Long Range Outlook

2021-2022 HVW Winter Outlook

by

The leaves on the trees are rapidly covering the ground across the Hudson Valley.  The air is beginning to feel a bit more crisp in the morning, and the children just celebrated Halloween.  As our minds begin to turn toward more wintry thoughts, it must mean that it’s time for the HVW Winter Outlook once again.  Each fall season, we compile observational weather data from across the globe, and begin to compare it against historical weather conditions.  We use a series of metrics and data elements to project our best bet for the coming winter.  Based on that data, and the long range computer guidance… lets take a look at what this coming winter may hold for the Hudson Valley.

Winter Temperatures : Near to Slightly Below Average (-1.5° to +0.5° vs. Average)

Winter temperatures in the Hudson Valley are likely to be near to slightly below average.  For the second straight season, a weak to moderate La Nina is projected.  This combined with similarly warm waters in the northern Pacific, should lead to persistant ridging over the northeast Pacific Ocean.  This should result in a persistent trough in the eastern US, and that should allow frequent shots of cold air to infiltrate the Northeast US.  The core of the cold is likely to be in the upper midwest, but even our area could be slightly below average for the season as a whole.  The other point worth mention, is that we could see a faster start to winter than in previous years.  A colder than average December is appearing increasingly likely.

Winter Snowfall : Near to Above Average (100% – 125% of Average) 43″ to 54″

A winter projecting a peristent trough in the eastern US is also favorable for a snowier than average winter season.  The regular access to cold air, combined with a trend that favors storms to track along and up the east coast, sets the stage for multiple east coast snowstorms.  Major snow events are a bit random due to the combination of ingredients that are needed, and thus are difficult to reliably project in a winter outlook.  However, based on seasons that had very similar conditions, this coming winter appears likely to have several opportunities for significant snow events, that could result in slightly above average snowfall.  The seasonal average in Poughkeepsie over the last 30 years is 43.0″.  We project between 43″ and 54″ of snow this season.  For perspective, Poughkeepsie saw 50.4″ of snow last winter.

Month by Month

For a little more detail, here is each month’s temperature and snowfall projection, for a more detailed idea of how the winter may play out, month by month.

December

January

February

March

Methodology / Discussion

By now, many of you may have a fairly good idea of the factors and indicators we usually look at when compiling the data for the Winter Outlook.  Everything starts with the Sea Surface Temperatures or SSTs.

When we look at the SST pattern, we are looking for years that have a similar SST profile to the current and projected pattern for the winter.  You can see the current sea surface temperatures and how they compare to average.  We like to focus in on 3 specific areas when creating the winter outlook.  The area labeled ‘A’ is the tropical Pacific Ocean or the ENSO (El Nino Southern Oscillation).  This is where the terms “El Nino” and “La Nina” come from.  The slightly cooler than average temperatures hint at a possible weak ‘La Nina’ for the coming winter.  The area labeled ‘B’ is the northern Pacific Ocean, and that is key because of the amount of warm water in that part of the ocean over the past several years.  The warm water favors an atmospheric ridge over the western US, which also then favors a reflexive eastern trough.  Then lastly area ‘C’ is the northwest Atlantic, which is important for the same reason ‘B’ is.  The warm waters tend to favor ridging over the Atlantic, which could have influence on the winter storm track, as well as the potential for warm air to sneak up the east coast.

So we can see the current pattern in place, next we want to see if we expect any dramatic changes in the pattern for the coming winter.

The seasonal computer guidance suggests the ENSO region ‘A’, may cool a bit more, signaling a weak to moderate La Nina for the coming winter.  The northern Pacific ‘B’ warm pool is projected to hold, with some cooler than average waters developing along the North American coastline.  The Atlantic Ocean ‘C’ continues to be warmer than average.  As a whole, the computer data does not change too dramatically from the current conditions… the biggest factor being the additional cooling of the ENSO region.

So now that we have good agreement between the current SSTs and the computer model SSTs, the next step is the most labor intensive.  We go through over 50 years of data, and look for winters that most similarly reflect the conditions shown above.  We select the years that are the best fit, and those become our analog years, from which we will develop the winter outlook.  After researching the data, we have pulled 7 winters.  When we blend those 7 winter SSTs, it looks very similar to the computer model projection for the winter.

Those winters are: 1995-1996, 2000-2001, 2005-2006, 2008-2009, 2013-2014, 2017-2018 and 2020-2021.  When blending those 7 winters, the analog SST pattern looks impressively closer to the projected winter pattern by the model.  The logic being, if the conditions this winter are the same as these years in the past, then we should experience similar results.  So this season, in addition to the SST pattern, we brought in a few new tools.  Three additional factors we analyzed for similarities were:  Atlantic Tropical Seasons prior to the winter; QBO (Quasi-Biennial Oscillation) index; and warmer than average Octobers prior to the winter.

In the interest of time, we won’t detail all of them here, but we will go into more detail in later discussions.  For purposes of this conversation, what is important to understand is that for each of these criteria, we went through roughly the last 30 winter seasons, and reviewed each criteria for years that were similar to the conditions we are experiencing, or expecting this winter.  We selected those years, and compared them to the winters we identified through the SST method above.  Several winters appeared multiple times in the additional criteria, especially the 1995-1996, 2005-2006, and 2017-2018 winters.  The reason this is important, is because we were able to develop multiple layers of assurance that our analog years are accurate.

Conclusion

Seasonal forecasting is a very imperfect science.  It relies on a set of assumptions, that are built on a forecast.  More simply, a forecast based on a forecast.  We know what our current conditions are, because we can observe them.  But we cannot be sure that those conditions will be the same several months from now.  So we rely on a computer model forecast data to extrapolate if those conditions will be the same 2 or 3 months from now.  Then… based off that data, we then use our experience and understanding of the science to develop the winter outlook based on a variety of additional factors.  With that said… last season, we actually did OK…

The actual temperatures were about a half a degree out of the range we projected.  That could be attributed in part to climate change, because global temperatures 20 or 30 years ago were a degree or so cooler, which we try (but may not fully) take into consideration when developing the outlook.  But our snowfall projection of 100% to 133% of average (42″ to 56″) likely seems like a large range… but when you recognize that in 19-20 Poughkeepsie saw only 16″ of snow… it doesn’t seem like that large of a range.  We saw 50.4″ (120% of the annual average) of snow… when the annual average in Poughkeepsie was 42″ at the start of last season.

So there is certainly ability to project correctly in these winter outlooks, it just relies on there being no surprises in the projected data.  Will we see surprises this winter?  Almost certainly.  The big question will be, what kind of surprises?  Luckily for you… we’ll be here to enjoy the ride with you, and discover what the coming winter holds.  We hope you enjoyed the 2021-2022 Winter Outlook.  Thank you all for your continued support.

-Alex and Bill

2020 – 2021 HVW Winter Outlook

by

As we push toward December, and the unofficial start of the winter season, the most common question right now is, “What does this winter look like?”  Each year we try to answer that question with our ‘HVW Winter Outlook’.  Our winter outlook is compiled by doing extensive research on a variety of atmospheric and oceanic factors that have been shown to influence our weather patterns.  These trends, teleconnections, and indicators can give us some insight into the possible weather patterns for the upcoming winter season.  We’ve compared our current and projected indicators for the 2020-2021 winter, to all the winters since 1960.  We’ve identified several years where the patterns and trends look very similar to our current year, and use those analogs for our winter outlook.  Using that data, lets take a look at what we expect for the coming winter season.

Winter Temperatures : Near Average (-1.0° to +1.0° vs. Average)

Winter temperatures are likely to be near normal across the Hudson Valley and northeast.  A moderate La Nina has set set up in the tropical Pacific Ocean.  This, combined with the overall sea surface temperatures (SSTs) suggest that we will see a winter with the potential to have several fluctuations between warm and cold patterns.  Historically, winters with a similar SST pattern have turned out colder than normal, so the potential for a colder than average winter is certainly there.  However, winters in recent years have been trending more mild, so we also factored that into our projection.  Based off our analogs, December is likely to be colder than average, followed by a milder January and first half of February, before colder than average temperatures return for the back end of February on into March.

Winter Snowfall : Near to Above Average (100% – 133% of Average) 42″ to 56″

Based on the analog years we have identified, we come up with the map shown above, which indicates above average snowfall for the Hudson Valley.  However, snowfall can be quite unpredictable.  A difference of 50 to 100 miles in the track of 2 or 3 coastal storms, can make or break your season in terms of snowfall.  So to account for the randomization of ‘luck’ with regard to winter storms and snowfall… we’re projecting anywhere from near average snowfall to above average snowfall… or 100% to 133% of average.  Since 1991, the average annual snowfall in Poughkeepsie is 42.8″, so that would me anywhere from 42″ to 56″ of snow in the Poughkeepsie area.  We’ll have to see if this year follows the expectations from the analogs more closely than recent years.

Methodology / Discussion

In previous years, we’ve spent considerable time digging into the methodology behind our winter outlook.  If you have viewed any of our previous outlooks, you have a good starting point for how we utilize Sea Surface Temperatures SSTs to infer their likely influence on weather patterns for the coming winter season.  You also know that we get quite technical in these discussions.  In the interest of saving time, this year we’re going to have a more direct approach.  But if you wish to review last year’s outlook for full context of how we develop the winter outlook, please check out the 19-20 HVW Winter Outlook.  If you’re a casual viewer, you may want to skip the in depth analysis… the choice is entirely yours.

Otherwise, lets start by looking at the actual SSTs.

This map represents the entire globe’s SSTs compared to average.  The blues and greens are waters that are much cooler than average… while orange and red are ocean waters much warmer than average.  We have placed an A, B and C over three (3) areas that we like to focus on.
A – Northeast Pacific , B – Tropical Pacific (la nina), C – NW Atlantic Ocean

These three areas all go a long way into helping determine the general jet stream pattern is likely to unfold.  The quick rationale is SSTs influence the air above.  The air above either rises or falls depending on whether it’s become more or less dense due to the influence of the ocean water below.  Those pressure changes then influence upper air patterns, and in effect determine our weather patterns through the winter.

So we’ve seen the actual SSTs… what does the CFSv2 computer model forecast look like for the winter…

Notice the similar pattern of SSTs between the actual SSTs and the computer model projected SSTs for the winter.  So if we want to try and predict what the coming winter might hold, we need to locate winters that had similar SST patterns to what we have, and what is projected.  Here is what we came up with.

Notice the pattern setup when we combine our analog years, looks very similar to the actual and projected SSTs for this winter.  The analog winters that combine to give us the SST patterns you see here, are:
1995-1996 , 1996-1997 , 2005-2006 , 2013-2014 , 2017-2018

These analog years then combine to give us the winter outlook temperature projection and snowfall projection seen at the beginning of the outlook.

Active Tropics : NW Atlantic Ocean

By now, everyone probably remembers that we’ve had a record active Atlantic Hurricane season.  The number of named storms has reached 30… with 13 hurricanes, and 6 of those becoming major hurricanes.  In trying to pinpoint trends to analyze for the winter outlook, we like to utilize the tropics as much as possible, and have found an interesting trend…

This graphic shows Atlantic Ocean ACE (accumulated cyclonic energy) for the hurricane season, for every year from 1851 to 2020.  ACE is effectively a measurement of the total energy output of a tropical system.  While the number of storms is important… a more detailed measurement is the combined strength of those storms.  An average hurricane season sees an ACE index of around 100.  This year, we’ve seen North Atlantic Ocean ACE of around 178, which is well above average.  However, if you look at the right side of the graphic, you’ll see the years with ACE over 150… the most active seasons on record.  One thing that jumps out to us, is that 3 of the 4 most active seasons since 1950 are:
– 2005, 1995, and 2017… 2020 ranks roughly 8th.

What’s the big deal?  of the 5 winters we identified using the Sea Surface Temperature (SST) method… 2005-2006, 1995-1996, and 2017-2018 are three of those winters!  So we have strong overlap between the hyper active tropical seasons and the SST pattern that is in place.  This simply helps strengthen our winter outlook confidence.  When we can look at 2 separate data sets, that compare 2020 to historical trends… and those two different data sets notice the same years being similar… we feel confident that we have a strong analog set for our winter outlook.

Month By Month

When we use the analog data from the 5 winters we’ve identified, here are the monthly temperatures compared to average that result.

December

January

February

March

The winters suggest that December starts off colder than recent years, with early chances at winter weather events.  January remains near average, with slight moderation, but still a similar trend to what we see in December.  February would appear likely to be the mildest month of winter, with some ridging in the eastern US trying to pull warmer air into the HV and northeast.  But March looks to try and close out the winter in cold and stormy fashion.  We’ll have to see if these historical monthly trends continue for the 2020-2021 winter.

Conclusion

Forecasting or projecting 3 to 5 months into the future is a serious challenge.  In fact, there are so many variables… the winter outlook becomes a forecast based on a forecast.  We assume certain conditions for SSTs, in anticipation that the maps shown earlier will be correct.  If certain areas of the ocean are notably warmer or colder than projected… the resulting forecast will be incorrect as well.  We’ve seen this several times in the past few years.  Where we put together the winter outlook, only to see areas of the ocean be much cooler than expected… and the result was actual temperatures that were much warmer than we anticipated.  So while the outlook is based in science, it’s still subject to considerable error.  We look forward to seeing if our ideas for this winter begin to unfold as expected, or if we start to see some extra curveballs thrown our way.  Either way, we hope you’ll be here with us… along for the ride… as we begin the month of December and unofficial winter season.  We thank you tremendously… for all your continued support of HVW.  You are what makes the HVW community so much fun to be a part of.  We hope everyone stays safe and healthy!  “Keep calm, and weather on…”

-Bill & Alex-

Tuesday Discussion : Seasonable Temps and Sun

by
Seasonably comfortable summer heat for the statistically hottest time of the year. A northwest wind will keep humidity in check, and keep temperatures from spiking much above the low 80s. A fine summer day for any activities you might have planned.
 
As we look to the remainder of the week, the wind shifts to the east and southeast, and that will allow more clouds and humidity into the picture for the 2nd half of the week. So while we don’t anticipate widespread rainfall… the chances for clouds and spotty rain showers. Scattered showers and thunderstorms are possible on Friday, in advance of what could be a pretty notable heat wave from Saturday through at least Monday of next week. A big ridge of high pressure will set up over the eastern US, and could allow for high heat and humidity, with temps in the mid 90s, and real feel temps near 100°. Something we’ll have to watch as we get closer for better details… but a hot weekend appears likely for the northeast and Hudson Valley.
 
Have a great Tuesday

November 4, 2019 : Long Range Outlook

by

The month of October split the country in half in terms of weather.  The northeast US was pretty close to historical averages, with a warm last week likely pushing temps just a bit above average.  While the western US was well below average, due to a persistent dip in the jet stream over the western US.

If you’ll recall during the month of October, there were record breaking snows in Montana and Idaho and the Dakotas, which makes these frigid temperatures much less surprising.  While here in the Hudson Valley, temperatures were pretty close to average.  We had some chilly starts, but our October was right about what we would expect.  The exception being the final days of October, and especially Halloween… where highs were in the low 70s (roughly 15 degrees above average).

So as we move into November, we’ve seen a drastic and sharp change to our weather pattern.  The 2 coldest mornings of the season so far occurred Friday and Saturday nights… sending temps near or below freezing across the entire region.  So with colder air injected into the valley… we decided to take a look at the upcoming pattern over the first half of November.

This is the current jet stream pattern over North America.  This type of jet stream in the middle of winter would be the harbinger of frigid temperatures… likely temperatures below zero.  This time of year, make no mistake, it’s got some cold air… plenty cold enough to support snow.  Locations centered under the deep blue and purple bulls eye, will be where the coldest air is located.  We’ve placed a purple and black star over the Hudson Valley for point of reference on this map.  As we start November, we’re seeing the cold air surge southward, from northern Canada, and it’s pushing into the northern half of the US.  Based on this trending push… here are the projected temperatures for the first week of November…

You can see we expect chilly conditions in the northeast and Hudson Valley… but nothing “brutally” cold.  Temperatures a couple degrees below average are expected through the first week.  With average high temperatures this time of year in the mid 50s… that means afternoon temps will struggle to climb out of the 40s in the afternoon.

But the truly cold air is building in Canada, and is poised to take aim on us.  Here is the projected jet stream as of Thursday…

An area of DEEP cold air is building in Canada, and pushing SE.  The huge ridge out west will allow for the Canadian cold to bowling ball its way into the eastern US.  At this same time, the SE ridge will be giving way to the bitter cold air.  A storm system is projected to form in the Midwest, and move eastward mid to late week.  Just how fast this cold air pushes south… will determine if the storm can make its way far enough north… to spread snow into the Northeast and Hudson Valley for Thursday or Friday.  We’ll talk more about that in a bit.

But in terms of temperatures… as this arctic air surges south… here are the projected temperatures for the 2nd week of November:

This is the 5 day period from Friday through Tuesday.  The average temperatures over that time… are projected to be 10° to 15° BELOW average!  That’s some wintry stuff.  Doing the math… those are afternoon high temperatures that do not get out of the 30s!  And overnight lows that could dip into the 10s and low 20s!  These conditions would be more typical of late December or early January!   So chop your firewood now…

As we reach mid November, you’ll notice that we don’t anticipate any major warm up coming our way… at least not based on the projected Jet Stream Pattern for mid month…

While the western ridge splits a bit, there is still tremendous blocking expected over Alaska.  That should continue to allow the cold air to push SE from northern Canada, and into the eastern US.  This type of pattern would continue to produce below average temperatures, and even potentials for winter storms to develop.  With regard to chances for snow… it’s certainly possible.  Timing is the key this time of year, and we need a storm to ride along the jet stream at just the right time, to combine with enough cold air.  The models have definitely shown that potential in the past few days…

European Model from Saturday Morning: Thursday Night – Friday

This setup appeared Saturday morning on the European model.  The cold arrives at just the right time, to allow the storm system to move into the Mid Atlantic and spread moisture into the cold air.  If this scenario were to unfold… 6 to 12 inches of snow could be the result in the Hudson Valley.  The timing looked to be Thursday afternoon into Friday morning.

However… like we mentioned, TIMING is the key.  Since that run of the computer model… data suggests the cold air will be more aggressive, and faster.  The end result is this…

European Model from Saturday Night: Thursday Night – Friday

The cold pushes south faster, and prevents the storm from moving as far north as the previous guidance suggested.  This would limit any major storm from developing, due to a flatter wave at the upper levels.  Light snow would be possible on the northern edge, but accumulations under 1 inch would be expected.  This scenario is a non-event for the Hudson Valley.  So in 1 run of the European model… we go from major winter storm, to non event.  Welcome to weather forecasting in the winter.

Beyond this Thursday night / Friday event… there will be more chances for snow.  The pattern just favors the potential to a large extent.  This pattern should hold with us for at least the first half of November… before things may begin to moderate.  But we’re going to kick this winter off early… whether it’s just cold temperatures… or if it includes snow.  We’ll have to wait and see.  One thing’s for sure… business is about to pick up at HVW.  We’ll be here to help guide you through all the twists and turns.  We appreciate your continued support, and can’t thank you enough!

HVW Preliminary Winter Outlook 2019-2020

by

Another summer is behind us, and the transition into autumn is upon us.  As the leaves begin to turn, many minds also begin to turn… to thoughts about what the coming winter season may hold.  Snow lovers begin to get excited, and winter haters begin to dread what lies ahead.  So as speculation begins to increase on the coming winter… that means it must be time for the 2019-2020 HVW Winter Outlook.  This year, we’re going to take a Preliminary look now, and follow up with a final update in late November.  This way we can discuss our ideas, and then update everyone on the trends, and if we need to make any changes to our preliminary ideas.  Will the coming winter be warm?… or bitter cold?  Will we see mountains of snow?… or will the winter be more wet than white?  These are some of the questions we’ll attempt to answer.  So enough small talk… let’s see what the coming winter might hold.

Winter Temperatures : Near Average to Slightly Below Average (-1.5° to +0.5°)

Temperatures this winter are likely to be near normal, to slightly below normal.  The range of -1.5° to +0.5° is due to the fact that we see some signals that could result in this winter being a chilly one.  The conditions heading into this winter are similar to the recent winters of 2014-2015, and last year 2018-2019.  Last winter was about 1.5° to 2.0° above average, while the 2014-2015 winter was about 3.5° below average.  So with both winters showing similarity to this coming winter, the tiny details will determine which previous winter our coming winter will look like.  The front half of the winter looks milder, with the back half of the winter looking more likely to be colder than average.

Snowfall : Near Average or Slightly Above Average (100% – 125% of Normal) 45″ to 55″

Snowfall for the coming season appears likely to be near average, to slightly above average.  The average snowfall in Poughkeepsie for the entire winter season is 43.7 inches.  We anticipate roughly 100% to 125% of normal snowfall totals in the Hudson Valley… that would translate into roughly 45″ to 55″ of snow in Poughkeepsie.  Seasonal snowfall is notoriously difficult to project, because along the east coast, nor’easters are wildcards that can drop 12″ or more in one shot.  The pattern that we expect to set up this winter should give us a few shots at nor’easter development.  If this season reaches its full potential, we wouldn’t be surprised to exceed the high end of our forecast range.  With that said, in general, we do believe that the storm tracks will be further east than last winter… leading to more snow events, and less wintry mix events than we saw last winter.  That should allow the seasonal snowfall totals to be higher than what we saw last winter, when Poughkeepsie saw 31.5″.

Methodology

(Disclaimer: This section may not be for everyone, it’s very heavy on the science behind the projection above.)
For anyone who may be new to HVW, we produce the Winter Outlook each year at this time.  Our goal is to provide a science based, long range projection of the coming winter season.  We combine our forecasting experience with computer data and historical trends, to produce the winter outlook each year.  Because there is computer data involved in creation of the outlook, it’s basically a ‘forecast based on a forecast’.  So if some of the assumptions we make to create the outlook turn out to be wrong… the winter outlook will be affected.  So lets take a look at how we arrived at the Winter Outlook.

Oceans absorb and store a tremendous amount of the earth’s heat, and that heat is transferred into the atmosphere.  Whether sea surface temperatures are influencing tropical storm development, helping to influence areas of high and low pressure, or whether an el nino or la nina are in place… oceans are one of the most (if not THE most) influential drivers of the weather pattern.  As such, how the ocean temperatures compare to average, are a fundamental key to deciphering what the coming winter may hold.  So, let’s see what our current SST (sea surface temperatures) look like…

The United States is in the upper middle of the global SST map.  The first thing that will jump out at you, is the bright red coloring in the Pacific Ocean.  That is indicative of well above average sea surface temperatures (SSTs).  We have highlighted 2 specific areas to focus on; (A) the Tropical Pacific, and (B) The Northeastern Pacific Ocean.

So now that we know the current conditions, let’s look at the projected conditions for the coming winter.  Below are the CFSv2 model, and the JAMSTEC model projections for Sea Surface Temperatures (SSTs) from December through February…

Now, comes the hard part.  We closely review the current conditions and the computer model projections for the winter.  Then we analyze ALL of the winters from 1950 to 2019.  Of those 70 winters, we identify those where conditions most closely resembled our current and projected conditions.  We combine them to create our “analog” winter… or the combination of years that most closely resemble the projected conditions for the coming winter.  Here is our analog for this winter…

 

When you compare our analog to the CFSv2 and JAMSTEC model projections, you’ll notice all 3 are very similar, with a few minor differences.

Area A, the Tropical Pacific is slightly different between the two computer models.  The JAMSTEC is slightly warmer than the CFSv2, but both are rather neutral in temperatures (near average).  So we don’t expect an el nino or la nina.  If the JAMSTEC is right, we could see a very weak el nino known as a “modoki el nino”.
Area B, the Northeast Pacific Ocean is very warm in both computer model projections, as well as our analog.  If you look at the Pacific Ocean as a whole (A & B together), all 3 maps display a very similar pattern of warmth, as well as the cooler than average water north of Hawaii.
While not labeled, if you look at the Atlantic Ocean (east of USA) you’ll notice our analog has warmer than average water off the east coast of the US, just like the models, and the cooler than average water south of Greenland on our analog, also consistent with the model projection.
We’ve combined multiple different winters, and weighted some more heavily than others, depending on how closely the conditions resemble what we anticipate this winter.

Additional Thoughts: Uncertainties & Wildcards

It’s probably the perfect time to do this section.  We just completed our recap of last winter and the comparison to the winter outlook.  So we’ve just analyzed last winter, and looked for weaknesses and errors in our methodology.  There are 2 factors we want to discuss, and the possible ramifications on the winter outlook.

First, there is the “cool” patch of water projected by the CFSv2 model off the western US coast…

We circled the area just to the east of the “B” on the map.  A small area of near normal temperatures… in a sea of well above average water.  It doesn’t look like much, but something similar to this was a major factor in the busting of last winter’s outlook.  You can read about all the details here: Winter Recap : 2018-2019 Winter Outlook Review.
Could it happen again? … Not likely.  The reasons are two fold.  First, this model is not supported by the JAMSTEC model, which has a solid band of warmth in the area in question.  Secondly, the area NW of Hawaii is cooler than average water, something last year did not have.  So we think this is not a big issue for uncertainty.  However, we still want to keep a close eye on this area as we move into the winter.  If this area is cooler than average… it could have major implications for the weather patterns that set up.

The 2nd “Wild Card” is much more interesting… let’s call it the “safety off” scenario.  When we began working on the Winter Outlook, and compiling all the winters that most closely represented this year, one winter stood out above all others.  The 2014-2015 winter.

 

On the left is the 14-15 winter… and on the right is this year’s JAMSTEC model for the winter.  The patterns are eerily similar.  For anyone who may not remember, December was very warm, and January was cooler… and then February was record breaking cold!  So we factored in the 2014 – 2015 heavily when creating this winter outlook.
So what’s the big deal?  … well, 2014 – 2015 was such a severe winter, to create our temperature projections for the winter, we had to tweak our settings.  The 2014-2015 winter was making our winter outlook too cold, and we needed to moderate the temperature outlook.  So we minimized the weight we gave to the 14-15 winter just a bit.

So when we discuss our concerns for a wildcard… the possibility that this winter is colder than what we are projecting, is on the table.  If we take the “safety off” and apply the proper weight to the 14-15 winter based on the similarities, this is what the winter temperatures look like…

This would be a much colder winter than what our actual outlook is suggesting.  You may ask, why don’t we use this as our outlook temperature projection?  The truth is, that was a record breaking winter.  And similarities aside, it’s not wise to use such an anomalous event as the basis for your winter outlook, or any forecast.  Many factors created the record cold February 2015, and to believe that will happen again, is irresponsible.  We only mention it here in this section, because we want to highlight that the temperature profile of the Pacific Ocean is very similar, and conductive to a potentially very cold winter.  We’ll have to see if the factors come together to give us the colder solution… but odds are much better that we’ll experience a more temperate winter.  That’s not to say it can’t be very cold at times, but lets wait and see.

So there you have it… the 2019-2020 Winter Outlook.  We hope you’ve enjoyed what we swore would not be as all encompassing as previous outlooks, but somehow became very, very thorough.  We greatly appreciate all your support, and can’t wait to help guide you through another winter in the Hudson Valley.  Keep Calm… and Weather On.  Thanks for reading.

2018-2019 HVW Winter Outlook

by

The days are getting colder, and the nights are featuring frost.  We have even just had our first snowstorm of the early season.  So with the 2018-2019 winter getting closer each day, the thought “what is this winter going to be like?” is growing more and more common across the region.  So that means it must be time for the HVW Winter Outlook.  This has become an annual event, where we combine science with speculation, and try to project the general weather patterns for the coming several months.  We focus on the temperatures and potential snowfall, based on what occurred when conditions were similar in the past.  Hopefully, you’re familiar with our method, but if you’re not… don’t worry, we’ll explain as we go.  So without further delay, lets get into the 2018-2019 HVW Winter Outlook:

Winter Temperatures : Below Average (-0.5° to -2.5° compared to Average)

Temperatures this winter are expected to be below average across the Hudson Valley.  We expect a persistent ridge of high pressure has set up across the western US, causing above average temperatures.  The offset to that, has been a persistent dip in the jet stream across the eastern US is likely… which leads to below average temperatures.  Our research suggests the winter could get off to a similar start to recent years, with a near normal December.  However our latest guidance is suggesting December could be colder than average and active.  Then, January is likely to be slightly colder than average, featuring some cold snaps, followed by a thaw or two.  February appears likely to be the coldest month of the winter, with the best chances for snow and sustained cold.  When all is said and done, temperatures are likely to be 0.5 to 2.5 degrees below average.

Winter Snowfall : Above Average (120% to 150% of average)

Snowfall this winter appears likely to be above average.  The 26 year average for snowfall in Poughkeepsie is 44.2″.  From first flake to last flake this winter, we expect to see snowfall totals roughly between 50″ and 65″.  The wide range in the forecast is a product of uncertain nor’easter storm tracks.  The reality of snowfall totals, is that luck is certainly a factor.  Patterns can be cold enough to support snow, but then the coastal storm hugs the coast a bit closer than anticipated.  The result can be a forecast of 6 to 12 inches of snow, becoming only 5 inches before mixing with rain.  Even with the variability of storm tracks, on average, we believe snowfall for the season will be above average.  The snowiest month is likely to be February, while December could get of to a slow start in terms of snowfall amounts.

Winter Outlook Methodology

This section may not be for everyone, as we tend to get a bit heavy on the science and details.  Many people just want the basic winter outlook, which is what you’ll find above.  However, if you’re looking for insight into how we arrive at our conclusion, you may find some interesting data below.  Each year, we use roughly the same technique in generating our winter outlook.  We focus on oceanic patters across the globe, with extra emphasis on the tropical Pacific Ocean.  Whether we are in an El Nino or La Nina has significant impact on the outlook.  We also use other data points and techniques… such as: early season Siberian snow cover; Atlantic Ocean tropical storm activity; solar cycles; volcanic activity; etc.

So lets begin with our current sea surface temperature pattern across the globe:

These are the global ocean temperatures compared to average, as of the end of October.  You’ll find the United States in the middle of the top right quadrant (to the right of the letter B).  So we’ve highlighted a couple areas which we will use for the rest of this section:

  • “A” indicates the Tropical Pacific Ocean, suggesting a weak el nino may be trying to set up… with warmer than average ocean temperatures.
  • “B” indicates the Northern Pacific Ocean, where waters are much warmer than average.

We have highlighted these areas, because the ocean also has influence over favoring higher or lower than normal pressures, depending on where the waters are warmer or colder.  Now lets look at the projected sea surface temperatures (SSTs) from the European and American (CFSv2) models for this winter (December, January, February):

You’ll notice that both computer model projections yield similar results.  If you look at location “A”, both models are projecting a weak el nino, with quite warm waters in the Tropical Pacific Ocean.  Location “B” suggests very warm waters close to Alaska and Canada, with cooler waters slightly further SW toward Hawaii.

So next, we went through the databases, and looked at every winter since 1950.  We isolated the winters that were most in line with the projected sea surface trends for the coming winter.  When we combined them together, we get this projected SST map:

If you compare our analog map of similar winters, you can’t help but notice the similarities to the computer model projections.  The Tropical Pacific Ocean “A”, displays a weak el nino, with wide sweeping warmer than average waters… all the way from the date line (180°W) to the South American coast.  In addition, the Northern Pacific Ocean “B”, is also very similar.  The waters off the coast of North America are quite warm, with slightly cooler waters expected pushing back toward Hawaii.  Here is the generalized 500mb jet stream pattern that unfolds when the oceanic pattern sets up as projected…

In similar years, a persistent upper level low pressure sets up north of Hawaii, which is reflexive of the cooler water temps in that region.  There is a responsive ridge of high pressure that develops over the NW US and Canada.  That ridge is responsible for above average temperatures consistent in winters like the one we are projecting.  Then downstream, over the eastern US, a persistent trough is likely.  That should translate into colder than normal temperatures, as well as a more active storm pattern than normal.  And Lastly, a strong area of high pressure develops over Greenland, thanks to a very strong negative North Atlantic Oscillation (-NAO).  This is a key feature with regard to the development, and track of nor’easters.  This pattern would be very favorable for an above average snow season.

Testing the Theory

Utilizing the ENSO and the Sea-Surface Temperature patterns that are in place, we develop the base forecast discussed above for the winter.   We have tested this methodology in previous winters, and it does give us a good starting point for analyzing the winter.  The projected winter pattern is quite similar to the current actual SST pattern… so then, let us test the method by looking at October, and see how the SST forecast analog for October compares to the actual October temperatures…

While it is not exact (these type of long range projections rarely are), the analog and the actual are quite similar.  When comparing the actual to the analog… the actual temperatures along the eastern US are generally warmer.  But it’s not the precision details we utilize when comparing legitimacy of the forecast method.  Keep in mind, we are using a generalized pattern over multiple months, to project patterns for the upcoming winter.  So rather than notice that the SE is warmer than modeled… lets focus on the general patterns in the analog and the actual.

  1. Both the analog and the actual are mild right along the west coast
  2. Both scenarios are below average in the center of the country, with the core cold in the upper mid-west
  3. Both scenarios are near to above average the further you go in the SE

Both the forecast October and the actual October were near or below average for most of the country.  The only warmth is found in the southeast, with the actual warmer in the SE than the forecast.  In addition, the warmth extends over a bit further NW, covering more area in the actual result.  A few of the details are different… but the overriding pattern is similar.   So it’s not a stretch to believe that the methodology used for the winter outlook holds merit.  However, should these factors change… and the pattern not unfold as expected… it would have major ramifications for the accuracy of the outlook.  Something to keep in mind, when taking the outlook into account.

In Conclusion…

We’ve already had our first snow of the season, quite a powerful punch for most of the region.  So the 18-19 Winter season isn’t looking so far away any longer.  The question is, what happens from here.  Does the winter moderate a bit, and give us a near normal… or even above average temperatures this winter?  Or does the winter play out like we expect… which means that our mid-November nor’easter was only a taste of what’s to come.  We’ll soon find out soon enough.  No matter what happens, we’ll be here with you the entire way.  We’ll forecast the coming winter, and focus on each and every storm… to give you the best information possible.  So you can make informed decisions, and stay one step ahead of the weather.  Thank you for reading our 18-19 HVW Winter Outlook… and thank you for supporting Hudson Valley Weather.