Thanks to an unbroken stream of atmospheric and oceanic measurements beamed to Earth from well-positioned satellites, meteorologists were able to predict Hurricane Sandy’s monstrous power and strange path (tropical storms typically turn right, not left) long before it hit the New Jersey coast.

It’s the kind of precision we might assume can only get better, as scientists fine-tune their instruments and forecast models. Instead, in a few years, weather predictions in the U.S. are in danger of becoming less accurate. Why? Because the federal government is unprepared, at least temporarily, to operate a full complement of satellites.

The gap could begin in October 2016, when a satellite put in orbit a year ago reaches the end of its expected five-year life. A replacement won’t be ready to launch until at least March 2017, and then it will take another year for its instruments to be checked out and ready to operate. That would leave a 17-month gap, during which three-to-five-day weather forecasts will be fuzzier.

To assess what such a loss of data could mean, the National Weather Service recalculated its forecast from early February 2010 of “Snowmageddon,” simulating a lack of information from polar-orbiting satellites. It came up with a prediction for a moderate to heavy storm, not the huge blizzard that swamped much of the East. The revised forecast underpredicted snowfall by more than 10 inches and missed the location of heaviest snow by 250 miles.

Longer Gap

What do we do? Well, there’s nothing that can be done in this particular case -- satellites simply take too long to plan. But we can find ways to minimize the possibility of such gaps emerging in the future. To see what’s needed, however, requires understanding how weather satellites are configured.

The U.S. operates two kinds of weather satellites: so-called geostationary ones, which stare constantly at one part of the country, and polar orbiters, which circle the Earth in a north-south path every hour and a half or so, viewing the whole planet twice over the course of a day. The geostationary satellites tell us what the weather is like at the moment, while the polar orbiters track systems as they travel west to east, allowing meteorologists to build forecasts as much as one to two weeks in advance, including blizzard and hurricane predictions.

As things stand today, U.S. forecasters use data from polar orbiters run by the National Oceanic and Atmospheric Administration, the Defense Department and a European weather satellite administration called Eumetsat -- a constellation that ensures they always have data no more than six hours old. As funding has gotten tighter in recent years, the U.S. government has twice reconfigured the management of its satellites, first combining the separate operations of the Defense Department and NOAA and then splitting them apart again.

But it still hasn’t come up with a national strategy for sustaining a robust satellite network to track Earth’s weather and climate. Instead, NOAA has done what it can to stumble forward, trimming the number of instruments included in its satellite plans as needed, as well as some data-processing capacity on the ground.

NASA’s Earth science program operates some climate-observing satellites, but in recent years this program has been significantly underfunded and, because of launch problems, has suffered the loss of two satellites -- one meant to monitor solar radiation and aerosols in the atmosphere and the other, carbon dioxide.

Satellite Management

A new report from the National Research Council argues that the White House Office of Science and Technology Policy should establish a person or office to oversee all weather and climate satellite operations and put together a long-range strategy to keep in orbit a full set of steadily improving Earth-observation satellites. We agree.

Support for such an effort could come from a National Climate Service -- an agency that NOAA needs but hasn’t been able to organize because of resistance in Congress.

One important thing this person or agency should soon figure out is whether it could be more economical to deploy smaller satellites more often, with each one containing only two or three instruments, than to send up a giant one carrying as many as eight or nine instruments every five years. More frequent launches of smaller satellites would allow scientists greater flexibility in planning and building instruments, and make it easier to recover from any problems or failures to launch. This strategy would require the development of smaller, less expensive satellite launch vehicles.

A backup satellite or two, ready for orbit, would come in handy right now. Better planning can make sure that, in decades ahead, forecasters always see the next Sandy coming.

To contact the Bloomberg View editorial board: view@bloomberg.net.