Weather forecasts have gotten quite a bit better since 1980 or so and are still improving. The DC snowstorm is a good example: the forecasts were sufficiently reliable for everyone to dedicate considerable resources to preparations well ahead of time. With hurricane Katrina there was pressure to order mandatory evacuations *3 days* before landfall. Neither would have been credible so far in advance 30 years ago.

*Climate* forecasting is still a rough area: by this I mean predicting a season's general trends (rainy/dry, cold/hot) ahead of time. Yet even here progress is seen: NWS correctly predicted last summer that it would be a wet winter in my area. Hurricane season forecasts don't seem to have gotten much better in the last decade, yet overall the field is a lot better than it used to be: 30 years ago you might have asked a meteorologist what dowsing rod he used in predicting a wet winter during a bone-dry summer.

My NWS office web page publishes a "Forecast Discussion" giving the basic reasoning behind a forecast and describing the situation in better detail than the rigid NWS format required for official forecasts. I generally skip the formal forecast entirely and go straight to the forecaster's commentary.

I guess my thesis here is that modern weather forecasts are reliable to the point of shaping preparations for activities. If forecasts show a front passing near a mountain maybe you shouldn't climb that day even if the airplane ticket is already paid, etc. Weather forecasts should be a basis for action and not merely passively collected as part of preparations.