>>16729966Tropical geometry has an unfortunate name in that the word tropical has nothing to do with the subject -- it just so happened to be named after a Brazilian/Hungarian computer scientist who pioneered the field.
Having said that, the core idea is very simple: in ordinary (algebraic) geometry we're often interested in the curves that describe zeroes of polynomials in some set, let's say. These things are called varieties. Well, what happens if you were to look at the same thing, but replaced the ordinary operations of addition and multiplication in the ring of polynomials by the minimum and ordinary addition, respectively? The barbed wires one anon mentions arise naturally because of these minimum operations. Now, you'll have tropical analogs of varieties, and tropical analogs of theorems about varieties, etc.
Why do we care? Well, I can only speak for myself, but in my area of expertise -- economic theory -- tropical geometry appears in allocation problems in the presence of indivisibilities (goods that cannot be sliced up in arbitrarily fine ways, like cars - you can only have integer multiples of cars, but not half a car). It has also found fruitful applications in auction theory.
I hope this helps.