Last month, the Meramec River forced thousands to evacuate their homes in a hurry. Flooding wiped out numerous residences and businesses along the Meramec and its many narrow tributaries, that on most days are nearly dry. Places like Union, Eureka, Valley Park, Fenton, and Arnold (all in east central Missouri) saw record crests between Christmas, 2015, and New Year’s Day 2016.
Dozens of major roads in the St. Louis area were closed for weeks. Even I-70, I-44, and I-55 were closed for days. Up to then, Interstate 55 had never been closed at the Meramec due to flooding; on the 30th and 31st, it was closed in both directions. As a result, the only north-south road between St. Louis and Jefferson counties to escape the flooding (Telegraph Road, or Missouri Route 231) was backed up for miles on both days. I saw Telegraph backed up for over three miles at about 6:00 p.m. on New Year’s Eve; people heading casually down to the river crossing to marvel at the flooding were making better progress than southbound drivers.
The primary cause was a record amount of rainfall in December – over 11.5” in the St. Louis area during the entire month , and just over 9.0” from December 26-28. The speed with which the Meramec and other rivers, like the Mississippi and Missouri, rose and spilled over their banks was unprecedented. Many flood victims had mere hours, and even less than an hour in some cases, to evacuate with only the clothes they were wearing.
The Meramec frequently floods in the spring; fall and winter flooding is much less common. The last winter flood of this magnitude on the Meramec occurred 33 years ago. Of the top 10 Meramec crests at Arnold, MO, all but two were spring floods. Of the top 20 crests, only two (1982 and 2015) occurred in the month of December. They are two of the top three.
(Note: The 2015 flood was particularly disheartening for St. Louisans, in general, because the lower Meramec River crested between Christmas and New Year’s Day. Fortunately, we have a wonderful community, in which people automatically help those in distress, but there is still lots of work to be done.)
Other causes of the 2015-2016 flooding in Missouri include commercial and residential development in flood plains. The St. Louis area has seen an enormous amount of flood plain development since the last major flood, in 1993. Storm runoff from developed areas has increased significantly in the last quarter century and it is becoming increasingly difficult and expensive to control.
Most notable is the commercial development of the Chesterfield Valley not long after the ’93 flood had passed, which is symptomatic of area development. What had been fertile bottomland along the Missouri River was “protected” up to 1993 by a levee designed to contain a 100-year flood. A hundred-year flood gets its name from the 1-in-100 (or 1%) likelihood that a flood of such magnitude will happen in any year, not because (as some people believe) a flood like that will not happen for another 100 years.
After the ’93 flood, much of the Monarch Levee was raised to an estimated 500-year level, and very large scale retail development was allowed to begin. Repair and build-up of the Monarch Levee (officially known as the Monarch-Chesterfield Levee) was mainly a local effort.
Because of the fragmentation of government in the St. Louis area – for example, besides the county government of St. Louis County, there are over 90 self-governing municipalities within the county – flood control measures like dams, dkies, and levees are not planned, budgeted, enacted, or maintained with the entire area in mind.
Flood prevention is dealt with on a local level, addressed with local “solutions” that don’t account for the effect of those solutions on communities upstream and down. Short-term actions target specific problems; they are corrections or fixes, but not corrective actions as they’re known within an ISO 9001 framework.
We rely on outmoded flood models and flood policy. We still rely on the past to tell us what’s going to happen, though past performance is not a valid predictor of future performance. Statisticians deal in likelihood, not certainty. Predictive analytics, while not perfect, would be a considerable improvement over traditional data analysis techniques. Time and cost are limiting factors in the adoption of predictive analytics, however.
Additionally, we do not account for the potential influence of climate change, earthquakes, etc., on floods. Modern risk management says this is wrong: We have to include climate change rather than ignore the possibility. Many fatal decisions are prefaced by someone in authority saying “That’s not important” or “It’ll never happen.” We need to use risk analysis tools like failure mode effects analysis, or FMEA, to get a better handle on risks we face. We also need to continually reevaluate our assessments and our management systems. Nothing is perfect for long.
Most importantly, we need to think of risk management – whether it’s in the context of flooding or any other threat – in strategic terms. Cities and districts cannot be allowed to implement whatever form of risk management they choose without concern for problems they can create for others. Absolutely nothing happens in a vacuum.
We have to take a utilitarian approach to risk management (i.e., the greatest good for the greatest number). We have to be mindful of risks as well as opportunities, and in more than the immediate sense. We have to work across boundaries – Valley Park with Eureka; Fenton with Arnold; St. Louis county with Jefferson, St. Charles, and Franklin counties; Missouri with Illinois – to effect meaningful, long-lasting change.
Risk management is an ongoing, evolving activity. We cannot afford to be divided on the subject.
 Rainfall is measured by the National Weather Service at Lambert International Airport.
 See http://water.weather.gov/ahps2/hydrograph.php?gage=arnm7&wfo=lsx for Meramec River flood data.