M: Well, you didn't expect it to be warm, did you？
W: Of course not, but I just heard the National Weather Service's prediction for the next 90 days. They said it's sup- posed to be much colder weather than usual.
M: Yeah, maybe. Personally, I think those long-range forecasts are useless.
W: Not as useless as you think.
M: Oh, come on, when you're talking about what's going to happen three months later, you might as well just pick a forecast out of our hat. W: Well, you are half-right. They aren't very good for the amount of rain. But they are a lot better for temperatures especially for this time of the year.
M: Really, so I should take them seriously about the cold but not count too much on a lot of extra snow？
W: Un-huh.
M: But what did you say about this time of the year？ Do some seasons really give them more problems in making forecasts？
W: Autumn is the worst, apparently weather patterns change so much then, just think how variable our weather has been the last three months.
M: Come to think of it. That's true. It probably would have been hard to predict all those changes back in the summer. You know, you're beginning to convince me there is more to forecast than I thought. How come you know so much about it？
W: I get my information from an expert. My sister's a meteorologist.
On what aspect of weather forecasting is the conversation about？
(20)

A. How much its accuracy has improved recently.
B. How reliable long-range forecasts are.
C. How difficult it takes to make a good forecast.
D. How the current forecast causes troubles.