Photos and and accompanying narrative by Roger Edwards.

For additonal information on this event, visit the following web page.



Elke and I moved north from Elm Creek to intercept the supercell shortly after it formed over northern Custer County, and had excellent views of its evolution, which included three fascinating and classical occlusion processes. The most illustrative shots have been selected for both the Sargent and Comstock tornadoes. [Elke shot slides, talked to
local farmers/spotters and admired the storm while I shot slides, digital stills and tripodded video. Best of all worlds!].

Although the photogenic tornado near Sargent looks "prettier," I think the rain-wrapped Comstock photos (both with and without digital enhancements) may prove more useful for educating both spotters and chasers. The images below are digital stills using the Nikon 880.


Our observing position for the Sargent wall clouds and tornado was 10 S of Sargent on US 183 looking WNW-NW. The supercell had developed one wall cloud (at left), which rotated vigorously -- at time with ragged, scuddy protuberances near the ground, but no confirmed tornado. By the time we set up for photography, this already was a classic supercell with flared updraft structure, "visual vault" and cloud tail at right, and considerable precipitation evident on the forward flank at right. A new wall cloud developed to its ENE (to my NW), which is at lower center in this initial wide-angle photo.


Both wall clouds were rotating slowly in this view, though the circulation in the older (left) one was weakening.


A wide angle view showing the wall cloud (lower center) associated with the strengthening mesocyclone which would produce the Sargent tornado. The old occlusion continued to weaken at lower left.


A wide angle view showing much of the storm structure, featuring the second and now menacingly low wall cloud (which was rotating vigorously) 2-3 minutes prior to the Sargent tornado. A new cloud base is evident in the foreground, in front of the wall cloud and to its near right. This new updraft eventually would wrap in rain and produce the Comstock tornado.


A zoom view of the pre-tornadic wall cloud. Individual scud elements on its left (SW) side were rising and moving rightward, while scud on its right (NE) side were moving left and up. The entire wall cloud was rotating quite furiously. Because of entrainment of increasing volumes of relatively dry, occlusion downdraft air, this wall cloud actually would raise its base and become more ragged before producing the tornado.


The Sargent tornado, which had begun a couple of minutes before as a
partially sunlit rope, widened to this shape. In this zoom, the occlusion downdraft's clear slot is evident as sun shines on the left side of the wall cloud. The lowering at right was rising but not rotating (therefore, no funnel).


Normal 50 mm focal view of the tornado and surrounding storm structure. Crepuscular rays shone before the vortex, emanating from the clear slot. The new (future Comstock) base remained in the foreground.


The tornadic wall cloud continued to rise as it entrained more occlusion
downdraft air; while the tornado's condensation funnel reached peak width. There is some evidence of partial vortex breakdown aloft in the fuzzy bulge, about halfway up the visible tornado.


A zoom of the tornado at apparent peak form.


Wide angle view of the tornadic storm, which by this time more resembled two supercells in one. While the Sargent tornado continued at rear, the new updraft in foreground gradually attained more bulk and was slowly rotating. Tail clouds are attached to the right (NE) of both updraft areas, as is their collective forward-flank precip plume.


Near the end of its 11-minute lifespan, the tornado gradually roped out. A short-lived funnel appeared above and to the left of the tornado in this view. The newer updraft is still in foreground with its "visual vault" region at right.


A zoom of the roping tornado and secondary funnel, the latter of which did not touch down or produce visible debris.


A lowering from the Sargent mesocyclone is evident at rear, to the left of the forward-flank precipitation area. This lowering exhibited mainly rising motion, with weak rotation, and no further touchdowns were evident from that occlusion. The new mesocyclone (base in foreground) began wrapping precipitation around its N side.


We had just repositioned to our observing location for the Comstock tornado, 11 S of Comstock on NE 70 looking NNW. The mesocyclone (at lower right) evolved slowly -- over 30 minutes had elapsed -- and became resoundingly HP in character, with considerable wrapping precipitation throughout.


Here is the unadulterated original, "as the spotter sees it," and an enhanced version of same. In this zoom view, a pronounced bowl-shaped lowering is evident behind wrapping rain, and behind a clear slot. Persistent visual continuity, over the span of several minutes, established that this lowering was rotating. The enhancement of the same photo shows the lowering well. There may have been debris under this circulation this soon; but it would have been diffuse and hard to discriminate from the precipitation.


A funnel developed and lowered (see annotation on the enhanced version). There might be debris beneath, partly or wholly smudged from view by precip. The foreground rain curtains were moving left to right, orbiting the mesocyclone and making it very hard to distinguish debris from wrapping rain, even on close concentration or in (low light and low contrast) video.

The same funnel continued moving ESE, as we watched rain curtains wrap around it. After this shot (with accompanying contrast-enhanced version), the condensation funnel touched down and firmly planted itself; but for much of the time, thick precipitation curtains allowed us to see only parts of the tornado. What we could see, however, showed obvious and vigorous rotation

Viewing was very difficult by this time, with the tornado deeply rain-wrapped, but growing bigger and moving in the general direction of Comstock. We could see only the left (W) edge of the tornado at the moment of the photo; but in the contrast-enhanced version, the right edge is faintly visible also. Outflow from the westerly occlusion downdraft raised some dust, best seen in the enhanced photo. I have observed similar RFD/OD dust plumes sweeping around the S semicircle of mesos associated with a few other fat, rain-wrapping tornadoes, including Spencer SD (30 May 1998).

After a few minutes, we could see the tornado somewhat better with our eyeballs; and it was a very wide vortex by this time


This was the last time we had decent visibility of the tornado, and even then, it was difficult to track through the increasingly dense, wrapping precip curtains.


We lost visibility of the tornado altogether after this shot -- where it takes contrast enhancement just to clearly make out one side (the E side). The tornado probably lasted at least a few minutes after this, but the entire mesocyclone became so thickly rain-wrapped as to obscure the tornado altogether. No power flashes, which we usually would seek, were evident. I am very curious as to when and where this damage path began and ended.


The HP supercell as it moved from Custer into Valley County, now seen from 11 N of Loup City. No other tornadoes could be seen. Shortly after this, we abandoned this storm for a brief intercept of a short-lived, smaller supercell that formed on its outflow boundary (and soon, was undercut by the same outflow). Only small, nonsevere hail was observed along NE 92 near the Sherman/Custer County line with the latter cell. After some sunset photos, we were ready to call it a day, and consume traditional, post-tornadic steak dinners in Kearney.



The End