over that same area, implying the PDO state is opposite the 500-millibar height anomalies over the northern Pacific. This would tell us that a positive PDO is present. However, moving into the Gulf of Alaska, persistent ridging has taken place over this time period, in an area that also has a negative correlation with the PDO. Therefore, that ridging off the western coast of North America implies a negative PDO. These are the same takeaways we found by analyzing SSTAs and surface wind patterns, and don’t really improve our understanding of what phase the PDO is actually in.

One could make an argument that it’s easier to just go by what the NOAA’s PDO index itself actually says, which indicates we are in a marginally-positive phase. In some situations, that’s certainly fine to do, but as was shown extensively in this article, the atmosphere is not totally reflective of a positive PDO state, meaning it would be misleading to just use the index value. The ambiguity in the current state of the PDO makes me question if it can be reliably used in seasonal forecasting for the fall months, and possibly for the winter months if these conflicting signals continue into late summer and early fall. Seasonal forecasters should take heed of the ambiguity and weight longer-term teleconnections accordingly.


Andrew