When a boat starts sinking, the prescribed reaction is to toss stuff overboard, beginning with those items deemed least important and moving progressively up the value ladder as the vessel continues to take on water.

The restructuring of the U.S. National Polar-orbiting Operational Environmental Satellite System (NPOESS) has followed a similar logic. The U.S. Air Force, the National Oceanic and Atmospheric Administration (NOAA) and NASA pared back the number of satellites and eliminated several instruments, presumably because they were peripheral to the primary NPOESS mission of gathering weather information for civil and military users.

The government had no choice but to restructure NPOESS, whose cost had grown from $8.4 billion to $13.8 billion, and given the life-and-death importance of operational weather satellites, it likely did not see a great deal of flexibility in choosing which sensors to keep and which ones to scrap. It is nonetheless very regrettable that key measurements for NASA’s climate-change research effort have been sacrificed to help save the program.

NPOESS was the cornerstone of NASA’s strategy for continuing its long-term study of human factors in global climate variability. The measurements this entails currently are carried out by NASA’s core Earth Observing System satellites: Terra, Aqua and Aura, which were launched in 1999, 2002 and 2004, respectively.

NASA had at one time planned to launch three rounds of satellites like Terra, Aqua and Aura to conduct a 15- to 20-year study of climate change. But that plan fell out of tune with projected budgets, so NASA scaled back several times, eventually settling for the initial Earth Observing System satellites and reliance on NPOESS for long-term data continuity.

To bridge any gap between its current satellites and NPOESS, NASA funded the NPOESS Preparatory Project, which originally was supposed to launch in 2005 to demonstrate the key operational weather sensors while also gathering research-quality data. Launch of the NPOESS Preparatory Project satellite has been delayed until 2009 under the restructuring plan.

During a House Science Committee hearing on the NPOESS restructuring June 8, NASA Administrator Mike Griffin said the agency has near-term replacements for some of the science sensors that were dropped. But he conceded that the agency currently has no plan, or funding, for continuing those measurements much beyond 2010.

Since that money is not going to magically appear anytime soon, NASA is going to have to get mighty creative if it wants to press ahead with climate-change research — which it absolutely should given the enormous implications of the phenomenon and the controversies surrounding it. The agency is going to have to rely more than ever on small satellites, low-cost sensors and international partnerships.

NASA has always been the junior partner in NPOESS and thus it stands to reason that its measurements would be among the first to go when the program was restructured. Unfortunately, the philosophy of the current White House — which seems to regard climate-change research as an anathema — doesn’t leave much hope for restoring the lost measurements in time to avoid a lengthy gap in the data record.

For NASA, the NPOESS experience is a warning against attempts to piggyback science measurements on operational spacecraft owned by other agencies. In the future, NASA should view such arrangements strictly as chance opportunities to fly low-cost experiments or technology demonstrations, and rely on its own satellites for high-priority scientific research where data continuity is imperative.

A broader conclusion in that same vein is that interagency space programs simply don’t work. For one thing, trying to meet the requirements of agencies with different mandates leads to big, complex spacecraft with too many sensors, which, as NPOESS showed, is a recipe for disaster because problems in one area can trip up the entire program. NPOESS also demonstrates how vulnerable interagency programs are to changing funding priorities and management breakdowns: it has become quite clear that nobody was in charge of this program.

This is perhaps the most unfortunate lesson of them all. Sweeping NASA’s research program under the NPOESS umbrella obviously was a mistake, but it is harder to see why NOAA and the Air Force, with so much overlap in their operational weather-monitoring requirements, were unable to get it together. True, the civil and military missions are not identical, but NPOESS is a failure of bureaucracy as much as anything else.

The experience does not bode well for another program that must transcend bureaucratic barriers to work: Space Radar, which is intended to serve both the military and intelligence community. The long-simmering tensions between the two camps over Space Radar requirements and funding suggest the program will be very difficult to execute, even if it gets strong congressional backing.

Federal budgets are limited, and all indications are that things will get tighter in the years ahead, and that makes it more critical than ever that U.S. agencies find ways to work better together to get the most of their space dollars. One of the enduring messages left by NPOESS may be the wrong one: duplication in space may be wasteful and expensive, but the cost of trying to break bureaucratic bad habits is even higher.