Sunday, May 19, 2019

Long Range Outlook: May and June 2019

This post will discuss my thoughts on the outlook for the remainder of May and into June of this year.  Click on any image to enlarge it.

We begin with an analysis of the current atmospheric pattern across the Northern Hemisphere.
Five day (May 11 thru May 15) mean 500 millibar height anomalies.
Source: Japan Meteorological Agency (JMA)
The last several days have seen a rather volatile pattern positioned over the Arctic and similar upper latitudes. Most predominantly, there has been a trio of ridges pushing into the Arctic Circle, placed over Siberia, north-central Eurasia and far western Europe. The ridge placed over Eurasia has forced itself northward over the North Pole, disrupting the tropospheric polar vortex and splitting lobes of cold air to lower latitudes. Indeed, we have seen such an evolution take place, with strong negative geopotential height anomalies over the northern Atlantic Ocean, northeast Eurasia/Russia, and marginally in the Bering Sea and Canada.

This is a similar process as what takes place when the tropospheric polar vortex is disrupted in the winter, with colder than normal air being forced down to lower latitudes as ridging forces the polar vortex to be displaced from its usual resting place over the North Pole. Consequentially, colder than normal temperatures have indeed been observed in the lower latitudes, including here in the United States.

The movement of colder air into the United States has been driven more-so by a pattern resembling the positive phase of the Pacific-North American (PNA) oscillation, which is seen best by the presence of a ridge along the West Coast of the U.S., which encourages troughing and cooler air to shift southward from Canada into the central and eastern U.S. Additionally, a ridge near Greenland as of late has provided further impetus for the transport of colder air into the contiguous United States, a pattern which some of you may recognize as a set-up typical of the negative phase of the North Atlantic Oscillation (NAO). Again, both of these are patterns typically emphasized during the winter, but this provides a good example of how they can still function across seasons.

In summary, we're entering this forecast period knowing the recent atmospheric regime is tilted in favor of colder temperatures for the United States. Let's begin with the outlook itself.

Forecasted 500 millibar geopotential height anomalies from the ECMWF model (left) and GFS model (right) for the 8-10 day forecast period.
Source: Pennsylvania State University
 My goal for this outlook is to begin with computer model guidance and then expand outwards to look at individual oscillations to try and discern if these computer models might be missing something in their forecast, or may be right on target. In accordance with that framework, we'll begin with the projected 500 millibar geopotential height anomalies in the 8-10 day forecast period from the ECMWF model (the "euro" model; left panel) and the GFS model (the "american" model; right panel). In general, warmer colors portray ridging (calmer and warmer weather), while colder colors portray troughs (stormier and cooler weather).

Model guidance is in pretty good agreement over the evolution of the pattern over the United States into the end of the month, with both models suggesting a near-Rex Block formation along the western coast of North America. A Rex Block is traditionally exemplified by a ridge placed "on top" (to the north) of a trough. In this graphic, such a set-up would make it look like warmer colors would be positioned directly north of colder colors. While this sort of set-up isn't precisely laid out here, both models anticipate the development of zonal flow downstream of the Rockies, typically what happens downstream of a Rex Block.

Such a zonal flow and ridging over the eastern two-thirds of the country is rather intriguing given what is forecasted in the upper latitudes, with both models forecasting a continued disruption of the tropospheric polar vortex as well as a continuation of high pressure near Greenland. As discussed earlier, both of these factors encourage colder than normal air to propagate to lower latitudes.

So what's happening here?

Let's take a closer look at the projected pattern over North America. I noted how it seemed like a vaguely Rex Block-esque pattern was forecasted to set up over the western coastline of the continent. But we missed a key item when looking at the positioning of the ridge and trough along the coastline: the trough is forecasted to sit in the Southwestern U.S. In other words, this prompts a pattern downstream more typical of the negative phase of the PNA oscillation, a flip from the recent positive PNA phase laid out at the start of this post. In a negative PNA regime, the trough in the West U.S. encourages the development of a ridge in the East U.S., which is indeed what we're seeing in the model guidance projections.
This isn't too surprising when we consider that these factors that affect the PNA oscillation are upstream of the United States, where as the NAO and other factors more encouraging of cold air in this time period are either well north of the U.S. or are downstream of the country, limiting their impact.

As outlined, let's now expand on this a bit and look at forecasts for four key atmospheric oscillations below.

Forecasted states of the Pacific-North American (PNA) index, top-left; North Atlantic Oscillation (NAO), top-right; Western Pacific Oscillation (WPO), bottom-left; and the Eastern Pacific Oscillation (EPO), bottom-right.
Source: Earth System Research Laboratory (ESRL)
The projected state of the PNA oscillation is shown in line with what we extracted before viewing this forecast, where a recent positive-PNA regime would flip to a negative-PNA situation in the 8-10 day period. This ESRL forecast builds on that by anticipating the PNA oscillation to remain in negative territory until the turn of the month. Be sure to note the strength of this negative-PNA regime, as we'll return to that later on.

Moving to the NAO forecast, our earlier suspicions are again confirmed with the agency projecting this oscillation to remain in negative territory throughout the entire forecast period (indicating that ridging near Greenland will persist), albeit to a weakening degree by the end of the period.

The WPO and EPO phenomena can be noteworthy and impactful when they're strongly oriented in one way or another, but they are both projected to flip states multiple times and remain at relatively-modest strengths during the forecast period. As such, we will focus more on the PNA and NAO for this outlook.

We now have a pretty good idea of what is forecasted to evolve over the next ten days or so (feel free to scroll down to the end for a summary if it was a little tough to grasp), so let's now move into the two-week-plus forecast period.

Forecasted GFS ensemble 500 millibar geopotential height anomalies (left) and spaghetti plot (right) at the end of the forecast period (June 3rd).
Source: Pennsylvania State University
It is imperative to first recognize the degree to which this particular forecast graphic is imperiled as a consequence of its long time horizon. Rarely, if ever, do forecasts made sixteen days in advance actually come to fruition. Unfortunately, our computer models are simply not yet advanced enough and do not have enough real-time observational data to input to make such accurate forecasts. This is the caveat that sixteen days out is a very long time, and conclusions drawn from here are not to be fully invested in to.

The GFS ensembles project the pattern 16 days out to be pretty similar to the pattern seen developing by the 8-10 day period. Indeed, the GFS ensembles anticipate ridging persisting over the Arctic Circle but a trough along the West Coast will sustain a negative-PNA pattern for the United States. In other words, such output indicates warmer conditions may be expected for the southern portions of the country, relatively seasonal weather for the country's midsection, and perhaps seasonal to cooler-than-normal conditions in the northern slice of the country. Cooler than normal and stormy weather would be anticipated for the western half of the country, with storm systems shuffling into the Southwest only to eject into the Plains and form the basis for severe weather threats in the Central U.S.

--

I'm a believer in the idea that the atmosphere takes all kinds of oscillations and phenomena into account when creating the weather for a given area - the United States does not exist in a vacuum where a negative-PNA state automatically produces warmer than normal weather for the entirety of the eastern U.S., for example, or where a positive-NAO state automatically produces zonal flow over the country. I also believe, however, that some phenomena can impact certain areas to greater degrees than other phenomena based on the location and intensity of these phenomena. For example, while the negative-PNA state may not automatically produce a 'textbook' negative-PNA outcome, its position as being upstream of the United States means the country's weather is more likely to be dictated by what the PNA does than what the NAO (way off in Greenland) does.
These are not controversial beliefs by any means to any weather enthusiast, but it is worth stating because my outlook here will rely heavily on the latter belief. While my forecast for the first ~two weeks relies substantially on the evolution of a negative-PNA state, as discussed, I'll now shift the basis for my forecast to another indicator.

--

Consider the ESRL forecast for those four oscillations once again. Note how all of them seem to dwindle down to weak levels of whatever their respective states are forecasted to be. That makes for a tough forecast for the start of June, if four of the most impactful oscillations are expected to be too weak to make a significant impact. As such, I'll now briefly zoom out and review the broader atmospheric pattern currently in place across the globe: an El Nino.

Multivariate ENSO Index (MEI). Positive (negative) values indicate the presence of an El Nino (La Nina).
Source: Earth System Research Laboratory
Frequent visitors to this blog in past winters will readily recognize the importance of understanding the current state of the El Nino-Southern Oscillation (ENSO) phenomenon. The ENSO phenomenon is identified by observing sea surface temperature (SST) anomalies along the Equatorial Pacific. Cooler than normal water temperatures typically indicate the presence of a La Nina, while warmer than normal temperatures indicate the presence of an El Nino. Each of these states of the ENSO phenomenon broadly drive weather patterns around the globe, giving you an idea of just how material this is to any longer-term forecast.

Using the MEI metric above, we see that positive values have recently increased, indicating the presence of warmer than normal water temperatures in the Equatorial Pacific and, potentially, an El Nino event. For the purposes of this post, the June outlook will operate under the assumption that an El Nino is in place. What does that mean for the forecast?

Surface temperature anomalies during summers where an El Nino was observed.
Source: Earth System Research Laboratory
The ESRL provides a neat tool whereby the user is able to analyze a variety of atmospheric variables that can be broadly expected given an El Nino or La Nina, given if it is occurring in the winter or summer. This provides a substantial advantage here, as we can glean what a "typical" El Nino in the summertime will give out in terms of temperature anomalies. The outcome is rather impressive, as it lines up well with what had already been established in the forecast through the end of May. Indeed, a "typical" summertime El Nino results in cooler than normal conditions in the northern U.S. as well as in the Rockies, with a small area of stronger cooler-than-normal anomalies in the Southwest. This fits well with the idea of a negative-PNA set-up, where troughs and cooler weather persist in the Southwest into the far northern U.S. as the jet stream then curves upward over the Plains due to ridging in the Southeast, a consequence of the aforementioned West U.S. troughing.

As such, it may be prudent to begin the June forecast with a basic outline akin to a negative-PNA state, with cooler and stormier weather in the Western U.S. extending to the northern Plains, severe weather opportunities in the Central U.S., and broadly seasonal to warmer conditions in the Southeast. Now that we've zoomed out to examine the overarching pattern, it's time to return to individual oscillations. While the PNA, NAO, EPO and WPO all may be too weak by the start of June to dominate the broad pattern for the U.S., another oscillation looks to be strong enough at this point in time to shape the June outlook for the contiguous 48. This is the Madden-Julian Oscillation (MJO).

Forecasted state of the Madden-Julian Oscillation (MJO) from mid-May through the end of May from the GFS ensembles (left panel) and the ECMWF ensembles (right panel).
Source: Climate Prediction Center
The Madden-Julian Oscillation (henceforth MJO) is categorized using a phase space diagram. In this case, that means that the MJO can take one of eight different phases. Why so many? The MJO phenomenon is identified by the placement of convective activity along the Equatorial Pacific, similar to the ENSO phenomenon. However, with the MJO, the zone of monitoring extends across the entire Equator, not just in the Equatorial Pacific. Each phase of this oscillation represents a different location for convection along the Equator, with Phase 8 and Phase 1 typically seeing convection placed over and around South America and/or Africa, and Phases 4 and 5 corresponding to convection north of Australia. Each of these eight phases has been found to produce different impacts on weather patterns around the globe, including here in the United States. We'll use that to our advantage a little later on.

Both the GFS ensembles and ECMWF ensembles anticipate the MJO to move near or into Phase 2 by the beginning of June. The models differ as to the strength of the oscillation by this point in time, however, with the GFS expecting a notable Phase 2 state but the ECMWF bringing the oscillation weak enough to enter that circle in the middle of the graph, indicating that the MJO is too weak to be definitively placed into a certain phase. Model divergence like this beyond two weeks out is to be expected, but because both models have the MJO either approaching or being within the Phase 2 state, we will continue the June forecast under the assumption that the month will begin with the MJO in a Phase 2 state.

As I noted earlier, each of the eight phases of the MJO affects global weather patterns in their own ways. The hard work of a number of scientists has enabled us to create composites for variables such as temperature, precipitation and upper-air flow for each phase of the MJO, so that we can have an idea of what kind of weather a certain phase of the MJO will bring for any given month. Let's view the surface temperature anomaly composite for a Phase 2 MJO state during the month of June.

Surface temperature anomaly composite for a Phase 2 MJO state during the month of June.
Source: American Weather (americanwx.com)
Allow me to restate again the importance of recognizing that no one oscillation controls the weather pattern - indeed, the atmosphere can be thought of as a web of different phenomena, all of them impacted or imposing impacts someway, somehow by/on to others. As such, it's critical to remember that we are still trying to glean a general framework for the June forecast, rather than definitively identify what *will* happen. After all, if we knew what would definitively happen, I would be on a private island with an expensive yacht somewhere!

As the composite image above shows, a Phase 2 MJO state occurring in June has historically resulted in warmer than normal temperatures over the southern and central Plains, extending northeast through the Great Lakes and Northeast regions. Slightly below-normal temperatures have been seen in the Pacific Northwest in such a situation.
This brings about a pattern similar to the negative-PNA pattern that we have discussed rather extensively in this post, and seems to encourage the idea that the pattern forecasted to end May could continue into the opening portion of June.

500 millibar height anomaly composite for a Phase 2 MJO state during the month of June.
Source: American Weather (americanwx.com)
Indeed, when looking at the 500-millibar geopotential height anomalies for a June Phase 2 state of the MJO, a pattern similar to a negative-PNA set-up emerges, with slight hints of a trough in the Western US aligning with ridging in parts of the Eastern U.S. Note the upper level low in place near Greenland, though, which represents the positive phase of the North Atlantic Oscillation (NAO) and tends to encourage zonal flow & warmer conditions over the contiguous United States. It remains to be seen if this will transpire into June, as ESRL forecasts have the NAO maintaining a slightly-negative state into next month, which would then suppress ridging over the Eastern U.S. somewhat.

Beyond this point, model guidance becomes too unreliable to rely on any combination of oscillations, and as such the remainder of June will expect a pattern broadly in line with a summertime El Nino, with recognition that individual oscillations could (and likely will) easily change such a pattern as next month draws closer and model guidance becomes more accurate as to the state of oscillations.


Forecast Summary

- Western U.S.: Expect warmer than normal weather to shift to stormier- and cooler-than-normal in coming days, with this regime change complete by seven days out. This stormy and cooler pattern should persist through the end of May. Tendency for stormy and cooler weather should continue into the opening days of June, with this same pattern encouraged by the El Nino for the entire month. Individual oscillations and other discrepancies may change this forecast for June, however.

- Central U.S.: Expect a continued active weather pattern through the end of May, with ample severe weather opportunities as storm systems eject eastward from the Southwest and ride an upward-curving jet stream over the Plains. Temperatures will vary, with a tendency for cool weather in the north Plains and seasonal to warmer weather in the south Plains. Precipitation should be expected to be above normal.

- Eastern U.S.: Expect warmer than normal conditions to evolve through the end of May and into June before a retrenchment to broadly seasonal conditions as the negative-PNA pattern breaks down. Precipitation anomalies are not seen dramatically going either way, though the Southeast U.S. may be monitored for somewhat below-normal precipitation. Broadly seasonal to somewhat warmer than normal conditions are possible for June in the aggregate.

Andrew