The following comes from: Fujita's paper on the Palm Sunday Outbreak
 |
| Figure 46 Click to Enlarge Shows the development |
When an aerial photogrammetric survey of the damage by tornado 5-2, which devastated the Midway Trailer Court (figs. 22 and 23) was taken, extremely severe damage could be traced along two paths that seemed to represent those of the twin funnels photographed by Huffman (fig. 2 4 ) . Standing at the same spot between US-33 and the New York Central Railroad tracks about 0.7 mi from the Midway Trailer Court, Huffman took a series of six pictures, which, gridded with azimuths and elevation angles for every l 0°, are shown in figure 46. The grid lines were computed after a visit to the photographed site by Fujita.
Huffman's site and the damage path are shown in the topographic map at the top of figure 47. His first photo- graph (fig. 46A) was taken looking almost due west. His second reveals a single funnel which in the third picture, when examined carefully, shows some evidence of splitting. The fourth one (fig. 46D) shows twin funnels on both sides of the highway, giving an impression that US-33 runs through a tunnel between the funnels. The fifth photograph very clearly indicates the patterns of stratified low clouds wrapping around the twin funnels, permitting us to determine the direction of both funnels to be identical and cyclonic. The patterns also give dimensions of circulation that are closely related to the tilt of the tornado axis. The best estimate of the tilt is 39' north-northeast.
 |
| Figure 47 - Damage path and debris marks near the Midway Trailer Court indicated by letter M (top) Click to Enlarge |
In figure 47, the middle chart represents photogram- metric positions of the funnels appearing in Huffman's six pictures (fig. 46). Black circles designate the initial funnel touching the ground; the small circles, the complemental funnel that appears in the third picture. The sequence reveals that the funnel near the ground increased rapidly in diameter between the first and second pictures, then began splitting in two. After the split, the two funnels rotated about each other around their common center. The bottom drawing in figure 47 illustrates the rotation and the photographic directions. The above offers a speculative explanation that the split in the funnel was caused by the rapid increase in the funnel diameter, while the tilt of the funnel axis was in excess of 30°. From the translational speed of the tornado, about 50 mi hr-', the estimated time intervals between successive pictures in seconds are : numbers 1-2, 17 sec; numbers 2-3, 27 sec; numbers 3-4, 13 sec; numbers 4-5, 8 sec; and numbers 5-6, 31 sec. Thus the increase in funnel diameter between the first and second pictures took place within no more than 17 sec. When a tilted column of rotating air increases its diameter very rapidly, the air parcels near the ground cannot move around the center. The motion beneath the tilted axis, especially, is restricted because of limited flow space and surface friction. As a result, the funnel may quickly take the shape of a shortcut circulation while the rest of the vortex starts forming another funnel. This began to occur at position 3, and when position 4 was reached, the initial vortex was dying out rapidly while the complemental vortex intensified. About 50 sec after the funnel had started to split, the twin funnels changed again into a n almost single one at position 6, the entire process having taken less than 1 min.