Interesting thing that popped up on social media today...
https://www.kark.com/weather/weather-headlines/ast-blog-radar-beams-towards-i-40/
Due to a temperature inversion in the atmosphere, a weather radar beam in Little Rock was able to pick up traffic in eastern Arkansas on I-40. This is apparently a pretty rare (but really cool) occurrence.
So an interesting thing about that, the Nexrad software removes obvious crap data most of the time from the base reflectivity. However, this map was the radial velocity one, where it leaves in the movement of large objects the radar pings, just in case those objects are real, like debris from a tornado.
Nexrad base reflectivity used to do stuff like this all the time, where inversions bend the beam back down at a distance, and things like mountains look like stationary severe thunderstorms. They still do, the software just removes it out, because really there can't be any real weather-made object falling from the sky that is bigger than the biggest possible hailstones the clouds can muster.
There is actually a formula for the degree of bend of the radar beam in my satellite and radar meteorology textbook from college, where temperature and dew point are the two primary variables that go into it.