Let’s say you’ve got a water molecule. It doesn’t have to be water, but it helps if it’s one we can easily picture:
You can imagine water vapor as an ensemble of many of these molecules flying and bouncing around in their container. This translational motion is not the only kind of motion they’re executing. They’re also rotating, and each impact with another molecule can change the speed and direction of that rotation. Not only are the molecules translating and rotating, they’re also vibrating. The relative positions of the two hydrogen and one oxygen atom are not fixed rigidly, but in practice behave more like they’re connected by little springs and so the vibrate with respect to each other.
Now this is a classical picture, and understood in quantum terms there’s a lot of differences and caveats to be aware of. Nonetheless, the true quantum model corresponds well in its broad outlines to this classical description I’ve outlined.
Now pretend we’ve got a photon of light, which in our picture we can treat as a tiny particle flying through this storm of water molecules crashing around. It runs into a molecule and in the process interacts with these rotations and vibrations – if you think of yourself as jumping on to a fast-moving merry-go-round it’s clear that you’ll gain speed if you land on the side rotating with your direction of motion and lose speed if you land on the side rotating against your direction of motion. Same sort of thing for the photon hitting the rotating molecule. A photon can’t change speed though, so its energy change is expressed in terms of a changing frequency. Quantum mechanically, it’s scattering off the rotational and vibrational states. A photon that loses energy is said to have undergone Stokes scattering; one that gains energy has undergone anti-Stokes scattering. These both fall under the general heading of Raman scattering.
These processes are fairly weak, generally only a tiny fraction of photons encountering an atom undergo one of these interactions. This is a difficulty for those scientists studying this behavior and it’s also a problem for the considerable and growing commercial applications of this scattering. One way around this is to use the properties of coherent light so that each scattering event emits light in phase, in the manner of a laser. It’s a third-order process requiring three separate lasers to produce the shifted light called CARS in the schematic quantum diagram below:
Why is it called CARS? Coherent Anti-Stokes Raman Spectroscopy. In a cute twist, the effect was first discovered in 1965 by physicists at Ford Motor Company. Yes, the auto manufacturer used to have a fundamental physics research division. To their credit, they weren’t the ones who named it. That came years later via other researchers.
We’re still discovering all sorts of interesting variations of this phenomenon. One of them is applying femtosecond laser pulses to produce the effect. This short-time-scale method was inevetably named “fast CARS”. It’s also possible to use femtosecond pulses to apply a combined technique that can simultaneously resolve the time and frequency behavior of CARS light. The name for this is “hybrid CARS”.
The possibilities practically write themselves. I have gently made fun of the ridiculous names in high-energy physics before, but I have to say I like the CARS puns. Me, I’m just hoping for a use in atmospheric science so we’ll have an excuse to put it in a plane and call it “flying CARS”.