Hindsight is very often 20-20, but sometimes foresight is, too. Mark Fischetti's article "Drowning New Orleans" in the October 2001 Scientific American all too accurately depicted the devastation that an inevitable strong hurricane would bring to that city, as have articles in many other publications since that time. Those predictions sprang from years of published scientific analyses. Any official who claims to have been surprised by the tragic events that unfolded in New Orleans after Katrina simply wasn't paying attention.

Realistically, even with this knowledge, a poor city in one of the nation's poorest states could never have done enough by itself to prevent Katrina's devastation or to cope with it afterward. "Catastrophic disasters are best defined in that they totally outstrip local and state resources, which is why the federal government needs to play a role," then FEMA director Joe Allbaugh told the New Orleans Times-Picayune in 2002. More's the pity, then, that in advance of the catastrophe the federal government severely underfunded the Army Corps of Engineers project to fortify New Orleans's levees. Rebuilding the city is sure to be considerably more expensive.

If federal priority-setters do not develop a healthier respect for the scale of damage that nature can and often will inflict, we will continue to be unprepared for another crisis now looming. "Preparing for a Pandemic," our special report starting on page 44, explains why scientists around the world are sounding an alarm and how governments and health authorities have begun belatedly to address the threat.

The leading candidate to cause the next flu pandemic is an avian virus dubbed H5N1 that has the potential to be even more lethal than the infamous 1918 pandemic strain. If it sickened a third of the population and killed just 5 percent of those, the death toll in the U.S. alone would exceed 10 times the pre-Katrina population of New Orleans. The U.S. pandemic plan that was due in October offers states much advice on how to prepare but little else with which to defend against an exceptionally infectious and virulent flu strain. Vaccines would take months to manufacture using archaic technology, and antiviral drugs that work against H5N1 would be in short supply. Both problems are fundamentally caused by a feeble market for these products in nonpandemic years.

That is why flu specialists have urged the U.S. government for at least a decade to prepare for a pandemic by taking the control of seasonal flu more seriously. Promoting yearly flu vaccinations for all age groups would be a good start, and mandatory vaccinations for schoolchildren are worth discussing: either measure could dramatically reduce seasonal transmission and thereby avert tens of thousands of flu-related hospitalizations and deaths annually. The policy would leverage a small government investment by giving vaccine makers incentive to improve their technology and expand manufacturing capacity. Similarly, if the U.S. would encourage doctors to use in-office diagnostic tests such as those that Japanese physicians commonly employ to diagnose flu on the spot and start treatments early, the rising demand for antiviral drugs would give pharmaceutical firms incentive to make more of them.

Flu season comes every year as reliably as hurricane season. If we shore up our defenses against both, we will be in a much stronger position when the "big ones" hit.