A Winter Storm in Late Summer - My Earliest Ever Thundersnow and
My First Lightning Trigger Success with Winter Precipitation
Southwest Colorado, September 8-10, 2020

by John Farley

During the period of September 8-10, Colorado and other western states were impacted by an unusually early winter storm and cold spell that produced record early snowfall and record lows in a number of areas. This event came on the heels of an unusual late-summer hot spell that produced record highs for September and records for so late in the season in many locations. For example, in Denver on September 5, a record high of 101 was recorded - an all-time record for the month of September and the latest in the season a temperature that high had ever been recorded. By September 8, measurable snow (an inch at DIA but considerably more in other parts of the area) fell, and a high temperature of 43 and a low of 31 were recorded. That 31 degree low was the earliest-ever below-freezing temperature on record in Denver. So from the latest 100+ temperature on record to the earliest below-freezing temperature on record in three days!

As the strong cold front approached, there was initially some disagreement among forecast models on how the big change in the weather would unfold. There was no doubt that there would be a quick swing from well-above normal temperatures to well-below normal temperatures in a short time, but there was considerable uncertainty regarding the amount of precipitation that would occur - a crucial point since much of Colorado was in severe to extreme drought, and wildfires were raging in several areas along and north of I-70. Some models suggested that as the front surged south, a cut-off low would develop somewhere near the Four Corners and linger from Tuesday, September 8 through Thursday, September 10. Other models suggested either that no such feature would develop, or that if one did, it would move off to the east or northest rather quickly. The former scenario would mean widespread rain and snow across Colorado, while the latter would mean relatively little precipitatoin. And some models wavered back and forth between the two possibilities.

By a day or two before the arrival of the front, however, the models began to come into agreement on the closed-low solution. So by Tuesday morning, I was optimistic that decent amounts of precipitation would occur, and given forecasts and the early-season nature of this winter-type storm, I was also pretty bullish that I would have an opportunity for a thundersnow chase, and perhaps more than one over the course of the potential 3-day storm. Below are summaries of my observations of this weather system for each of the three days it impacted the Pagosa Springs area and the southern San Juan Mountains of Colorado.

Tuesday, September 8

By mid-to-late morning Tuesday, a couple lines of thunderstorms had developed in southwest Colorado, ahead of the main area of precipitation occurring to the north over west-central Colorado and southeast Utah near the main upper low. Both lines became quite electrically-active, but the leading line appeared stronger on radar. It was not initially a squall line, but a line of discrete storms extending roughly from Durango south to the New Mexico state line. And the radar began to show mesocyclones and hail markers with some of the storms. Intense bursts of small hail occurred near Durango, and a special weather statement was issued for southeast La Plata County and southwest Archuleta County mentioning the potential for thunderstorm wind gusts to 55 mph and small hail. At this point, I was seeing potential for observing both severe weather around the Pagosa Springs area as the storms moved east and then thundersnow in the mountains once the storms got there. One challenge of observing storms on this day would be the high amount of smoke in the air from wildfires in California, Utah, and northern Colorado. Indeed, an air quality alert was in effect, and it was clear just from looking out the window that visibility would be significantly reduced. In fact, the smoke was so dense that some of the mountains that you normally can see from Pagosa Springs were invisible.

I decided to initially position myself just east of Pagosa Springs after briefly looking at the sky from the west edge of town. Already as I was leaving the wind was very strong and pine needles and small twigs were flying all over the place in my neighborhood. Visually, it looked like most of the activity was to the north and northwest, but radar showed the storms surging into western Archuleta County, and the aforementioned special weather statement mentioned eastward movement at 50 mph. I did not want to get overrun by the storms in town and then have to try to play catch-up, nor did I want to be under the trees in much of town with high winds blowing. However, by the time I stopped to look at the sky and check radar when I got through town, most of the storms south of U.S. 160 had disappeared from the radar screen. So I figured to get into position to photograph lightning and storm structure, I would need to get farther north than the east edge of town, so decided to proceed onward to the scenic overlook on U.S. 160 between Treasure Falls and Wolf Creek Pass - a place I had intended to eventually go anyway; now I just needed to go there sooner. As it turns out what apparently happened at this time was that the storms in the southern part of the line gusted out and collapsed, with their precipitation quickly ending, but with a surge of strong outflow wind through and east of Pagosa Springs. Pagosa got no rain from this line of storms, but a 46 mph gust was measured at the airport. And the wind was almost certainly stronger at the northeast edge of town and for a few miles northeast of there, as I later when returning home I would observe wind damage - a large tree that went down on U.S. 160 a couple miles northeast of town, which by the time I got back a little before 3 p.m., highway crews had just finished cleaning up, and also a tent that was brought down by the wind at the Malt Shoppe in the River Center at the east edge of town.

By around 1:00 p.m., I was at the aforementioned scenic overlook, but alas it was closed for construction so I stopped to watch the storms in a parking area along the road just above the scenic overlook. This picture will give you a sense of how the smoke (and perhaps by now, also some dust) in the air was imparing the view - both of the scenery and of the sky. It was quite murky:

I watched here for a while, and soom it became evident that a new storm was developing to my south. At this point, I decided to move on up to Wolf Creek to try to intercept thundersnow, which I thought might occur there as the storm developing to my south moved up over the mountains. I decided to position myself in the Wolf Creek Ski Area's lower parking lot, to try to get away from the traffic noise on the highway. Soon after I got there, a mix of rain and graupel (snow pellets) began, light and intermittent at first, but soon heavier and more predominantly graupel/snow pellets. Enough to get my video camera lens wet, and wouldn't you know, just as I was wiping the lens off, I saw a bolt of cloud-to-ground lightning not far away to my east or southeast. I missed that bolt, but was able to get the camera back in position in time to get the thunder, as the snow pellets continued to shower down. My earliest-ever thundersnow, September 8! Here is the video:

In the video, you also hear a second, much less loud rumble of thunder from lightning farther away, and see a picture that apppears below, which I will say more about shortly.

About five minutes after I got this video, the precipitation stopped at my location, as the now banded shorms shifted a little to my east. The storms were moving northeast, but apparently a little more east than north as the band had shifted to just east or southeast of the ski area. At this point, I decided to switch to my still camera and lightning trigger. Although I have gotten bolts from thundersnow storms on video in the past, I had never gotten a picture of lightning with a lightning trigger from a storm producing winter precipitation. In fact, I do not think I had even ever tried to do that. But with a clear view of the line of storms just to my east, this seemed like a good time to give it a try. And it turns out it was:

The outing was worth it just for this shot! Note how, although this is a cloud-to-ground (CG) bolt, it is not visible all the way from the cloud base to the ground - the upper portion is obscured by the snow and graupel. With rain, you could probably see it, but it is obscured more easily by frozen precipitation. Apparently, the bolt occurred at an angle, with it striking the ground outside the area of heavy precipitation, but where it came out of the cloud it was back in the graupel and snow. In fact, the invisibility of the upper part of the bolt shows why it is hard to get pictures of lightning bolts associated with snow, even though they are there. In many cases, all you can see is a flash. But your chances of seeing a bolt are better with the more isolated thundersnow squalls that occur in the Rockies than with a thundersnow event in the Midwest or East - which is part of why I decided to try the lightning trigger. And I am glad I did.

I stayed at this site a while longer, but with no more precipitation and increasing noise from construction the ski area was doing in preparation for the ski season, I decided to move to a parking area along U.S. 160 a mile or so farther east. When I got there, things were quiet at first, but it soon became evident that a new area of precipitation was forming just to my west, so I decided to fire up the video camera again, but I was a little too slow. Just as big chunks of graupel began to fall, there was a loud crash of thunder - but my camera was just coming on, and by the time it was ready to record, the thunder was over. And then there was no more. Oh well - with the above picture and video, and with the experience of two thundersnow storms on September 8, I could not complain. So I decided to head home.

The storms were far from over, however, and I knew that I could still get some thundersnow later on after I got home. I thought it was too warm for that when I got home and a thunderstorm moved in within about fifteen minutes, around 3 p.m. However, the precipitation began as a mix of rain and either sleet or very small hail - I am not sure which. So I grabbed the video camera again, but by the time I got it going, it was all rain. This storm continued for more than 3 hours, with rain and thunder and lightning until around 5:45 p.m., at which time it mixed with snow and then turned to all snow around 6 p.m. I thought this offered yet another opportunity for thundersnow, but alas there was no more lightning. So this time, the thunder and lightning and the snow were about 10 or 15 minutes apart - close, but no banana! The snow turned back to rain around 6:15, and the rain ended around 6:30. When I checked the rain guage, there was 0.39" of much-needed rain. At that time I brought in the guage, because the temperature was forecast to reach the 20s overnight, so I was unable to measure the moisture content of the rain and snow that fell later in the night or, more accurately, early the next morning.

Wednesday, September 9

There was no more precipitation between the time I brought in the rain guage and sometime after midnight, but toward morning on Wednesday a new band of precipitation moved in, mostly wet snow. By around 8:30, there were spotty accumulations of around a quarter-inch of wet snow on elevated surfaces and some grassy surfaces, as shown in this picture:

Of all the various places I have lived in my nearly 71 years, this is the earliest I have ever had snow on the ground at my home. It was not much, but it was snow on the ground. Snow continued intermittently until around 9, after which it changed to rain and ended by around 10 or so. Since I had brought in the rain guage to keep it from damage from freezing, I don't know what the rain and snow water equivalent from this morning precipitation was, but other reports from around the area would suggest somewhere around .2" of rain and melted snow. I put the guage back out around 11 a.m., but only intermittent rain occurred later in the afternoon and early evening, adding another .05" of moisture in the ensuing 24 hours. In the late afternoon around 5 p.m. I went out to see what I could see of the snow in the mountains from town (near Village Lake). The mountains were still shrouded in clouds, but through the clouds you could see parts of the snow-covered mountains:

Thursday, September 10

The morning stayed dry, but forecasts and models insisted that by afternoon, new precipitation would develop in New Mexico and move northward into the area as energy rotated around the closed low. While no more precipitation had fallen in town, a report from Wolf Creek Ski Area indicated that, as of early morning, a total of 13 inches of snow had fallen since the start of the storm Tuesday afternoon. Through the morning, the closed low remained near the southern part of the CO-UT state line, but by afternoon it would start moving to the north-northeast after very little movement for nearly 48 hours. Several rainshowers did occur during the early afternoon in Pagosa Springs, and by shortly after 2 p.m., heavier rain and snow showers were developing east and southeast of Pagosa Springs in and near the south San Juan Mountains. I noted a report of heavy snow falling at Wolf Creek Pass, and the forecasts still indicated a possibility of thunder with any heavier cells. Since things seemed to be ramping up again, I headed out, hoping I might get another shot at thundersnow in the mountains, but also figuring that even if that did not happen, I could get some nice snow pictures there. As I headed out the east end of town, it appeared visually that some pretty heavy convective rain and snow showers were occurring, so I felt that there was a decent chance they might get strong enough for thundersnow. On the other hand, these showers were clearly not as high-topped as the ones a couple days earlier, since none of them were showing up on either the Albuquerque or Grand Junction radars, whereas the Tuesday activity clearly did. As I drove up to Wolf Creek Pass, I stopped a couple times to take advantage of openings in the clouds to see views like these:

These pictures were taken looking up into the mountains to the north and northwest from along U.S. 160 southwest of Treasure Falls. No snow in the valley there, but plenty up higher in the mountains. I continued on toward Wolf Creek Pass, noticing spotty snow on the ground once I got past the scenic overlook a couple miles past Treasure Falls. Gradually the snow on the ground increased, with the ground completely covered and snow on the trees starting around 9000 feet above sea level. By the time I got to Wolf Creek Pass, the temperature was at or below freezing and snow and graupel/snow pellets were falling intermittently. I stayed there for over an hour. I was hoping that the heavy activity I had seen moving in that general direction would give me some thundersnow, but it was not to be. Showery snow continued the whole time I was there, but it was never really heavy - light or very light most of the time, moderate for a couple short periods, with some accumulating on my windshield, but never heavy. When I got back later where I could get an internet connection, I learned that the closest lighting was well to the north, around the Gunnison/Crested Butte area. But despite the lack of thundersnow, the fresh snow was beautiful in Wolf Creek Pass:

These pictures were taken just a mile from where I caught the lightning bolt at the beginning of the storm just over 48 hours earlier - quite a transformation of the landscape!

Wolf Creek Ski Area sent out an email that included an update of their snow total to 15 inches by the morning of the 11th. The 15 inches of snow that fell there was one of the larger accumulations from this storm in Colorado, but not the biggest. The Monarch Ski Area west of Salida was the big winner with 18 inches of snow. A similar amount was reported near San Isabel, in the mountains northwest of Colorado City. But even some high mountain valleys did very well. Alamosa, in the San Luis Valley, had multiple reports of around 14 inches of snow. All in all, quite an impressive storm for September 8-10 when it is still, by the calendar, summer! The final band of precipitation that produced snow at Wolf Creek also produced some rather heavy rain in the area around Treasure Falls, mixed with snow at times. This was still going on as I passed back through that area around 4:40 p.m. on the way home. However, this band totally missed the immediate Pagosa Springs area. The earlier showers added another .05" of rain at my house, bringing the estimated storm total precipitation there to around 0.7 inches. Not a huge amount, but welcome nonetheless, and nearly as much as we got in the entire very dry month of August.

Return to 2020-2021 Winter Weather Observation Page