A radar does not see a tornado.
It sees air moving toward the antenna and air moving away, side by side, in a tight pair.
Find that pair low enough to the ground and you have a warning worth issuing.
Which works right up until something else in the same patch of sky is also spinning.
Central Illinois has a lot of turbines.
Two of its tornado outbreaks crossed directly over them.
Why the filter that removes buildings lets turbines through
Radar has always had to deal with ground returns. Towers, silos, grain elevators and hills all bounce the beam back.
The fix is a clutter filter, and it works on one simple rule. Anything returning a signal with little or no motion is not weather, so remove it.
A wind turbine breaks that rule on purpose.
The blades reflect the beam and they are moving fast, so the processor reads them as a target with reflectivity and motion, which is the definition of weather it was given.
So the returns stay in. They land in the reflectivity field as false precipitation and in the velocity field as false wind.
And a rotor turning through a vertical circle produces a spread of speeds toward and away from the antenna, which is exactly the pattern a forecaster is hunting for.
Where the beam actually is when this happens
The obvious assumption is that the turbines are blocking the view. They are not.
Over the affected counties the two lowest weather radar scans sample the atmosphere at about 1,200 and 2,000 feet above the ground, well clear of the machines below.
The beam is roughly one degree wide, and the antenna steps through elevation angles from just below the horizon up to nearly 20 degrees.
What reaches the receiver from a turbine is sidelobe energy and returns from the edge of that beam, not a shadow.
Distance decides how bad it gets. The federal radar office draws a no build ring at about 2.5 miles, where a nacelle can genuinely block the beam and the return can exceed the receiver protection limit.
Closer than about 660 feet, maintenance crews would be standing in microwave energy above workplace exposure limits.
What happened over Macon County
On the first of December in 2018, a supercell crossed central Illinois during an outbreak that produced 29 tornadoes statewide.
One of those tornadoes came down inside a wind farm in northern Macon County. The office reported that the contaminated data obscured the rotation.
It happened again on the third of January in 2023, in a rare winter outbreak that put six tornadoes on the ground and two of them in the Maroa area.
The warning coordination meteorologist at the Lincoln office said plainly that the data from the lowest level of the storm was contaminated.
A colleague there put the geography on it. The disruption is worst where the machines sit close to the antenna, in Macon, Logan and DeWitt counties.
The same contamination distorts rainfall estimates. The federal office has logged 4 inches or more of estimated rain falling on ground where none fell at all.
What this is not
Both warnings went out. In the January case the tornado warning was issued more than twenty minutes ahead of the tornado.
Forecasters climbed to higher elevation angles, read the storm above the contaminated layer, and leaned on trained spotters watching from the ground.
That is the honest version. This degrades a tool rather than switching it off, and it costs seconds and certainty rather than warnings.
It is also not the same problem as a turbine field putting phantom rain on a screen, which is annoying, while a hidden couplet is a different order of failure.
And the machines really do change the air above them, which is a separate finding again, with one Oklahoma farm pushing up a river of wind into the atmosphere.
Those three things get collapsed into one argument constantly, usually by people who want a conclusion out of them.
The part with no rule behind it
The federal radar office maps four zones around every antenna, from a no build ring outward through mitigation, consultation and notification bands.
Developers submit proposals into that process and the office negotiates. Projects get moved, lowered or reshaped, and some operators sign agreements to slow their machines during significant weather.
All of it is voluntary and advisory. The office can analyze a site and recommend against it, and nothing obliges anyone to listen.
Which leaves the fix where it has landed in central Illinois, on people. The weather service there is recruiting more trained spotters in the affected counties and running free classes to do it.
That is a real answer and a strange one, a volunteer with a phone standing in for a national radar network in the exact places the network cannot see.
The turbines are not going anywhere and a working group has been grinding at mitigation for years, so meanwhile the gap gets covered by whoever is willing to stand outside and look up.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
