Denver, Colorado

Winter Outlook 2014-2015, Part 2: Forecast Models

Posted by @BrianBledsoe

In Part 1 of our winter outlook, we laid out the different analogs and surmised, in general, where we are likely heading for winter and spring. The tough part about using just analogs in this forecast is that there is no perfect analog for what is about to happen. Some are pretty close, but within those particular analogs something always comes up that makes it less than perfect. So, if we can use analogs to get an idea about the forecast, we should then supplement those analogs with computer model information to increase potential accuracy.

Many of you are thinking that this winter and spring may be similar to what has happened during October -- i.e., very warm and dry. I am going to show you why I haven’t jumped on that wagon, and why I think being patient with this forecast is what it is all about. One model has been consistent with its forecast. It likely has the best handle on what’s going on with the sea surface temperatures in the oceans, and that is as important as anything. It is the JAMSTEC Model…

JAMSTEC Model Sea Surface Temperature Anomaly Forecast (Red = Above Normal Blue = Below Normal)

December - February

DEC-FEB SST Forecast JAMSTECJapan Agency for Marine-Earth Science and Technology

Notice the warmer than normal water across most of the Pacific Ocean. Two areas of interest are near the west coast of North America, and along the ENSO region off the west coast of South America. While water temperatures in the Gulf of Alaska and off the west coast of the US have been cooling, the model still forecasts SST anomalies to be above normal in this area. However, this warm pocket is much cooler and a lot less broad than what has been observed during the past year. Warmer than normal water is also observed in the ENSO region off the west coast of South America. While we are technically not meeting El Niño conditions right now, the JAMSTEC says we will likely move into El Niño territory during the early part of the winter. The interesting thing about the JAMSTEC is that it keeps the warmest water in the central part of the ENSO region. This would still be classified as a traditional El Niño, though not for long if the warm anomaly continues to move westward. Let’s take a look…

March - May

MAM 2015 SST Jamstec
Japan Agency for Marine-Earth Science and Technology

Warmest water in the ENSO region continues to favor the central region, but is expanding to the western region. This would suggest that any El Niño that develops would develop in a traditional manner, then shift to what we’ve discussed in the past...Modoki El Niño.

June - August

JJA 2015 SST forecast
Japan Agency for Marine-Earth Science and Technology

By summertime, the warmest water still resides in the central and western ENSO regions. The warm anomaly has cooled somewhat since the spring, and this is typical. Since the PDO shifted to its cool phase, we have not had an El Niño maintain its strength or strengthen during the summer. I see no reason to suggest that the summer of 2015 will be any different. Bottom line, I think we are still looking at a weak El Niño event that starts out traditional and evolves into Modoki. The map below shows the difference between the two:

El Niño and El Niño Modoki EOF Maps from JAMSTEC

El Nino vs El Nino Modiki

A weak El Niño doesn’t have the same classic signals that a moderate or stronger El Niño portrays. Plus, this El Niño will likely end as a Modoki and that makes the forecast signal different too. I think you will see that in the JAMSTEC temperature and precipitation forecast.

JAMSTEC Model Temperature Forecast (Red = Above Normal Blue = Below Normal)

December - February

Temperature forecast winter 2014-2015
Japan Agency for Marine-Earth Science and Technology

March - May

MAM Temperature Forecast
Japan Agency for Marine-Earth Science and Technology

June - August

JJA temperature forecast
Japan Agency for Marine-Earth Science and Technology

According to the model forecast, Colorado would experience colder than normal temperatures for the winter and spring. The pattern that would put this in motion is a ridge (not as strong as last year) in the Gulf of Alaska, and a weak, but stronger than usual, subtropical jet stream riding along the southern tier of the country. We would have a steady supply of colder than normal air from our north, combining with decent moisture just to our south. That is a recipe for more frequent moisture producing storms...storms that could potentially be good snow-producers and keep the ground and surrounding air temperatures cold.

JAMSTEC Model Precipitation Forecast (Green = Above Normal Brown = Below Normal)

December - February

DJF precipitation forecast USA
Japan Agency for Marine-Earth Science and Technology

March - May

Japan Agency for Marine-Earth Science and Technology

June - August 

JJA PrecipJapan Agency for Marine-Earth Science and Technology

The JAMSTEC keeps Colorado in above normal moisture for the winter and keeps most of Colorado wet for the spring. This is especially true for the southern part of the state. The model then starts to dry us out a bit for the summer and actually has a negative moisture anomaly from southwest Colorado southward during monsoon season. If this were to pan out, most of Colorado would do very well in the snow department this winter and spring. Given the likely setup, it is feasible.

I should have included these in Part 1, but here are the analog year snow totals for both Denver and Colorado Springs.

Denver snowfall totals analog years

Colorado Springs
Colorado Springs snowfall totals analog years
The analogs generally give Denver “normal” snowfall while they give Colorado Springs above normal snowfall. Some nice snow totals for both cities during those analog years are seen, but also a dud or two...which many of you are worried about. ‘76-77 and ‘02-03 were both downers for snow lovers. Why do I say this about ‘02-03, despite 62” of snow for Denver? Because most of that winter was free of snow, until the monster mid March blizzard. So, don’t think for a minute that I am ignoring those two down years among our analogs...they bug me everytime I look at the numbers.

Here’s the bottom-line: this is a tough forecast and one that requires some patience. There are many conflicting signals, we’ve had a very warm and dry October, no analog matches our current conditions terribly well, and it is a long range forecast. However, I do feel good about Denver and the northern Front Range seeing AT LEAST NORMAL SNOWFALL. There is a decent chance it could be above normal, too, though with Colorado Springs and the southern Front Range more likely to see above normal snowfall for the winter. This is also true for the southern mountains and the Southeast Plains.

Over the coming week’s we’ll continue to hone in on which analogs appear to best fit the pattern going forward and weight accordingly as we head into the winter months. We’ll also be taking a look at what impact the
biggest sunspot in over two decades may have on our winter forecast, and long-ranging forecasting all together.

Despite a warm fall so far, I also feel we will get pretty cold. This is especially true for the late winter and spring. With an active storm track, I believe we will struggle to warm up during the spring. So, if you thought winter was going to be a no show this year, think again. As I said at the Weather5280 Meetup, it is likely going to be a late show and possibly a good one.

The State of the Atmosphere: Sunday, October 26, 2014

Posted by @brendansweather

MISSING: Winter.

Okay, not entirely. While Denver’s average first measurable snowfall came and went last week, there are still many snowy months ahead. We were spoiled with a cool(er) wet summer, so a streak of warm, calm weather only seems fair. Right? Now we’ll see if we can follow this with a more interesting winter than recent years.

We finished this past week with record warm temperatures. Denver recorded a high of 82 degrees at Denver International Airport Friday, breaking the old record of 80 degrees
set in 2011 for the date. The normal high low for the this time of year is 62 and 34 -- numbers we’ve been nowhere near for some time now.

While I know a lot of the snow lovers that follow the blog were as eager as some of us for an October snow, a snowy October does not always bode well for a snowy winter. On average, Denver receives just 4.2 inches of snow during the month of October, which equates to our
seventh snowiest month of the year. The next six months all average more snow than October does, with next month (November) taking the four-spot.

Of the five analog years outlined in our
Winter Outlook, Part 1, only one of them had a impressive October snowfall -- 2009, which has dropped lower in the list of analogs over recent months. Additionally, all but one of these years (1976) had average or above average snowfall for the season, though many of the years the snow came during the middle to second half of the winter (or in the case of 2002-2003 virtually all at once in March).

Much cooler start to the week
We’ll finally say goodbye to our well above average temperatures to start the coming week. As discussed in last week’s State of the Atmosphere, this cold front won’t be anything to write home about -- but we’ll certainly feel the punch after temperatures running so far above average for so long. It’s also possible that we see our first below freezing temperatures of the season by early Tuesday morning.

This system will carry some moisture with it, but not a ton. For the time being we’re going with scattered showers for the eastern Plains, and a chance for snow/snow showers at higher elevations for tonight and Monday. The National Weather Service has issued a
Winter Weather Advisory for the north-central mountains for 2 to 5 inches of snow possible above 9500 feet.

Here are the forecast 2 meter temperatures for yesterday afternoon (Ieft), and then Monday afternoon (right), much cooler across the west.

GFS Temperature Forecast Monday October 27, 2014

Model mess
By the middle of the week back comes the warmth as the real cold slips into the east (sound familiar?). To be perfectly honest, there’s not a lot in the medium range that looks all that convincingly cold, snowy, or for that matter, interesting. Models have been a pretty big mess over the last few weeks in the long-range pattern and with that it’s pretty hard to get behind any one solution that they spit out. My prediction is that we get a decent cold shot before too long, and that it could sneak up on us a bit as I’m not all the impressed with medium range ensembles.

The next best chance at snow could come during the first week of November -- and even that could be a stretch, though it does look generally colder. At one point models were trying to bring in a deeper trough for the Oct 30 - Nov 2 time period, and now this looks to be more like Nov 5 to Nov 9, really a tell-tale sign they don’t have a good handle on things.

An excellent example of this is the Canadian model. The top image is the 00z run from October 24 for forecast hour 348, or November 7. Below is the 00z Canadian ensemble run on October 25 for the same forecast hour. Notice the positioning of the lower height anomalies… a few days ago across southeast Colorado on November 7, now they’re still back off the west coast for the same period! It still does progress a trough across the region, but it’s got a much different look to it and is several days later than previously forecast.

CMC Ensemble example one

CMC ensemble example twoSource: WeatherBell Analytics

The GFS and ECMWF ensembles continue to vary on bringing multiple disturbances across Colorado next week, the first by next Sunday, the second by the following weekend. Again, hard to get behind one solution or another at this point, but next week certainly has the look of being (in general) more interesting. The EURO ensemble control is at this time the coldest of the three models really bringing a blast of cold air across the state by next week.

Bottom line
Bottom line is there’s not much to watch at this time, at least for us. Weather stories for the coming week will be a cooler start to the week across Colorado before seasonal to warmer than normal temperatures return for the middle of the week. The cold will slip into the eastern United States for the Halloween weekend, and the Pacific Northwest which has been really getting hit with rain and snow will continue to be impacted over the coming week and into next weekend as energy from what is now Hurricane Ana get wrapped into the flow.

Here’s the current GFS precip + MSLP look for 12z next Saturday (hey look, rain and snow for northern California!):

GFS Precip Forecast Northern California | Ana
And afternoon 2 meter temperature maximums for Sunday, November 1 (chilly in the east).

GFS Temps

I’ll resist the urge to share with you CFSv2 temperature outlooks and some of the more grim ensemble model runs for now as I really do think that while models may be right on keeping us pretty quiet over the next 5 days+, overall they’re not handling things all that great. It’s best to have patience and see how things start to shake out as we head into the final days of October and the start of November. Next week could very well be colder.

For what it's worth, I still believe we have a good chance for a snowy winter, even if it's painfully slow to arrive.

Winter Outlook 2014-2015, Part 1: Analogs

Posted by @BrianBledsoe

If you didn’t attend the Weather5280 Meetup last week, you missed a great hang, but fear not: a) we’re likely to have another one in the coming months; b) I’m going to share a lot of the same material that I talked about that evening, and how it relates to what will happen for our winter/spring 2014-15. So, here we go... 

Sea Surface Temperature Anomalies:

SST September 2014Source: UNISYS

The above image represents how the sea surface temperature anomalies looked on September 21st, just one month ago.

Here is how they look now:

SST October 2014Source: UNISYS

Notice how much the northern Pacific Ocean has cooled! It is completely different than it was a month ago. This has allowed a blocking ridge of high pressure to break down, which is allowing more storms to come into the West Coast. This is
potentially quite good news for California, as it could well allow for more storms to impact the state. Part of this is due to seasonal change, but the overall shift that is occurring in the Pacific is mostly responsible due to the current PDO phase and some sort of El Niño episode that is trying desperately to get underway. You hear me mention the PDO a lot, because it plays such an important part in our weather. While we are predominantly in the cool/negative phase of the PDO, we have taken a short detour to the warm/positive phase in the PDO since January. We still have a +PDO, and I believe it has been instrumental in helping erase drought conditions for the whole state. In fact, check out these moisture totals since April across Colorado:

Total Precipitation Colorado April-October 2014Source: WeatherBell Analytics

Some areas along and north of I-70 have received between 20 and 30 inches of moisture! Conversely, areas south of I-70 haven’t received as much. However, the moisture that they have received has gone a long way in helping cure the awful drought. The most recent drought monitor suggests that too.

Here’s how things looked just five months ago on May 20, 2014.

Drought Monitor Colorado May 2014Source: Drought Monitor

Much of the drought that you see above was the result of two years of below normal moisture. In fact, that two year period was one of the driest on record for many areas in far Southeast Colorado. I firmly believe that it was a result of the PDO being in a long term cool/negative phase. That phase is more prone to La Nina episodes which are notoriously windy and dry for many areas east of the mountains.

Here’s the most recent drought monitor:

Drought Monitor Colorado October 2014Source: Drought Monitor

Bottom-line, the short detour in the PDO to a warm/positive phase has been very instrumental in helping us in the moisture department.

Here is a look at sea surface temperature (SST) anomalies associated with the two phases of the PDO:

PDO phases
If you look at the SST anomaly maps above, you will see that we are dealing with a warm/positive PDO. However, I believe that this phase of the PDO will be pretty short lived. Thus, we had better continue to capitalize on it in terms of moisture, before we go back to a pattern that will be more conducive to drought conditions once again. Let’s see if that is possible…

Analog Years
You have heard me mention analog years before. These are simply other years in the past that have similar global weather patterns to those that are occurring currently. Contrary to the belief of the Climate Prediction Center, they are a very important part of long range and seasonal weather prediction. In fact, through about mid-to-late summer this year, 2009-10 was acting like it would be the perfect analog year to project the rest of 2014 and into 2015. While it is still in the running, there are other years that look even better. We have to break them down in terms of what the SSts looked like during each particular year, then we can see how the temperature and moisture panned out.


Sea Surface Temperature Anomalies October - May


This has a similar look to what’s happening now. We see colder than normal water extending eastward from Japan, warm water in the ENSO (El Niño Southern Oscillation) region extending westward from South America, and some residual warm water along the western coast of North America.

Temperature Anomaly October - May

NOAA NCDC Temps 1957

Yellow to red means warmer than normal, while the greens and blues mean colder than normal. In general Colorado was normal to slightly colder than normal, with the coldest weather being just east and south of us.

Precipitation Anomaly October - May

1957 Precipitation Anomalies

Yellows to reds mean drier than normal, while greens and blues mean wetter than normal. For this period, Colorado was slightly wetter than normal, as was most of the southern tier of the country.


Sea Surface Temperature Anomalies October - May

1976-77 SST

Same story here, with a very cold pool of water in the northwest Pacific Ocean and relatively warmer than normal water near the West Coast and in the ENSO region.

Temperature Anomaly October - May

1076-77 Temperature Anomalies

We see colder than normal for the southern and eastern tier of the country, with warmer than normal temperatures in the far Northern Rockies/Plains. This winter was legendary in the eastern part of the country due to how cold it was.

Precipitation Anomaly October - May

1976-77 Precipitation Anomalies

Above normal moisture was hard to find here. The west was dry, as was the Mississippi and Ohio Valleys. At least the eastern part of Colorado had normal moisture during this time, with above normal moisture in parts of the Nebraska and Texas.


Sea Surface Temperature Anomalies

1986 SST

This is another year that looks very similar to what we are currently experiencing. However, the ENSO region was pretty warm in late 86 and 87, and I’m not sure we will see that strong of an El Niño. More on that later…

Temperature Anomaly October - May

1986-87 Temperature Anomalies

We see this analog year was quite warm in the northern tier of the country, with mostly normal temperatures for the rest of the country. The one exception was areas of west Texas and New Mexico.

Precipitation Anomaly October - May

1986-87 Precipitation Anomalies

This is where the signal really shows up...Dry California, Great Lakes and Ohio Valley. Wet Colorado and most of the Plains states, eastward into the Southeast.


Sea Surface Temperature Anomalies:

SST 2002 2003


While a lot of the Pacific looks very similar to where we are and likely where we are heading, that warm pool just off the coast of Western Canada may be too warm. Otherwise, I really like this particular SST analog year.

Temperature Anomaly October - May

2002-03 Temperature Anomalies
Precipitation Anomaly October - May

Precipitation Anomalies 2002 - 2003
This year has a less clear cut signal unless you live in the far south and southeast. That slightly wetter than normal signal in the northern part of Colorado was largely due to the major blizzard in mid March 2003. Otherwise, the winter and early spring was quite dry. The lack of moisture for the whole state was largely made up in the spring.


Sea Surface Temperature Anomalies:

2009-10 SST


This analog used to be our favorite, but is now running behind a bit. The overall SST structure isn’t as good as the other analogs, but there are some signals that keep it intriguing. The less than warm North Pacific and a weak El Niño make it one that we have to give some weight…

Temperature Anomaly October - May

2009-10 temperature anomalies

Almost the entire country was generally cooler than normal during this time. The obvious exception was the Great Lakes Region and The Northeast. Otherwise, this was one of the colder fall/winter periods in recent memory for folks in the Southern Plains and Southeast.

Precipitation Anomaly October - May

2009-10 precipitation anomalies

The wettest areas were the Western High Plains just east and southeast of Colorado, parts of the Mississippi Valley, and of course the Southeast. Colorado had decent moisture, but nothing excessive after a wet and cold fall.

Here is the composite analogs factoring in the years from above:


Composite temperature anomalies


Composite precipitation anomalies
Overall, I think it is safe to say that our analogs produce cooler than normal weather in the south, warmer than normal weather in the north, and generally “normal” temperatures for Colorado. They also produce slightly wetter than normal weather in the South/Southeast, drier than normal weather in the Pacific Northwest and Great Lakes Region. Again, Colorado shows up as having “normal” precipitation. Do I think it will be like this? Not exactly. Whatever type of El Niño episode that develops will likely be weak. That means that some of the traditional signals will be a bit weaker, and possibly displaced. Coming up in a few days, I will post part two of this outlook, which focuses on model data and why I think that ONE particular model has a good handle on the overall forecast.

The State of the Atmosphere: Monday, October 20th, 2014

Posted by @coloradowx

Recap of month so far
If it feels like we have had a lot of warm days so far in October, you’re not mistaken: Denver is now a not-insignificant +2.2℉ above normal for temperatures month-to-date. In fact, the last 7 consecutive days (including today) have featured well above-normal temperatures, and over this one-week period, we’re actually averaging a sizable 7.7℉ above normal.

What has caused this prolonged stretch of above normal temperatures? Quite simply, we’re in an amplified -- or “blocky” -- atmospheric pattern aloft, which has generally featured a trough to our west and a trough to our east, with a ridge overtop of us, in between. The downstream trough has prevented much movement in the jet stream, so we’ve been stuck with a plethora of ridging dominating our weather. These blocky patterns can be notoriously difficult to decay, and it looks like we’re in for at least another week of continued above normal temperatures and drier-than-normal conditions until then. (We’re running somewhat behind on precipitation for the month, as Denver International Airport (DIA) has reported 0.50” of rain so far this month, or ~49% of our normal total monthly amount, with approximately 65% of the month having already passed.)

The week ahead
There are few forcing mechanisms to push our current blocky jet stream configuration out of its current orientation. That said, we will have a very short-lived respite from above normal temperatures during the middle of this week as a weak trough of low pressure and associated cold front pushes through the area Tuesday night. That likely means our well-above normal high temperatures today and Tuesday (in the mid-to-upper 70s, which is >10℉ above normal) will be pushed to closer near 60℉ on Wednesday. Also, scattered showers may accompany this system late Tuesday into early Wednesday, but most models are quite bearish regarding precipitation potential in #Denver --  probably <0.10” of rain can be expected.

Behind this weak, and short-lived cold front, a ridge of high pressure will build into the area anew by Thursday into Friday, bringing temperatures right back up to well above-normal levels by late week into the weekend. In fact, I wouldn’t be shocked if we flirt with the 80 degree mark again by this coming weekend -- which would imply close to record-high temperatures once again.

Note forecast temperature anomalies over the next 10 days from the CFSv2 shows above-normal temperatures continuing.

CFSv2 Forecast Temperatures USASource: WeatherBell Analytics

What changes going forward?
We all know above-normal temperatures can’t last forever, and we are now beginning to see some indications that this warm, dry pattern over the western US may begin to break down towards the end of the month. This is a bit more technical, but we’re currently under a strong +PNA pattern, which favors a ridge over the western US; you can see, however, that teleconnection forecasts weaken the +PNA considerably over the next few days, and even suggest it may go negative by next week, which could support a turn to colder temperatures.

Forecast teleconnections October 2014Source: NOAA

As our +PNA begins to collapse, we can expect an upper-level trough developing off the coast of the western US. Indeed, all medium-range models do show an offshore trough by next weekend. Most models believe this trough will come ashore by around Monday of next week, but there are some fairly large differences in regards to timing, strength, and location of this trough. Those details will need to be sorted out before we can get a clearer image of exactly what effects this trough may have on weather around our neck of the woods. Here is a look at what the Canadian GEM control run is showing valid 6pm Sunday, October 26.

GEM forecast 500 hpa USASource: WeatherBell Analytics

As such, the period next Monday through Wednesday may offer unsettled weather, with the
possibility of cooler temperatures, more clouds, and the chance for some precipitation, particularly in the high country. It is too soon to know with any more confidence exactly how this trough will impact our weather, but we’ll probably have a much better handle by mid-week.

The end game?
I don’t think, however, that it is this trough the puts a decisive end to our warmer-than-normal pattern. Indeed, most models start to warm us up again by the middle of next week, as high pressure once again builds right back into the region. Our eyes will then be on the next upstream trough which is likely to pay us a visit during the first few days of November. I believe that it is this system which offers better chances for finally squelching our warmer-than-normal pattern, as well as bringing notably cooler temperatures and the chance for snow -- perhaps even on the Plains. In short, while it’s a ways out, I believe that the period November 1-3 needs to be monitored for possible impacts from a (deep?) trough aloft, which could finally bring central Colorado a taste of wintry weather!

Here is the latest Canadian ensemble control run valid 6pm on Saturday, November 1.

GEM 312 hour forecast CONUSSource: WeatherBell Analytics

...we will surely be talking about this, probably higher-impact, trough in next week’s State of the Atmosphere! Until then, enjoy our continued warmer-than-normal temperatures, and the probability of a brief respite in the warmth early next week, before one last (perhaps brief, only a few days) stretch of mild temperatures before we head into November. Note that the CFSv2 shows cooler-than-normal temperatures attempting to establish themselves over the western US during days 10-15 (which would be the first week of November.)

cfsv2 forecast temperature anomalies