About a month ago I talked a lot about the course we are likely on as we head into fall. I addressed the potential for and El Niño and what type, and reiterated that our top analog for the next several months is late 2009 and early 2010. What happened in late 2009 and 2010? See for yourself…
The above map shows the temperature anomalies (or departures from normal) from September 2009 through March 2010. All of that blue and purple represent colder than normal temperatures. Conversely, all of the yellow, orange, and red represent warmer than normal temperatures. Basically, the southeastern two-thirds of the country was quite cold during that time, with Colorado being on the northwest fringe of the coldest air. What about precipitation?
The map above shows the precipitation anomaly for the same period. Blue represents wetter than normal conditions, while yellow, orange, red represent drier than normal conditions. During the winter of 2009-10, Denver received 60.6” of snow (~105% of normal). Half of that came in October (17.2 inches) and March (12.8 inches).
Given the pattern that we are in and how things played out in late 2009, my gut says that we will hit fall pretty early with the potential for some very active weather in October. Now, we base our forecasts more than just on gut feelings, so let’s see what one of our better computer models thinks of this theory.
The JAMSTEC Model is based out of Japan and seems to have the overall pattern nailed, ocean temperatures included, which is a major key to this puzzle.
JAMSTEC Temperature Forecast:
September - November:
The blue shading represents cooler than normal temperatures and the red shading represents warmer than normal temperatures. If you look at Colorado, you see that the state is shaded in light blue. This means the model is calling for cooler than normal conditions for the September through November period.
December - February:
Get a load of all that blue across most of the United States! The model is going crazy with colder than normal weather, especially for the central and southern part of the country. This includes Colorado, which is shaded in more than just light blue. The darker the blue, the colder the forecast. This would suggest a higher than normal probability of below normal temperatures during the heart of the winter months for Colorado.
March - May:
The spring of 2015 is also showing up as being colder than normal for the same locations that the model believes will have just endured a very cold December through February. To me this suggests three things: 1) a steady tap of cold air from Canada, 2) an active storm track bringing several rounds of snow, and 3) the snowpack on the ground being VERY reluctant to melt. All three of those things in combination make a perfect recipe for a colder than normal winter. Speaking of snowfall, lets look at that (by way of precipitation anomalies) a bit more closely.
JAMSTEC Precipitation Forecast:
September - November:
On this map, the green represents wetter than normal conditions, and the brown represents drier than normal conditions. As you can see, the model has a lot of the United States wetter than normal for the fall. This includes Colorado. Plus, do you notice the slightly darker shade of green from just off the coast of Mexico into Arizona and Colorado? That enhancement means that those locations will likely be even wetter than the surrounding areas. I have been saying for a while now that the active Eastern Pacific hurricane season MAY supply us with a tropical connection of moisture that may meet up with a storm coming in from the west or northwest. That “phasing” -- in effect, when two jet streams link up -- has the potential to produce a major late summer flooding event or an early fall big snow event...or both. The pattern suggests that it is possible, and certainly something to watch.
December - February:
The heart of the winter shows up with generally normal moisture for Colorado, with a slightly wetter than normal signal for far Southern Colorado. I think most of us would take a winter with normal snowfall, as they have been hard to come by of late. Either the mountains get hammered and the lower elevations don’t, or vice versa. Those winters when both the plains and mountains see good snowfall have been rare to say the least. However, the signal the model is set on is dry from central California into the Pacific Northwest (bad) and wet from Eastern New Mexico into the Mid-Atlantic. Look familiar? Pretty identical to the second map I showed you in this post.
March - May:
The spring looks very similar to what the model is forecasting for the fall. It shows Colorado as being pretty wet, along with all of the Western High Plains. Getting good and consistent spring moisture has also been a rarity for Eastern Colorado lately. However, we had excellent spring moisture in 2010, and Denver even measured 1.3 inches of snow in May. Not unheard of, but is a testament to how long the cooler than normal weather hung around that year.
I always try to give a confidence reading with the forecast, and I must say that my confidence in this forecast is higher than normal. I’ve been talking about it for months, I haven’t seen any of the major drivers change much, and something else...can you feel a difference in the weather already? The nights are getting more crisp, some of the Aspen trees are already starting to show color, and we will see additional cold fronts move through during the next 10 days. Fall is coming, and it is likely coming early…
As we mentioned in this week’s State of the Atmosphere, it’ll be an up and down week for storm chances along the Front Range and eastern Colorado.
We saw several strong storms develop Monday across the eastern plains, with only a few scattered brief storms during the early afternoon hours in the Denver area. Today those storm chances go up a bit, with 30-40% chances for eastern Colorado.
While the best chances for strongest storms will likely be west and again east of the Denver area, our chances will be greater as well. Precipitable Water values will be higher than what we saw on Monday -- up from ~0.6 inches Monday afternoon to 0.8 - 1.0 inches today (models have been slightly overdone for Denver). Greatest PW according to the 18z NAM will extend from Denver northeast into southwestern Nebraska.
Source: WeatherBell Analytics
CAPE will also be greater across northeast Colorado, with pinpoint for greatest values across the northeast corner of the state where values could climb to 1500-2500. We’ll need to watch for capping across the eastern plains, but should see the cap break later in the afternoon as outflow boundaries push east.
Source: WeatherBell Analytics
As for precipitation totals… Both the GFS and EURO are pretty spotty with QPF through Tuesday night across northeast Colorado, generally ranging from 0.5 inches across the Front Range foothills, to a few tenths out east. We expect enough ingredients to be in place to allow for better storm coverage -- and in combination with the time of year, could see spotty cells put down heavy rainfall at times. As is always the case, impossible to pin down exactly how much any single location will see with these widely scattered storms. Overall severe threat should be low today, but can't rule out a few storms becoming severe across eastern Colorado.
As for timing, Denver’s best shot at getting a storm will be during the early to mid afternoon hours Tuesday, decreasing into the evening. A few storms may linger in the foothills into Tuesday evening, as we also see the best chances for strongest storms developing across the northeast corner of the state by late in the day.
Storm chances decrease a bit for Wednesday, but enough moisture should be in place to keep a chance for scattered storms going. In fact, we have a chance for at least isolated storms in the forecast through next weekend, with slight better chances returning for Thursday and Friday.
I recently went out with a couple of Forest Service firefighters to learn how they calculate the fire threat across Colorado. My interest in the Forest Service’s efforts to gauge wildfire threat levels started this winter when I was at a SNOTEL site measuring the snowpack atop Berthoud Pass. My co-worker and fellow meteorologist Cory Reppenhagen was with me when we saw the far greater than average snowpack and made the assumption that the fire threat will be much lower as a result… which was an erroneous assumption to make. As Cory and I left the SNOTEL site I thought, “I wonder how they know how much water actually gets into the trees. Lots of snow doesn’t mean that the water is actually getting into the trees, does it?”
As it turns out, the season snowpack can help keep the wildfire threat low, but only so long as there is snowpack. As soon as the snowpack melts, the trees rapidly dry out. Coming out of dormancy and going through the summer, all fire fuels (like trees, grass, and ground cover) need a steady supply of moisture to maintain that low fire threat. I should point out, a good snowpack can lead to quick growth of the entire vegetation, not just the trees, so consistent rain is needed to keep that vegetation from turning into fuel.
The Forest Service measures moisture content in these fuel sources, and does so regularly. Cory and I went along with a crew in the Pike National Forest in early August to see what it was all about.
We hit their Trout Creek fuel site, one of many spread throughout the forest. Samples were taken of the ground cover (litter), the ground layer just beneath a tree canopy (duff), living pine needles, and 3 various sizes of a tree (large twigs, medium to large branches/trunk). The collected samples are then “baked”.
This process meant adding all the collected fuel samples to a convection oven and baking them at 170 degrees for about 24 hours. The weight of the fuel sample is then compared to its weight before going into the oven. The difference in weight is the moisture (water) loss during the dehydration, which indicates the fuel’s moisture level.
These samples are collected at dozens of sites in Colorado, and hundreds across the Western US every two weeks. The compilation of the sites produces quick outlook images such as the one below, showing the moisture levels in the small to medium size branches (the 100-hour fuel as the Forest Service calls them). This image is from early August when Colorado’s steady supply of rainfall had high moisture content in this fuel source, and all other fuels measured quite wet as well.
I asked the Forest Service how long a period of dry weather it would take to see the fuels trend drier, increasing our fire threat. Their response: only a week!
I had been waiting for a good period of drier days to see the fuel moistures change. This past week, and especially weekend, was just what I needed. Statewide temperatures were near to above average, and rainfall was near to below average. See the comparison in the fuel moistures to two weeks ago.
100-Hour Fuels (the small to medium branch sizes):
Notice the drying that occured in both fuels, in a time period of two weeks. It makes sense that the smaller fuel dried in a more pronounced way, but even the large fuels showed drying across Colorado.
It is the combination of these fuels that make up our fire threat. These fuel samples help determine staffing for the Forest Service as well. On the day of our visit, several firefighters from that station were already working the fires in the Pacific Northwest where the threat was clearly greater. Yes, the fire threat in Colorado remains low, but that can change more quickly than I had previously thought.
As my colleagues and fellow Weather5280 bloggers have pointed out, the moisture and temperature outlook will change very little over the next few weeks. However, if continued drying occurs, these fuels will be more prime to ignite. Most of Colorado’s fires are started by lightning, though careless campfires are also a threat.
For the next week, the most likely region for an elevated fire threat (including factors of dry thunderstorms, wind, and heat) will be in Southern Wyoming to Northwestern Colorado.
My thanks to Lawrence Lujan, Jay Karle, and Samuel Urffer with the Forest Service.
Looking back at the first half of August, 2014
Now that more than half of the month of August has passed, it is prudent to take a look back on the conditions we experienced during the first half of the month in order to see how the month is shaping up overall and to inform our forecast for the remainder of the month.
On August 2, our average high started falling once again (from 90 to 89), and on August 9 our average overnight low also started falling once again (from 60 to 59). Our average high/low as of Monday is now 87/58, and that will fall rather quickly to 84/54 by the close of the month.
So far this month, Denver International Airport (DIA) has experienced slightly below normal temperatures (-1.4 ) month-to-date, and has recorded ~0.80” of rain versus full-month average of ~1.7”. (As is typical of hit-or-miss summer thunderstorms, however, some spots across central Colorado have picked up several inches or more of precipitation.)
Sunday marked just the 3rd day this month with 90 -temperatures, and the 26th so far this season. The average number of >90 days Denver sees per year is 40, so it appears likely that we will fall well below this number by the end of the warm season -- perhaps about 30 days total (or 25% fewer >90 days than is typical.)
CPC’s analysis of monthly temperatures shows the much of the country has seen slightly below normal temperatures month-to-date (including Colorado), with California and the Pacific Northwest having seen well above-normal temperatures.
It would be fair to say that up through press time, most of Denver metro has seen slightly below normal temperatures with near to somewhat above-normal precipitation month-to-date. That said, we've seen several days in a row (including Sunday) that have featured slightly above normal temperatures, and we’ll look to continue that trend over the next 48 hours or so.
GFS ensembles show that we’ll experience a modest step down in temperatures through the remainder of the month: above-normal over the next couple days; near-normal for several days thereafter; finally, the probability of somewhat below-normal temperatures to finish up the last week of the month.
To add confidence to our forecast, let’s also take a look at what the CFSv2 has for the next 10 days in terms of both temperatures and precipitation. The model shows metro Denver just on the cusp of slightly below-normal temperatures. The southern extent of a push of cooler-than-normal temperatures during the last week of the month is somewhat unclear at this point, so I think the remainder of the month may either end up near-normal or slightly below-normal for temperatures. When we put that all together, I believe the August, as a whole, will wind up just slightly below normal for temperatures -- perhaps no more than -1 for the month.
As for precipitation, the CFSv2 suggests that Denver will be very close to near-normal values for the remainder of the month. There are some positive precipitation anomalies (though not a terribly strong signal). I think we’ll end up near- or just slightly below-normal for precipitation through the rest of the month. When we put that all together, I believe August, as a whole, will end up very close to ‘normal’ for monthly precipitation: likely slightly below in some areas and slightly above in others.
In short, we’ll experience a blend of above-, near-, and below-normal temperatures (likely in that order as the rest of the month proceeds), with a fairly seasonally-normal precipitation pattern. The week ahead should feature daily afternoon storm chances (some greater than others), with an eye on how models want to handle the potential for a cooler and wetter few days by next weekend.
Quick glance at September, October and November
Looking further ahead, I will leave you with a tease of what may be in store for us for September → November (SON) period according to a constructed SST anomaly model run by CPC. We see a zone of below-normal temperatures over the central part of the country, but the western extent is somewhat unclear. Other models show this zone of cool a bit further west to include the eastern half of Colorado.
And here’s what our top analog year -- 2009 -- looked like during the September → November (SON) period. You can see a small area of below normal temperatures over the central and southern Plains, including eastern Colorado. The Weather5280 team will have more in-depth look at how the fall months may shape up for us over the next couple of weeks.