I love this report in the Yomiuri:

South Korea's Yonhap News Agency announced on 29 March that there is a possibility that the launch of North Korea's long-range ballistic missile under the guise of an "artificial satellite" will take place after 6 April due to weather conditions.

North Korea has announced that to international organizations that the launch will take place between 4 and 8 April, but according to the Yonhap wire service, The [Republic of] Korea Meteorological Agency has forecast that, at the launch base in Musudanri, North Hambyong Province, weather conditions will be "overcast beginning 3 April, with rain or snow falling on the afternoon of 4 April, and heavy cloud cover on 5 April also."

However, ROK forecasts have a bad reputation with citizens as "often inaccurate."

Oh. All right, then.

Another Yomiuri article, this time posted to the English site, says that intercepting the missile could be difficult for Japan because, of course, no one knows exactly where it will go. This handy diagram is appended:






If you're having a hard time reading that, the red lines represent paths in which the rocket falls on land in Japan--the solid line if it's the first booster rocket to separate, the dotted line if there's just not enough thrust off the launchpad and the whole thing flops.