Perham to Melrose, Minnesota and Northome, Minnesota
Like a lot of weekend chases, I watched this event several days ahead in hopes of having a chase day. On Thursday (August 1), the SPC added marginal risk over central Minnesota. I thought that looked good. On Day 2, they maintained the marginal risk but had no risk outlined for tornadoes. I thought there was an outside chance and was slightly surprised there wasn't a 2% area. On Day 1, the risk was upgraded to slight, and a 2% tornado area was added in the 1630 outlook. Also, a hatched area was in the outlook for hail, which means some very large hail is possible.
An event summary from the Twin Cities office of the National Weather Service can be found here.
Categorical severe weather risk outlook from SPC issued at 11:30 CDT.
Tornado risk along with tornado reports.
Hail risk and reports.
In the morning, some storms were expected over northern Minnesota, and this area of elevated thunderstorms was expected to gradually work southward during the day, and surface-based storms were forecast to develop in the late afternoon just south of the ongoing, elevated storms. As always, I delayed as long as possible as I was reluctant to make any drive south of I-94. I wanted to get some maintenance work done at home, and I wanted to come home and have a good night's rest so I could get things accomplished on Sunday.
I allowed the elevated storms to move through. These elevated storms produced an elevated shelf cloud, which looked pretty nice over my house. These storms were occasionally warned for large hail, and a report of two of 1.75" hail was received. I also let this area of storms get to the south of me, which meant I might have to maneuver around hail cores on my way to the target area.
Elevated shelf cloud moves over my house. I got 0.10" rain from this storm.
I left around 2:00 PM, but I didn't have any problems with hail. Other than some wet roads, it was a pretty easy drive to the target area. I stopped in several places to reassess the surface observations and satellite picture. Satellite showed a considerable lack of cumulus in the warm sector, but there were some pretty flat cumulus along the warm front in the Fargo area. Surface obs showed dewpoints in the 70s along the Minnesota/Dakotas border along with south winds. My goal was to keep track of this moist axis and adjust where I needed until initiation of surface-based storms occurred. My initial guess was around Detroit Lakes, but I adjusted southeastward a bit as the front approached. I ended up in Perham. There was a Shell charging station, so I made an attempt to top up on charge before the chase started. I failed to get the charger started before a surface-based storm appeared immediately to the northeast. The chase was on!
The storm as it developed just east of Perham, MN. Initially, the base was pretty flat, and the storm looked linear at first. That was about to change!
The storm was initially barely discernable from an elevated storm. It had a flat, somewhat high base, but the radar presentation looked more surface-based in that it moved more slowly than the elevated storms to its north. It didn't take long before the storm began to look better. We moved southeastward on U.S. 10, and when we stopped near New York Mills, a wall cloud developed. From that point, the visual presentation of the storm was very much supercellular.
Here is a radar capture of the storm right before it developed a wall cloud and more supercellular visual features.
Visually, the storm is pretty clearly taking on supercell characteristics.
The storm was just north of Highway 10, but its movement would probably take the core across the road at some point. U.S. 10 also curved a bit more east, so if the storm did not turn right, the highway turned left. The radar echoes started to reach over the highway, but visually, the core was a few miles to the north, so I stayed on U.S. 10 until Aldrich, and then stair-stepped southeastward on county roads. At the first stop on the county roads, the storm had developed a nice, beefy RFD shelf cloud, a low-level inflow tail, and a notch in-between the two to our north. The inflow tail had some nice, rapid motion into the updraft base.
The RFD shelf cloud is looking pretty interesting at this point.
I wanted to stay ahead of all this so I could see into the notch. The county road network was dense enough to allow us to do that. A couple stops later, we were in a great position to do this, and I was getting out of the car to take stills when some cloud-to-ground (CG) lightning strikes happened nearby, and in rapid succession. It was a CG barrage! The barrage didn't last too long, but I got video of most of it. That was fun!
This was the site of an intense CG barrage. I have it on video at the end of this account.
Radar had shown a new, smaller storm west of ours, but it was now showing this second storm to be intensifying rapidly. At some point, it might be a good idea to drop back to this second storm, but for now, our storm still looked good and had some nice structure. Plus, our storm still appeared to have better rotation. Sometimes, the eastern storm has more uncontaminated air and then lays down and outflow boundary that reduces the moisture quality in the air being ingested by the second storm. On the other hand, the moist axis was a bit west today, and if there was no major outflow from the first storm, the second storm might be better.
Radar was now showing two distinct supercells. We were on the eastern storm but starting to think about going to the western storm.
Another stop to look north at the storm two minutes after the above radar image.
Our storm was looking great and continued to have rapid motion in its inflow tail. It also had some hint of occlusions within the partially rain-filled RFD within the mesocyclone. At this point, the road network had a gap between south options as well as a more considerable tree-filled area. This forced me to run east to Highway 10 and then turn southeast on 10.
Last decent look under the Staples mesocyclone before we ran southeast on U.S. 10.
We got to 10 and made a bit of a long jump to the southeast. I stopped to have a look back at the storm, and at this point, it began to look a little more featureless. There appeared to be a longer, less distinct shelf cloud. I decided to continued southeast on 10 and charge for a bit at Little Falls while the storm approache. I told Julia that I was about to do this, and she asked about getting to the western storm-- when might be the best time to do that. We needed a direct south option followed by a good west option. We could go to Little Falls and go west on Minnesota 27.
Just past Randall, I changed my mind and took a county road back west a couple miles and caught another county road south. From here, I picked county roads, as they became available, until one county road was closed and under construction. On the map ahead, I saw a diagonal road, and that would be a good one to take to a better south option. It turned out that diagonal road was gravel. Julia decided she didn't like that and exited that road for a different south option. I kept going and found the roads that would stair-step me southwest in front of the western storm. It was necessary not to get too close to it. There were 4-inch hail reports in Wadena!
A picture of the Melrose storm as I was getting in front of it.
I gradually got in front, but I was now running low on charge. I had a few options: 1) I could go back to Little Falls and charge there (only really an option on the way back home, and I was still in chase mode), 2) I could follow the western storm southward and charge at Willmar, but this was the most distant option by a few miles, and 3) I could swing around the southwest periphery of the western storm and drive northwest to Alexandria where I would end the chase. Reluctant to drive farther south, I decided on the third option. I hoped to simply hop on I-94 westbound to make a swing in front of the action area, but I wanted to give this big hail-producer a bit more room, so I continued south through Freeport then took County Road 30 westbound through New Munich. I got through New Munich and turned right on Country Road 173 to look north at the storm.
There had been a somewhat long shelf cloud when I initially got on the storm, but now there was a nice mesocyclone and wall cloud to my north. I got out to take some stills. Once I had clicked off a couple pictures, I gazed at the storm a little longer and realized that mesocyclone to my north had a lot of rotation. I quickly hopped back in the car and zoomed my video camera in on the action area. I thought there was a developing tornado since the rotation was certainly fast enough for one. If the circulation tightened, there could be a condensation funnel to the ground at any moment. However, rain was quickly filling in, and I wondered if I was getting excited too easily. I kept the camera rolling, giving the storm about a minute to do its thing. The rain- (and possibly hail-) filled RFD was approaching, and I didn't want to take chances with hail, so I was anxious to leave. Reflectivity was also increasing rapidly in some new cells to my west, and with rain filling in the area to my north, and my car getting low on charge, I decided to make my exit.
I hopped out of the car to take a quick still image then realized this area was rotating quickly.
Here's a still image capture from my video. Rapid mesocyclonic rotation was visible.
I drove west and made another, final stop before entering the heavy rain to see if anything might reveal itself to my north, but the wet RFD had advanced, so I could no longer see far enough around the RFD precip arc. I took a couple stills of the dramatic shelf cloud and hopped back in the car, entering the rain area. Sure enough, my phone rang with an EAS tone. The national weather service had issued a tornado warning. The radar velocity signature was pretty indicative of a tornado, now showing a fairly tight couplet, albeit a little broad for a tornado. At this point, I lamented my decision to drive to Alexandria rather than Willmar because now I was fully committed.
I was interested in looking at what the video had caught. Also, geolocating some of the damage pictures that showed up on Facebook, X, and on the National Weather Service web page, I was able to find the tornado on my video, after a significant amount of contrast enhancement. At first, I thought a cone near the center of the video was probably the tornado. However, given the direction I was pointing in that part of the video, it did not line up with the damage. What did line up was a second area of lowered condensation (to my visible horizon) farther west. I labelled these two areas on the image below.
The tornado is the dark area on the left. The cone on the right looks like a tornado, but its rotation is less intense.
Annotated view from my position that I used to label the image above. The blue lines are my lines of sight to the tornado on the left and the other "cone" on the right.
Once I left my initial tornado viewing spot, I stopped to get another look into the notch before the wet RFD hit me.
Here's a still image I took right after I passed through the wet RFD core.
This is the radar loop around the tornado time.
It didn't take me too long to exit the heavy rain. Once I was out, I looked back to my east and watched the mesocyclone to my east, which had some fast north to south motion. There was a tornado embedded in there, but its damage track was probably at its end at this point. The wind was strong from the northeast, and my car told me I would need to charge BEFORE I reached Alexandria, due to these strong winds. I decided to wait a few minutes to allow the storm to get farther southeast and take those strong winds with it. I began a slow drive to Alexandria, keeping below 50 mph on the country roads up to I-94 and then staying at or below 60 mph on I-94. The winds cooperated, diminishing to nearly calm, and I also entered a construction zone, so I had to drive 60 mph anyway. I made it to Alexandria with 8 miles to spare. Making it to Willmar from where I saw the tornado would have been a mile beyond that. However, had I decided to go to Willmar, I likely would have taken a slightly different route.
I charged up to 90% at Alexandria while I walked over and grabbed dinner at McDonald's. I got back in the car, finished my fries, then drove off toward home. I forgot about my burger for quite a few miles.
This is the part of the chase on August 3, 2024. Then, the clock struck midnight.
The Chase Goes On!
I made the mistake of checking the solar wind data. It showed there might be auroras. Bz had been hovering around -5 nT for a few hours, and although the solar wind speed was only around 350 km/s, the duration of southward Bz would likely make for auroras. Additionally, the moon phase was nearly new, so the skies would be pretty dark wherever there was any clearing, and satellite loops showed clearing to my north.
Indeed, when I got on the stretch of Minnesota 27 between Long Prairie and Little Falls, it had cleared, and I could see an auroral arc, out my driver side window, to my north. I pulled over on 20th Avenue, next to the Twelve Mile Tavern, and shot a few aurora pictures to document the arc.
Auroral arc first seen on Highway 27 between Long Prairie and Little Falls. This shot is from 20th Avenue at Twelve Mile Tavern.
I continued northward, but satellite pictures showed I would run into some broken clouds. I decided to continue north into whatever cloudiness there was because there was a chance there could be breaks, and since Bz wasn't too far south (around -5 nT), the show would probably not be spectacular, and if the clouds ended up being too thick, I would not be missing much. Also, there was a fair bit of artificial light contamination this far south (although my Twelve Mile Tavern shot didn't look too bad).
Well, so much for that idea. Bz dropped south to the mid negative teens. I was now more interested in how things were evolving. I stopped just northeast of Brainerd to take another shot, and I also stopped on Highway 6 at the Roosevelt Lake public access near Outing. These both indicated an auroral arc relatively high in the sky for the -5 nT Bz, but now Bz was dropping pretty fast. The clouds were also thickening as I drove north.
Auroras just northeast of Brainerd.
A small clearing over Roosevelt Lake near Outing reveals auroras, but the clouds are thickening.
I kept checking satellite and looked up, hoping for a break when I reached Remer. However, the overcast was now solid, and there were no auroras or stars visible anywhere in the sky. Satellite showed the closest clearing would be in northwestern Itasca County. Did I have enough charge to make it there and home? I definitely would if I drove slowly. I was feeling reluctant to go so far, knowing how tired I would feel the next day. However, I skipped my turn home on County Road 63 and took the cutoff to Deer River instead. From there, I could make it to at least Northome, which was a bit of a safe estimate because satellite showed the clearing should be closer to Squaw Lake, which was not as far.
Indeed, between Squaw Lake and Alvwood, I looked up to see stars, but the Avenue of Pines is named that for a reason. I had trees on either side of me and not many views to the north. At Alvwood, the highway curved due north, and I could see the auroral arc ahead of me.
Auroras over the road at Alvwood, MN.
After snapping a few shots, I continued further until I had a more open view. There is a farm between Alvwood and Northome, and that is where I stopped. I spent about an hour or so there, and I took a bunch of images pretty much right on the road. A couple cars drove by while I was there, but otherwise, I had an uninterrupted shoot at this location.
Auroras over a farm south of Northome, MN.
Auroras get brighter.
A shot as the second car begins to appear to my north.
Here's a cropped version of the same shot above.
The clouds suddenly redeveloped and moved in from the northeast, so I started back toward home. Satellite showed the clouds should be disappearing now farther southeast near Winnie Dam, so I decided to stop there for one last set of images before calling it a night. On the final stretch of road to the dam, I drove back into the clouds. When I hopped out of the car at the dam, I could only see the tail ends of the arc on the northwestern and northeastern horizons. However, when I snapped a couple images, they revealed the clearing line was getting closer, so I kept shooting as the clouds backed away. By this time, the north-northeastern sky was brightening quite a bit and washing out the auroras. It was time to put the camera back in the car, go home, and get a few hours of sleep!
Auroras at Winnie Dam before the clouds cleared there.
Auroras over Winnie Dam. 4:30 am and twilight is here. Time to drive home!
This part of the drive occurred after 12:00 AM August 4. I got home about 5:00 AM.
