Wrong estimates of flood levels by US army corps

In 2013, Hannibal had a 50-year flood, a high-water event only expected once every 50 years. In 2014, it had another 50-year flood. Somehow, the river has reached its 10-year flood stage in Hannibal—which should happen about once a decade—in seven of the last eight years. And if the years seem to be passing with unusual speed, so do the centuries. Hannibal had a 200-year flood in 2008, considerably less than 200 years after an even larger deluge in 1993.

Evidence is mounting that Hannibal’s statistical anomalies have been caused not by glitches in the space-time continuum, but by a combination of floods getting worse and government estimates of flood risks being wrong. Those estimates of flood severity and frequency come from the long-troubled Army Corps of Engineers, and they determine how high to build levees and floodwalls, where to approve development in floodplains, and who needs to buy flood insurance.

But a new study of major Midwestern rivers in the Journal of Earth Science found that the Army Corps dramatically underestimates modern flood levels, leaving communities consistently underprepared for potential disasters at a time when flood protection is becoming a more urgent necessity. And climate change only seems to be a small part of the problem.

As the tenth anniversary of Hurricane Katrina this month refocuses attention on flood dangers in coastal communities—and on the foibles of the Army Corps, which played a central role in the drowning of New Orleans—the study by Washington University hydrogeologist Robert Criss suggests riverside communities are also much more vulnerable than they’ve been told. Criss calculated that official 100-year flood measurements are about four and a half feet too low in Hannibal, five and a half feet too low in St. Louis, and six feet too low in Omaha. After analyzing 150 years of data, he essentially concluded that Army Corps river managers have taken the advice of the legendary Hannibal native Mark Twain: “Get your facts first; then you can distort them as you please.”

 “They go through a bunch of convoluted razzmatazz, and they end up with absolutely crazy numbers,” Criss said in an interview. “They’re lying to people. They’re saying it’s safe to build when it clearly isn’t.”

 The Corps says that Criss erred by measuring water levels instead of water flows, the amount of water that passes a certain point in the river at a certain time. Criss said that’s bizarre, since high water levels are what create floods, and there’s such detailed historical data about water levels; trying to estimate flooding from flows requires much more complex models of how the river works. He said the official models, last updated about a decade ago, essentially exclude the possibility that floods could get worse over time, as if the same amount of rainfall would always create the same amount of flooding. And it must be said that Army Corps has a history of fuzzy math.

Source: http://www.politico.com/agenda/story/2015/08/are-americas-flood-estimates-dangerously-wrong-000203