Hurricanes ...are quite well understood in their basic anatomy.
That's not what a tornado is.Here, a warm air mass goes up and over a cold air mass (it's generaly lighter). Friction between the two creates a roll of rotating air and this tips over and becomes a column which may touch down (as when it sees a trailer park).
The above explains nothing in functional terms. So ...
[ Note: The following is a possible model for an air effect similar to a tornado. But it's not how a standard tornado forms. That comes from a downflow through a low pressure area in a vortex formed by two air masses moving in different directions. The tornado forms in the "hook" as the meteorologists call it.]
What is an air mass?
One place is hot ... another cold (or dry or humid or any combination). There is a boundary between them. One might then inquire as to why the two don't mix. After all it's just air and air mixes easily from anyone's experience, doesn't it?
Yes and no ... Our experience is with room size bodies of air and these do mix quite well. But it takes some amount of time. If the room were the size of Connecticut, the time needed to mix completely is enormously longer than a room. It's the same stuff a basically the same temperatures so it mixes at a fixed rate (fixed in the sense of the 'big picture').
Therefore, for large areas, the dividing line between two differring areas might as well be solid given the small percentage of mixing that occurs over, say, two or three days time. If you've ever been in a jet, you may have seen "cliffs" of clouds showing obvious demarcation between "air masses".
What happens when two different masses collide?
There is friction between the cold and hot masses and the roll develops. This constitutes another "air mass" (differentiated by it's rotation aspect). Note here also that this new mass is lighter (less dense = generally lower air pressure) than the cold air mass.
Fricton at the hot/cold boundary slows relative air velocity there ... creating a very low pressure area in front of where the roll will form. This causes the "over the top air" to rush ahead even faster to fill that "vacuum". The air on the top of the roll is therefore faster than on the bottom and the whole thing is pushed along by the over the top air mass at no less than it's velocity. The roll then has a rotation velocity faster than the relative velocity of the colliding air masses.
By itself, this "roll" is harmless. We need a mechanism to get to the ground. It needs to be tipped over 90 degrees as a single unit and brought down to the trailer park.
The mechanism involved is not electromagnetism.
What mechanism then?
All the lighting and thunder is mere storm dressing ... all sound and fury signifying nothing (like the head gear worn by a Las Vegas showgirl). We're just interested in the "body" of the mechanism. The true mechanism is simple air pressure considerations and a third collision.
We need another air mass collision parallel to the ground but perpendicular to the "roll". Somethinng to give it a "turning moment" ... then ... it will precess like a top and get pointed in the right direction - at the trailer park.
Now, you might think ... "Precess like a top? Why doesn't it just "disintegrate"?
Because to disintegrate it would have to change shape. For instance, what if it just precessed in small slices like a "slinky".
No good ...
The volume of the cylinder laid down in this manner is less than the former. The gas must then be compressed but there is no force to do compression at this stage. Therefore it may only rotate as an integrated, independent "air mass".
Once rotated, even though it has less weight than a comparably sized air mass near the ground, it will still "float" on and in that mass like a heavier-than-water bar with a cork on one end. Some of it will extend down to ground level and bob up & down wherever "atmospheric waves" may take it.
As it drops down it should have a funnel shape because the lower, denser air mass is compressing the "vortex". When it stops dropping the funnel shape should disappear into a smooth walled cylinder at maximum compression.
Note: Angular momentum equals mvr ... so, since the mass is constant, the velocity of rotation must increase in proportion to the decrease in radius. If the funnel is compressed to 1/4 it's original diameter it will have a velocity 4 times original [if 60 mph at it's initial "roll" ... then ... 240 mph after compression].This model then requires three things for a tornado to occur
The hot air, the cold air mass and another air mass (of undetermined size & type ... possibly just an asymmetry in the cold or hot air mass causing the roll to stick a little here and there) skewing the rolled air to make it "stand up" like a top.
As air in the storm roils about, there will be swirls where the t-roll will sit comfortably, i.e. rotate in the same direction. Most swirls will tend to be affected by coriolis forces (as in hurricanes and water going down your tub drain) so that non-reinforced rotations will be less likely to occur ... hence, the (non-exclusive) preference for counterclockwise tornados in the northern hemisphere.
Why don't tornados disintegrate by friction with the surrounding air?
They do. They don't last toooooo long. Not like a hurricane. But it does take some time ... after all it's a gas-gas interface, i.e. pretty friction free. They can be stopped much sooner, however ... if you put lots of aluminum trailers in their path.