Mini-terraforming against tornados

There are regions in the world, also in the US, that are prone to devastating weather phenomena, such as tornadoes. There are probably numerous physical models attempting to describe this in the attempt to predict their formation. I propose using these models in an attempt to subverting them, not only predicting.

Some regions, even near those that are prone, have smaller probabilities of being harmed. I believe that one of the possible reasons is the landscape, such as hills, mountains and valley. These weather phenomena are highly dependent on air pressure, and change with height of the landscape. Furthermore, who has not heard of the "butterfly effect"? While its origin is in the chaos theory and it is usually misinterpreted, I want to deliberately make a "bad" use of it for this project.

I propose that a "sleight" change in landscape may cause a drastic change in the development, progression and severity of weather phenomenon, such as tornadoes. While I have no proof of this, I believe current models can, or should if they don't, analyze effects of changing altitude on such events. For example, given a known prone area, what would a small hill or a valley do to the formation and progression of a tornado?

Going one step forward, if/when such models exist, one can also try to optimize the process, by finding "what is the smallest change in landscape that could have a beneficial effect on the phenomenon, such as guidance or diminishing of magnitude?" In other words, if I could create a hill somewhere, where should I put it to guide a tornado out of residential or other important areas?

While this question seems bizarre, it has both scientific and operational significance. The science is that one could better understand the phenomenon by analyzing effects of landscape, which is a crucial (in my opinion) parameter on the dynamics. The operational one is somewhat more challenging. I have seen a "garbage mountain" that was several hundred feet high and a mile long. Furthermore, tunnels, highways and railroads completely change the landscape in drastic manners unfathomable a century before. I believe we now have the technology and the means to actually change the landscape of the environment around us. This is what I call "mini-terraforming": while it doesn't change the entire Earth habitat, I believe it can change some of it large factors, such as weather phenomenon.

Taken together, the optimizing landscape change for the mitigation of tornadoes and the ability to actually create or drastically change the landscape, I propose that a test be performed in the following manner:

a.       Take a tornado-prone area.

b.      Analyze the effects of possible landscape changes on the development and dynamics of tornadoes.

c.       Optimize landscape changes, i.e. what is the minimal change required to mitigate the devastation such a tornado does.

d.      Try to create/change the minimal landscape in the optimal position.

e. Wait for next tornado season and hope for the best.