Understanding Uncertainty
in Floodplain
Mapping Procedures

<< Back to Main Page


Nick IrzaBy Nick Irza

Recent extreme flooding events have highlighted the limitations of what is currently known and expressed on nationwide FEMA floodplain maps. These maps are used to determine building codes, site development, set flood insurance rates, and are the primary source of information about residential flood risk, but, like most engineering estimates, include uncertainty that must be acknowledged to fully understand flood risk. The danger of flood maps is that they can create a false sense of security when they are accepted as the “hard and fast” rule instead of a guideline.


True flood protection requires increased awareness about flood risk and transparency about flood risk uncertainty in our maps.


Uncertainty in Studied Areas

Accurately mapping the floodplain is an engineering challenge which, even with the inclusion of advanced statistical and computational models, includes uncertainty and data bias. For example, we are frequently asked to extrapolate a 100- or 500-year rainfall from what is typically only a few decades’ worth of observed data. This, coupled with a myriad of other variables that can create a “perfect flood,” limit accurate floodplain prediction. Engineers do their best to overcome these obstacles, but floodplain boundaries are often much more uncertain than the seemingly tidy lines often depicted on maps.

Addressing such uncertainty requires embracing it as an actuality in floodplain mapping and flood risk communication. Evidence of poor understanding of flood risk has become more evident in recent years, and there has been discussion of moving away from quantitative descriptions of risk on floodplain maps (e.g., 100-year, 500-year) and toward more qualitative descriptions (e.g., low, moderate, or high risk). It is also important to note that the current shaded Zone X (“moderate flood hazard” including 500-year floodplain and shallow ponding areas) floodplain boundary is not a hard-stop for flooding. Though not currently delineated, if included, less probable floods would show that residual flood risk extends well beyond currently mapped boundaries.

Flooding Outside of Studied Areas

In addition to uncertainty within the mapped floodplain boundaries, recent storm events have brought attention to street flooding issues outside of delineated flood zones. Floodplains are currently mapped in areas that are considered most susceptible to flooding: regions adjacent to major rivers and streams or coastal zones that are vulnerable to storm surge. Street flooding and localized drainage issues, the primary sources of flooding outside of mapped areas, are not typically taken into consideration.

Widespread mapping of street flooding is not always economically feasible with current methods. However, future delineation of areas with historical local drainage issues or street flooding, and more detailed mapping of locations where the problem is known to be particularly severe may solutions to improve understanding of flood risk. Additionally, areas affected by dams/reservoirs (including the maximum pool extent and possible flooding from a dam breach or uncontrolled releases), may need to be included on future floodplain maps, as the potential devastation became overwhelmingly apparent during Hurricane Harvey — a Category 4 storm which devastated the Texas coast.

Aging Floodplain Maps

The previous issues affect even the most recent floodplain maps, but the time period in which floodplain maps were developed has its own impact on accuracy and uncertainty. Like any engineering study, creating a floodplain map requires managing the tradeoff between project cost and level of detail. Many older floodplain maps did not have the opportunity to leverage the latest technology or longer periods of observed data, either because it didn’t exist or was too cost prohibitive. Older floodplain maps also often exclude tributary floodplains and the 500-year floodplain. Regardless, until they are updated, these maps still influence development in some areas. Based on approximate methods, these maps will improve in accuracy as they are restudied, but they currently (more often than not) over- or under-estimate flood risk for surrounding communities.

While a level of uncertainty will always exist in our understanding of extreme event floodplains, addressing aging maps is the only the first step. Increased funding and improved mapping techniques go a long way toward ensuring that communities have the best available information. Ultimately, however, true flood protection requires increased awareness about flood risk and transparency about flood risk uncertainty in our maps.