Rare Alaska Funnel Cloud on Independence Day
Social media lit up on 4 July 2026 with video of a strong rotating condensation funnel from Cantwell, Alaska, on the southern edge of the Alaska Range. You can see the video in this news story from KTUU-TV in Anchorage. This funnel was seen around 5 PM local time, or around 0100 UTC on the 5th. It didn’t didn’t quite reach tornado status as it never appeared to come into contact with the surface, and thus no damage was associated with this event. The NWS Weather Forecast Office in Fairbanks released the following public information statement about the event (reproduced here for accessibility and permanence:
498 NOAK41 PAFC 060027 PNSAFC AKZ701>704-711>714-721>729-731-732-735-741>757-761>766-771>773-781-785-787- 791-795-061227- Public Information Statement National Weather Service Anchorage AK 427 PM AKDT Sun Jul 5 2026 NWS Anchorage received a report of a funnel cloud spotted south of the Alpine Creek Lodge along the Denali Hwy on July 4, 2026 at about 5pm; the report was received at about 5:49pm. The funnel was visible for about 15 mins by onlookers, extending from the base of a towering cumulus. Referencing videos and photos, there are no visible signs that the funnel was in contact with the ground. Radar shows that the parent cloud reached about 30kft and produced a single cloud-to-ground lightning strike at 5:14pm. Based on area observations and local topography, this funnel was likely generated by Susitna Valley winds flowing against storm motion winds, creating enough wind shear to generate a funnel cloud. If more information is provided that confirms contact with the ground, this event could be reclassified. $$ PP/AB
Geostationary satellite observations are challenging at this latitude. Cantwell is at 63.4 degrees north. The geostationary spatial resolutions are defined at the sub-satellite point over the equator. As long as you’re close to the equator the 2 km infrared resolution is pretty close. However, as the satellite scans further north, the pixels become more elongated. By the time it reaches Cantwell, the pixels are stretched to almost 6 km in the north/south direction. We can see the effects of this in the relatively coarse pixels of the visible imagery in the following loop. This is Band 2 from GOES-18, so its nominal resolution is the best we have at 0.5 km. However, at Cantwell the pixels are about 3 times the nominal resolution. Cantwell is in the center of the map, due east of the Denali National Park.
It may be challenging to interpret this loop as we’re seeing low convective clouds over a snowy mountain range, and thus we’ve got white on top of white. This is why it’s worth looking at different products to gain a fuller understanding of what’s going on. This next loop shows the Day Convection RGB, which is used to discern if convective clouds are actively growing deep. In this case, the Day Convection product is useful because it’s helping to discriminate the clouds from the snowy surface.
From this, we can see that this is becoming quite an interesting case. Normally, we’d expect deep convection to drive such a well-defined funnel cloud. But this is Alaska, and things are different up there. Prior to this event there have been only six recorded tornadoes over Alaskan land (Wikipedia has a list of tornadic events in Alaska, and the 19 April 2024 event was documented on this blog). The typical Great Plains supercell driven by the intersection of warm, moist Gulf airmasses and the edge of cooler, dry air masses generally doesn’t form in Alaska. Instead, the driving force for this particular event, as the NWS statement above describes, is atmospheric motion interacting with the local terrain. Take a closer look at the motion of the clouds, using this image as a key. The Alaskan Range arcs across the middle of the loop. There’s some lower level easterly flow that appears to be trapped by the terrain while the deeper flow is coming up from the south. This creates some significant localized wind shear that can force rotating storms even in the absence of deeper convection.
There was a VIIRS overpass a few hours before this event (around 22:30 UTC), and that true color view of that is shown here. Note how there is much greater detail in the VIIRS view, although there’s nothing in this image that makes it obvious that such an unusual event will be shortly taking place. Still, it is good for forecasters to consult the VIIRS imagery when it’s available so that one can evaluate the coarser geostationary imagery with some better context as to what’s going on.
