The recent flooding of Indian Creek that led to the damage and closing of Coach’s Bar and Grill has brought much attention from local papers and news stations. Several local media outlets have presented the flood as a “100 year flood event.” But what makes a flood a “100 year flood”?
Due to the nomenclature of the 100-year flood, many are often confused as to the likelihood of such an event happening, and, as a consequence, may under prepare or under insure with the illusion that a flood of this magnitude is a once-in-a-lifetime event. A 100-year flood event should statistically happen once every 100 years over a long time period, or have just a 1 percent chance of occurring within any given year. When people hear that a 100-year flood has occurred on a particular river, they may incorrectly assume that a flood of that magnitude should not happen again for roughly another 100 years. This is the same logical fallacy that would lead one to believe that flipping a coin and it coming up heads four times in a row is far more unlikely that flipping alternating heads and tails four times (statistically speaking, each are equally likely). Probabilities of this time scale should only be applied over long time horizons and cannot be accurately applied to short time periods. Therefore, back-to-back rare events may be more common than one would originally believe.
In addition to misinterpreting the definition a 100-year flood, there are a few other issues that can lead to error in attempting to predict such an event.
First, there is a common misconception that a 100-year storm event will always lead to a 100-year flood. While the classification of a storm event will affect the significance of a flood, many other factors contribute to a river’s flood levels, such as the current soil saturation or the coverage of a particular storm within a watershed. For example, a 100- year storm may only affect a small portion of a dry watershed area, thus only leading to a downstream river reaching 20 year or a 50 year flood levels. On the other hand, a less intense rain event on top of fully saturated soil may lead to more significant flooding.
Second, a 100-year flood event is only an approximation based on past data and is no guarantee of future outcomes. Because of the uncommon distribution of flood levels, there is no significant correlation with any common mathematical distribution to accurately estimate the actual probabilities of flood events. Whereas a human’s height or test scores can be normally distributed and graphically represented with a bell curve, the distribution of flood events cannot. Although the U.S. government has stated that flood events should be estimated using a Log Pearson Type III distribution, this distribution is by no means a perfect way to accurately predict flood events and must be used with caution.
Finally, it must be noted that the 100-year flood level is dynamic and can be reduced with good engineering. The probability of the 100-year flood level being reached within any given year is estimated at 1 percent. Because peak river levels are largely affected by increased runoff, an increase in urbanization and development may lead to an increase in the 100- year flood levels downstream, putting more homes and businesses at risk for flooding. However, through the use of detention basins or retention ponds and technologies like permeable pavements, runoff quantities can be reduced and subsequently lead to lower river levels during 100-year flood events.
In conclusion, extra caution must be taken when designing for low probability events, such as the frequency or level of a 100 year flood, regardless of how unlikely the event may seem at first glance. Through the intelligent use of detention and retention systems, the impact and loss of major flood events can be decreased for all homes and businesses downstream of urbanized areas.
— Derrick Price, Project Engineer