River Sentry T100. Trust factor and preventing flood tragedies. (Part Seven)

From 2010 to 2020, Under the EU Copernicus program, Germany and Belgium worked together to build a cutting-edge Flood Early Warning System (FEWS) that would be collectively known as the European Flood Awareness System (EFAS). It was for good reason as floods have long plagued the region. The worst occurred in 1962 killing 362 people in Hamburg. EFAS was the pinnacle with tens of millions spent on the most modern sensors and software available.

On July 11^th, 2021, EFAS began the first of days of increasingly severe flood warnings with associate rain fall predictions. In the Valley of Ahr, these warnings were met with little apparent effect. On the afternoon of July 15^th, rain predictions foretold a river rise of 19 feet, an apparently acceptable risk for local leadership. Evacuations were not ordered. An additional flood warning was issued at 3PM. The local population seemed unfazed, perhaps desensitized by lesser floods in what was a common flood prone area. The multitude of warnings were simply not heeded.

Far upriver, at approximately 8pm; installed water sensors recorded a peak water height. In the upstream valleys, rain continued to fall reaching up to 7 inches in places. Downstream, the Valley of Ahr slowly fell asleep, fatefully unaware of the danger growing upriver.

At approximately 11PM local, the monster arrived in darkness. Predictably the power grid failed. Many were awakened to the unfamiliar roar enveloping their town center. Stone homes were swept of their foundations. Those retained trapped sleepers in their basements. A late evacuation attempt resulted in many drowning in their vehicles. At midnight, the water crested at 33.5 feet, nearly 15 feet above estimated levels. In the town of Beyenberg, the newly installed warning sirens failed leaving a local citizen to ring the medieval town bell in hopes of saving his townsfolk. Regrettably, 134 people lost their lives this night.

What went wrong? Why didn’t EFAS prevent this? “Climate change, unprecedented rainfall, extreme convective weather, a focus on larger tributaries, failed sensors, inoperative sirens and the local leadership’s failure to act in time were all provided as causal factors. Let’s discuss these.

A quick google search identifies a prior 24-hour rain fall record of approximately 13.9 inches recorded approximately 300 miles away in August 2002. The terms “climate change” and “unprecedented” are too often used as accountability off ramps. Prior rain records existed in proximity. The models were deficient. Just own it and stop making excuses.

Fragile sensors and a lack of focus on the smaller tributaries was a calculated guess with higher focus on more densely populated areas located on larger tributaries. EFAS guessed wronged and left a critical river without adequate detection. Own that too.

The lack of action by local officials is far more complicated and reflects a significant human factors issue. Why the indifference to the most severe warnings issued by EFAS? Did the local officials or population not trust the system? This behavior is indicative of warning fatigue. What was the Wolf Ratio in the Valley of Ahr? Had EFAS reduced itself to a never-ending nuisance? Did it generate common sense flood alerts every time it rained slowly grinding away any legitimacy or trust.

That data is not easily sourced as most post incident reporting is by government or government adjacent organizations perhaps ill positioned to criticize high visibility expensive projects. Regardless, the Valley of Ahr holds ominous warnings for the Kerr County Flood Commission. The same solution providers featured in EFAS are present in the FEWS plan being considered. Is this a good idea?

Texas is in tornado alley. This is Thunderdome. The 24-hour rain record for Germany is under 14 inches. The Texas record is over 43 inches, three times Germany. Texas is the land of mesoscale 50,000-foot monster storms featuring unpredictable wet microburst “rain bombs” inundating one valley while sparing the next. If the best modeling software couldn’t account for 7 inches of rain, how will it perform here? Will the false alarm ratio spike as it struggles with highly localized fast occurring flash floods? Will a high Wolf Ratio cause it to become ignored and discarded, thus setting the trajectory for another tragedy?

Most importantly, will the audience trust it?

 Sources:https://nhess.copernicus.org/articles/25/581/2025/