Scientists not involved in the study said Tao's proposal is based on outdated ideas about how tornadoes form and ignores important geographic features.
"He's got the basics almost completely wrong," said Harold Brooks, a research meteorologist at the National Oceanic and Atmospheric Association's (NOAA) National Severe Storms Laboratory in Norman, Oklahoma. "The reason there are so many tornadoes in the central United States is southerly flow at the surface bringing warm, moist air northward, and westerly winds from the Rockies bringing relatively cool, dry air above that, not some surface 'air flow clash location.'"
Markowski also chafed at the notion that stopping "air mass clashes" could stop tornadoes. (In fact, he recently co-wrote a paper in the Bulletin of the American Meteorological Society that offers a long argument about why people should drop the term "clash of air masses" when explaining tornadoes.) Air masses clash all the time, and they don't always spawn tornadoes, Markowski told Live Science. What's more, tornadoes can develop even without an air mass collision. Rather, storms with tornado potential form when warm, humid air near the ground gets trapped under drier air. This instability can produce a spinning updraft of air, which sometimes leads to a tornado.
Semantics aside, the scheme ignores already existing east-west mountain ranges, such as the Wichitas, Arbuckles and Ouachitas in Oklahoma, Brooks said. And roving air masses that go on to seed tornadoes would be able to clear a wall of the size Tao proposed, other scientists said.
"Air masses routinely pass over the Appalachian Mountains. This is true in winter, when the air masses are much colder and heavier than they are in the summer," said Matthew Parker, a storm researcher and associate professor at North Carolina State University.
One of Parker's graduate students at N.C. State, Brice Coffer, actually put Tao's proposal to the test in computer simulations. Coffer used the Weather Research and Forecasting Model, a system commonly used to make high-resolution thunderstorm forecasts, to re-create a storm that spawned deadly tornadoes in Oklahoma in May 2013. He watched how the storm virtually unfolded in three different scenarios: one with the type of walls Tao proposed, one control with no walls and one with 1.6-mile-high (2.5 km) walls — just to play devil's advocate. [Skyscraper Storms: 7 Big City Tornadoes]
Coffer found that the 1,000-foot walls had no significant impact on the formation of storms in the region; that simulation looked nearly identical to the control. The exaggerated, 1.6-mile walls, however, do block the air — but with unintended consequences.
"If you put a rock in a stream, the same type of thing happens," Coffer told Live Science.
If the rock is small enough, the water will just flow over the rock. But plop a big enough rock into the water, and the stream will just go around it on the sides. That's would happen with a mile-high wall. The air would move around it instead of over it, shifting the storms eastward to the Mississippi River Valley instead of the plains and turning much of Texas into a desert, according to Coffer's model. What's more, strong circulations would occur at the edges of the walls, producing landspout tornadoes, which don't emerge from a supercell like regular tornadoes do, Coffer found.