how does that relate to Popper?

When a weatherman's prediction is falsified, the model itself is not disproven. The fact that the practitioners of that discipline stick with it even when a prediction is falsified starts to look like the pseudoscience side of Popper's falsifiability criterion.