For extreme events is there a better
A perfectly predicted 125 mm area having a position error
may be a terrible forecast. Or is it?
Imagine a non-hydrostatic model forecasting 125 mm (5 inches) or
more of rain (magenta).
It verifies as the light blue area