This case was a repeat of an earlier experiment but with a doubled along-line domain size to test the impact of the periodic boundaries (which impose a scale) on the simulated convection. The squall line is oriented north-south, and the view is from southeast to northwest. Small random temperature perturbations were placed, and along a cold front in the middle of the domain a series of storms formed, becoming the squall line you see here. Snapshots of the squall line at 2 hours, 200, 220 and 240 minutes are shown below. Note the viewpoint for 220 min. is somewhat different.

Shown in the top 4 images is cloud+rain (in white) with a colored "floor" indicating surface wind speed with respect to to the line (blue = easterly winds toward the line, and red = westerly winds out away from the line, towards the viewer). The deeper blue represents even stronger "inflow" winds found just east (ahead) of individual storms - the winds feeding the storm updrafts. The deeper red shows the westerly outflow winds behind the leading edge (gust front) of the storms. In the bottom 4 images, wind vectors (colored by speed, so blue represents lighter winds and red higher wind speeds) replace the colored floor. Resolution: 1.8 km (horizontal) by 250 m vertical; vertical scale exaggerated here by a factor of 3.

n95E: frontal squall line, 300-km domain

120min-floor	200min-floor	240min-floor	220min-floor
120min-vect	200min-vect	240min-vect	220min-vect

In the second case a smaller along-line domain was used, and a line of warm thermals were placed along the cold front, in addition to the random temperature bubbles used above. This line of thermals become a line of rather steady, rotating storms - supercells. All images are of the same simulated time in the evolution of the storms - just from different viewpoints. The view is considerably more "zoomed in" than the case above. Higher winds aloft are carrying the storm anvils off to the east above the viewer.

The transparent cloud+rain surface allows us to view the large-scale rotation within the storm, shown here as red surfaces. The red represents vertical vorticity of 0.002/sec - these are not tornadoes, by size or intensity, but large-scale rotation within the storm¹ (tornadoes cannot be simulated at the resolution used here). We have only shown the rotation from the surface up to 3 km, hence the cutoff "top" of the red tubes. Also shown are a few surface temperature contours (every 0.5 degree C) indicating the leading edge of the cool downdraft air, the gust front. The gust front is also visible in the surface wind vectors, which shift from westerly to southerly. Resolution: 1.8 km (horizontal) by 250 m vertical; vertical scale exaggerated here by a factor of 5.

n95d: line of supercells

View 1	View 2
View 3	View 4

¹The vorticity in this case is at most 0.005/second, or one-half the criteria for a mesocyclone. Of course, the horizontal resolution here is 1800 meters. At higher resolution 0.01/second would probably be met. The rotation is fairly steady, with continued updraft growth on the south side of each tower.