Flood waters surround buildings in Houston on Wednesday. Will Harvey finally prompt changes in flood management plans? (Jabin Botsford/The Washington Post)

The waters stretched as far as you could see, and then farther, in every direction. Rainfall totals were reported in feet, not the usual inches. Dozens of people died. Highways were submerged; thousands upon thousands of square miles of land were deluged. The damage totaled billions of dollars.

It happened 24 years ago, in the Great Flood of 1993 in the upper Midwest.

After that disaster, the Clinton administration directed an experienced federal interagency task force to report on the flood and its causes. That report, "Sharing the Challenge ," was prepared by Army Brig. Gen. Gerry E. Galloway and released in 1994. It made more than 100 recommendations for policy and program changes to address and reduce flood risks and improve the nation's floodplain management everywhere, not just in the area along the Mississippi River that had been underwater. The government found that many policies were encouraging — rather than discouraging — people to build homes and businesses in places with increasingly high risks of flooding by allowing new building in those areas, constructing insufficient flood-control projects that give residents a false sense of security and subsidizing redevelopment after disasters without mitigation. That often compounded the costs and problems caused by floods.

Ultimately, though, very little changed. The lessons of 1993 were largely ignored, especially in parts of the country that were most vulnerable to flooding — such as Houston. Experts and policymakers have known for a long time that we need to change the way we approach flood mitigation and prevention, but that hasn’t stopped the nation from making the same mistakes over and over. Now, as the federal government prepares to spend billions more cleaning up from catastrophic floods, we’re in danger of doing it again.

* * *

The Clinton administration's report seemed like it might change things at first. It suggested the government should offer voluntary buyouts to owners of buildings that flooded repeatedly, clearing the most at-risk land of businesses and residences and leaving it as open space that could be devoted to flood-tolerant uses such as parks, recreation areas and wetlands. Especially in states such as Missouri, Iowa and Illinois that had been hit hard by the 1993 disaster, governors supported this new approach. More than 10,000 buildings were bought so their owners could move outside floodplains. The federal government spent $121 million on this type of mitigation after the 1993 floods — acquiring land or elevating, relocating or flood-proofing buildings. That investment probably saved $600 million in disaster relief: The National Institute of Building Sciences estimates that each dollar spent on flood mitigation saves $5 in future flood damage.

Four years after the Great Flood commission issued its report, the two of us collaborated on a follow-up paper released by the National Wildlife Federation, "Higher Ground ." We looked at the nascent successes of the Midwest floodplain buyouts and relocations that took place after the 1993 floods. In Missouri, where there were major buyouts, one study showed that disaster relief after a 1995 flood cost 99 percent less than after the 1993 flood, even though 85 percent of the same area was affected. We also looked at the class of properties considered by the National Flood Insurance Program to be "repetitive loss properties" — places where the program had paid repeated claims within a 10-year span. We found that substantial benefits for property owners and taxpayers could be gleaned by simply removing damaged buildings, rather than repairing them only to see them flooded out again. We also concluded that many flood insurance policies were heavily subsidized and underestimated risk, leading to premiums that were far too low to make homeowners realize just how flood-prone their locations were.

Still, we also found that paying to relocate a relatively small number of people could significantly reduce the risk. About 2 percent of the flood insurance policies with repetitive losses were generating 40 percent of the claim payouts — many of them making expensive claims frequently. But these properties were seldom subject to risk mitigation (such as elevation or relocation), even after a flood caused damage worth more than 50 percent of their value, which is considered "substantially damaged" and is supposed to require mitigation. We also found that about 20 percent of these properties were located outside the designated "100 year" floodplains, areas where the chance of a flood in any given year is predicted to be 1 percent. The maps, in other words, were badly out of date, even then.

But owners often found few realistic options for getting out of harm’s way, even with some of the new programs that were established, because the government was slow to provide assistance and there seemed to be little real resolve for making difficult changes.

The city where we chose to release the report was Houston.

At the time, Houston ranked third (and surrounding Harris County fourth) in the nation in the number of repetitive-loss buildings, behind only Jefferson Parish, La., and New Orleans. It was too late to change policies in a way that would help New Orleans, which was already mostly built out — and mostly at or below sea level. But Houston and its adjacent areas, even with explosive growth and development underway, still had clear options to prevent future floods. They could conduct watershed planning and management. Officials could buy out or relocate the most endangered buildings, or put them on stilts, and could adopt higher building-elevation standards and stronger building codes. They could regulate the expansion of impervious areas in a way that would limit paving over pastures and wetlands to construct parking lots and endless subdivisions, and they could establish shoreline protections to reduce flood risks and costs going forward. This was the array of approaches needed to stay and hopefully reverse the growing costs of flooding. The nation’s floodplain organizations, at the same time, pleaded with city leaders to heed these warnings and act.

* * *

They didn't. Houston did some buyouts, but repetitive losses continued to mount as development pushed along mostly unfettered. Over the past week, many news reports have highlighted long-ignored studies chronicling the steady encroachment on Houston-area bayou floodplains, and the paving and building over of pastures and wetlands that used to act as sponges absorbing and slowing runoff. At least 4,000 residential and commercial structures have been built within the identified 100-year floodplain since 2010, and, according to one Texas A&M University study, 30 percent of Harris County's coastal prairie wetlands were paved over from 1992 to 2010.

Now we have yet another chance to get things right. Congress will return this coming week to a flood insurance program requiring reauthorization by Sept. 30 and owing $25 billion to the treasury. Lawmakers will have to approve billions in disaster assistance funds and an increase in the cap on federal debt. The sheer scale of the crisis might finally be enough to force politicians and policymakers to abandon what failed in the past. True, Harvey’s rain fell in such volumes that it might have defeated any flood management plan. But the damage could have been lessened. And the floods also underscored problems we’ve long known about.

Congress should consider forgiving the flood insurance program's current debt, most of which can be traced to storms such as Katrina, Rita and Wilma, which hit in 2005, before Congress gave FEMA better tools to manage financial risk for these types of events. Forgiving the debt owed by one part of the government to another is no different from paying for disaster relief without requiring offsetting budget cuts, which only die-hard fiscal conservatives advocate; clearing it off the books will mean old costs from previous storms don't burden the agency's ability to manage floods now.

After the most damaging flood in U.S. history, we also need to fund accurate flood mapping for the entire nation. We need better maps out ahead of development, so people know how to build and won't get caught with costly flood insurance premiums if maps are drawn after an area is developed. It would take an estimated $7.5 billion to map all the floodplains in the nation.

Congress should also reinstate the federal flood risk management standard President Trump recently revoked in an infrastructure executive order to assure a reasonable building standard for states and municipalities that use federal money to rebuild after disasters. This could be as simple as constructing most buildings one or two feet above the 100-year flood level (and two or three feet higher for critical facilities such as hospitals and police and fire stations). That way, taxpayers won't have to pay to rebuild these facilities as often. Houston already required putting new buildings one foot above the 100-year flood level. Studies of Harvey's rain may soon show that to be too low for the city', which may instead want to consider the three-foot standard Dallas has in place might be worth considering.

The government must clarify, either through law or regulation, that Department of Housing and Urban Development disaster relief funds can be used for voluntary building acquisitions, removals and relocations. We can’t keep paving over watersheds and filling them with parking lots and rooftops, taking away the ground’s capacity to absorb rainfall. We must make wise use of nature’s natural ability to store floodwaters to protect us.

Finally, leaders should immediately focus on improving and streamlining voluntary buyouts and other mitigation options to help residents move out of harm’s way within reasonable time frames, especially after disasters, allowing communities to reform their land use in response to nature’s realities.

The nation has not been particularly good at learning from past floods, which have occurred with discouraging frequency. This time, we need to actually fix some of the problems. If we don’t, Americans facing some new devastation in the future will be looking back at Harvey and wondering why we didn’t act now.


Read more from Outlook:

Poor Texans are going to suffer the most from Harvey, thanks to state politics

Hurricane Harvey threatens more than you think

How farmers convinced scientists to take climate change seriously

Follow our updates on Facebook and Twitter