Seen on the AP wires:
An aging weather satellite crucial to accurate predictions on the intensity and path of hurricanes could fail at any moment and plans to launch a replacement have been pushed back seven years to 2016.[...]
If the satellite faltered, experts estimate that the accuracy of two-day forecasts would suffer by 10 percent and three-day forecasts by 16 percent, which could translate into miles of coastline and the difference between a city being evacuated or not.[...]
QuikScat, launched in 1999 and designed to last two to three years, provides key data on wind speed and direction over the ocean. Weather aircraft and buoys can also obtain similar measurements near a storm, but they do not provide a constant flow of data as QuikScat does.
Last year, the satellite suffered a major setback — the failure of a transmitter used to send data to Earth about every 90 minutes. Now the satellite is limping along on a backup transmitter and has other problems.
The article mentions that an additional wrinkle in the situation is that the replacement satellite is allegedly less precise than the current one.
I suppose, in the interest of finding a silver lining in this hurricane prediction cloud, I could point out that increased evacuations should provide an additional disincentive to living in storm-prone areas, hopefully weakening demand for properties in those areas, and thereby easing pressure on property insurance rates due to the growth in cat risk.
Sadly, I suspect a more realistic outcome is that there will be an increased likelihood of folks becoming desensitized to storm evacuation orders, leading to increased casualties.