This year's hurricane season in the Atlantic Ocean has so far seen just two named storms: Andrea and Barry. But a much bigger tempest has been brewing in Miami, Florida, for several months — let's call it Hurricane Bill.

In January, Bill Proenza took over as director of the US National Hurricane Center, the forecast facility in Miami, Florida, that aims to keep the US public safe from storms. At best it was a lateral career move from his previous post at the National Weather Service into a position that Proenza had not applied to fill.

Within weeks of assuming the centre's directorship, Proenza landed himself in hot water (see Nature 447, 514–515; doi:10.1038/447514a 2007). Among other things, he criticized how much money the centre's parent agency, the National Oceanic and Atmospheric Administration (NOAA), was spending on anniversary celebrations. More pointedly, he publicly bemoaned the lack of a detailed plan to replace the QuikSCAT satellite, which among its many jobs provides data on ocean winds to the team that forecasts hurricanes in the Atlantic.

This criticism did not sit well with his bosses or his employees, many of whom eventually called for his ousting. On 9 July, in the midst of a special assessment of his performance, Proenza was placed on leave.

Last week, he got his chance to tell his side of the story in Washington DC, to the House Committee on Science and Technology. Democrats on the committee pressed the question of whether Proenza had been sidelined because of his whistleblower activities on QuikSCAT. Meanwhile, Republicans griped about the committee spending its time investigating what they dismiss as a routine personnel matter.

Nick Lampson (Democrat, Texas), who chaired the hearing, got at least one thing right. “The only storms the centre should be dealing with are those that form out in the ocean,” he said. At the hearing, both sides acquitted themselves well: Proenza delivered an impassioned defence of his leadership, and Conrad Lautenbacher, NOAA's administrator, gave a lengthy and reasonably convincing explanation of why the agency felt it had to remove Proenza from his position.

The Proenza affair is not something that hurricane researchers and forecasters really need at this point, as the storm season begins to gear up. Government agencies are still reeling from their failure to cope with Hurricane Katrina in 2005, and scientists are dealing with the fall-out from their very public spats over the possible link between hurricanes and global warming (see Nature 441, 564–566; 2006).

So NOAA needs to show coherent and firm management. First, it should find a relatively quiet spot to assign Proenza to — outside of the hurricane centre, where employee resentment is apparently too high for him to continue to function as an effective leader.

Spats among staff should not distract from the larger task of focusing Earth-observation priorities.

Lautenbacher is aware of the need for NOAA to raise its public profile and assert itself as a powerful scientific agency, rather than just a backwater of the Department of Commerce, of which it is part. But it must make sure that its public-relations efforts don't get in the way of its scientific work. The very existence of a multimillion-dollar anniversary celebration is a cause for concern. And NOAA scientists have also been unhappy in recent months about management decrees suggesting, for example, that they improve the agency's branding by substituting 'NOAA' for 'National' in the names of centres such as the National Weather Service and the National Hurricane Center. Both of these outfits have distinguished histories and identities of their own, and NOAA needs to find ways of asserting itself and its mission in the public eye without diminishing them.

More substantively, the agency needs to address gaps in its satellite systems. Proenza is only the latest to highlight these. The National Academies had already done so, most recently in January, when it set out a national strategy of Earth-observing missions for the federal government. That plan would include an ocean-winds mission to do much the same job as QuikSCAT. An area of further concern was, the academies said, the level of coordination between NASA, which has customarily developed research satellites, and NOAA, which takes them into operational use.

NOAA needs to make sure that spats among staff at its hurricane centre do not distract from the larger task of focusing Earth-observation priorities in the coming decades. Researchers have already clarified what they need. Now it's time for the government to follow those recommendations and make sure the next generation of satellites is in place for the storms yet to come.