Monday, February 24, 2020

Mid-Winter 2019-20 Recap (A Little Late)

I am late in posting this as we are in the middle of the February rating period so on-air duties have come first.

So far the winter has been much warmer than we indicated in our winter outlook back in late October. What did we miss?  Why did the seasonal outlook end up being "milder"? To answer this we will look at the many drivers of the winters and compare what they were indicating in late fall versus what actually happened.

Most of the metrics describing the winter (temperatures, snowfall, etc) will come from the northern Ohio perspective. The other analyses will focus on other elements like continental and global drivers. My goal is to always give historical perspective. S

Let's start with Lake Erie ice cover:

Photo courtesy: Tom Horsman
Lake Erie ice coverage has been at historic lows (data since 1972). The ice coverage as of February 19 is now at 4.8% (daily data from GLSEA--Chart from Environment Canada) The only years close to 2019-20 on this date were 1998 (none), 2002 (7.1%), 2012 (1.7%), 2017 (2.2%)

What were the years with the lowest MAX ice levels? 1998 (8%), 2002 (10%), 2012 (14%) and this year (4.8%). Data from @NOAA_GLERL is daily data for 2020 and 2012. Other years' data is weekly ice data from Environment Canada.

How does this winter compare to last winter?

Rest of the Great Lakes are also showing little ice coverage vs last year

The winters with at least 10% or more with ice coverage over the last 10 years. We've had some lean years recently with some cold and icy Lake Erie periods like the winters of 2013-14 and 2014-15.

Here is are the seasonal snowfall maps starting with this winter 2019-20 all the way back  to the winter of 2013-14 for comparison

Another way of comparing winters is to use the AWSSI (Accumulated Winter Season Severity Index). It takes into account snowfall, extreme temperatures and other factors. Notice how low this winter has been even compared to last winter. Below is the winter 2019-20 vs recent winters, winter of 2004-05 (snowiest winter in northern Ohio), winter 2013-14 (one of the coldest recent winters) and the coldest winter (1977-78)

This winter started out very promising with early November snowfall:

November temperatures ended up below normal across the eastern US. Notice the ridge of high pressure over the Gulf of Alaska. We thought this would be the precursor to the overall winter pattern...

Ridge over GOA started to break down. Bermuda high retrograded west. Temperatures in early December started to trend up.

Low over Alaska/western Canada developed in mid-December. Temperatures continued above normal across much of the US.

Our overall winter outlook highlighted temperatures above normal for December.  The holiday season across Ohio was one of the warmest on record. Warm was similar to the same period in 1982

The northern Pacific pattern (EPO - East Pacific Oscillation) is highly connected to our temperatures in northern Ohio. Notice how it was below normal in November and early December.  Then the EPO started a sharp increase as the northern Pacific ridge started to shift.

We're we worried about the warm December temperatures/lack of snow pointing to a warmer winter overall?  Nope.  At this point, we anticipated the pattern to gradually shift to more sustained periods of cold.  We highlighted these locations where we analyzed ocean temperatures. This configuration would be more favorable (among other factors) for a weakened Polar Vortex and high pressure "Blocking" over Greenland.

This November/December temperature set-up happened during the winter of 2014-15 so it wasn't without precedent. Notice that we had a very mild winter then the bottom fell out that winter in January and February.

So what actually happened in January this year?

We had a weak, warm El Nino signature (still do as of late February). Notice the warmth below centered near the dateline with cooler water in the ENSO 1.2 region (closer to South America).

 The dateline warmth increased slightly over the last 2 months (see graph below)

BY ITSELF (key words here), the placement of ENSO warmth during El Nino winters can play a significant role on the type of winter (cold or not-so cold)

Based on the past warm/neutral ENSO years (AGAIN BY ITSELF), January and February looked colder than normal (top row). But that didn't happen.....

The ridge in the northern Pacific became MUCH stronger in the first half of the month. The effects downstream were pronounced: Above normal temperatures eastern 2/3 of US

Why did this happen? The MJO (Madden Julian Oscillation) played a big factor as the MJO was in favorable high amplitude phases for cold initially (November) then warmth in January and became a large driver of the warmth.

November: MJO was in a favorable COLD phase

The MJO went around the dial in January with high amplitude/phases 4-5-6 which favored warmth

Eventually, the high pressure settled over the central US. The result was continuous above normal temperatures. A bubble of "cooler" air in the central US started to pop up as a panhandle storm track started to develop by the end of the month

The pattern started to change in early February. The ridge in the northern Pacific was strong similar to early January as was the Bermuda high. However the placement of the ridge centers made all the difference in the storm track. The jet stream (center box below) steered storm systems out of the panhandle of Texas with actual storm origination in the central Pacific. A weak northwest track across the northern states shifted the track further east into the Ohio Valley.

As of this writing, we have had 6 panhandle storm systems cross the Ohio Valley

The storm track throughout the winter has been on target. (Image courtesy Larry Cosgrove)

In summary here's the globally ocean temperatures comparing October to late January:

The cooler water in the eastern Indian Ocean started to warm #1 below. The ENSO region near the dateline (#2 below) remained consistent along with cooler water near South America (#2 below).The Gulf of Alaska warm pool cooled slightly (#3 below). The Indian Oceanic Dipole (IOD) which was at record positive levels in late fall (this significantly influenced the Australian fires) started to fall which is common as the water in the eastern Indian Ocean started to warm.

This takes us to the arctic where the cold air has been locked up overt the North Pole. The Polar Vortex hasn't become dislodged for long periods. The polar jet has been strong.

The Stratospheric Vortex (upper atmosphere) has also been strong. It's been sitting and spinning in a uniform pattern.

In years when the Polar Vortex is weak, cold air outbreaks are more frequent and stronger.

We look for rapid warming in the upper atmosphere and sharp wind reversals. Last winter (2018-19) the temperatures rose at 50 hPa and the zonal winds dropped and shifted direction (see below). 

The jet stream was strong in spots in December 2018. The jet started to buckle in January which resulted in well below normal temperatures across the Great Lakes in January 2019

This winter (2019-20) the 50 hPa temperatures stayed at near record low levels with zonal winds at near record high levels.  

Notice the difference between last winter and this winter. The jet stream was strong in spots in December just like 2018. Unlike January 2019 when the jet started to buckle, the jet in this winter intensified shutting down the cold air across the US.
January 2020 temperatures

Look at the surface pressure over the arctic. Lows around the North Pole. Ridges below in the lower latitudes.

Also no Greenland Blocking (high pressure ridge). In fact the exact opposite. 

The teleconnections (AO/Arctic Oscillation and NAO/North Atlantic Oscillation) have been in a dominant positive phase. Notice the LOW over Greenland similar to this winter

POSITIVE Arctic Oscillation

Dr. Judah Cohen from AER has a great blog in which he describes the arctic behavior each winter and the variables that influence its changes week to week. Here are his thoughts on the feedback mechanism that has kept the Polar Vortex so strong from his February 10th entry:

"...But the incessant stretch of positive to strongly positive AO since late December, based on the polar cap geopotential height anomalies (PCHs) plot originated in the polar stratosphere with cold/negative PCHs/strong PV back in mid December that propagated to the surface over a course of two weeks (see Figure i). And ever since, the two (strong PV – positive AO) have been locked in an infinite loop. Cold/negative PCHs/strong PV favor a positive AO and a positive AO with the lack of high latitude blocking/zonal (straight west to east) jet stream favor a strong PV. And to be honest there is no built-in circuit breaker to stop this infinite loop other than internal/natural variability where wave breaking can help to disrupt the PV (for such an example see Leet et al. 2019)."

He continues (and echoes what I have surmised as a possible driver (IOD and MJO) of the strong PV):

"...Arctic sea ice extent is also not conducive to disrupting the Polar Vortex. For much of the winter the greatest negative extent anomalies have been on the North Pacific side of the Arctic and not the North Atlantic side. This asymmetry in sea ice anomalies favors a strengthened PV and not a weakened PV, at least based in some modeling experiments (e.g. McKenna et al. 2017). Please don’t interpret this last point as trying to attribute the strong PV/positive AO solely to Arctic forcing. It seems highly plausible to me that the positive Indian Ocean dipole and strange behavior of the Madden Julian Oscillation (stuck in an infinite loop of its own between phases 4-6) have played an important role in the positive AO winter, especially given that it was predicted by the models this past fall. I

The overall results for this winter in northern Ohio and a large portion of the US has been one of the warmest winters on record:

3rd warmest in northern Ohio since 1932

Snowfall across Ohio Valley and Great Lakes vs normal

Here are the metrics we use in comparing seasons. Data is for Hopkins Airport. All data found HERE

We forget the very mild winter of 2015-16 and some of the snowy winters between 2010 and 2015 along with the early 2000s winters which featured more snow than the winters of the late 1970s!


We are left with several questions. Why did the winter end up much warmer than our October winter outlook indicated?

1) How much did the Indian Ocean (IOD) shift influence the MJO during the winter influence the strength of the northern hemispheric (polar) jet and the stability of the arctic?

2) How much--if any--did the central weak, neutral "El Nino" play a role?

3) How much did Climate Change influence this winter?

4) How much did the Arctic sea ice play a role?

5) Any other variables?

If you have any research on these ideas and would love to shed light on these question, contact me on Twitter (scottsabolfox8).  Thank you in advance!