Rebuilding Houston: Considerations for communicating flood risk to the public

Featured image by Sgt. Daniel J. Martinez for the Air National Guard

By Andrés A. Salazar, Ph.D., P.E., D.WRE.

As the Houston region grapples to recover from Hurricane Harvey and prepare for future storms, one continuing point of contention is whether the floodplain maps illustrate a realistic level of flood risk.

A report prepared by the City of Houston indicates that during Harvey, a larger number of structures flooded outside of the 100-year and 500-year floodplains than within the floodplains. Because flood insurance is not mandatory for homes outside the 100-year floodplain, most of the impacted homeowners are now facing costly repairs without insurance benefits. Frustrated homeowners who thought they would be safe from flooding are now questioning the accuracy of these maps.

What hasn’t been transparent is that the flood risk mapping process includes unavoidable underlying uncertainty and limitations. Even if the maps are updated with new information, the underlying uncertainty will always exist. Rather than relying on the accuracy of the maps, the community must work toward a more complete and robust understanding of flood risk and the maps that portray this risk.

In order to understand flood risks, we need to go beyond a single map. This list outlines some ideas that should be taken into consideration when we communicate with the public about flood risk to help them make more informed decisions:

1.  There are an infinite number of combinations of how hard, how long, how much and where it rains that could cause flooding.

Rainfall varies widely from place to place and time to time. Several metrics can be used to describe a rainfall event, all of them having a wide range of possibility:

• Intensity: How hard is it raining? Rainfall can vary from an annoying drizzle to heavy downpours that make driving impossible and can’t drain fast enough, creating flooding.
• Duration: How long does it rain? Rain can come in the form of a 10-minute delay on your favorite outdoor activity, or the five-day event we experienced during Harvey.
• Spatial coverage: Where does it rain? There can be isolated or scattered showers and thunderstorms over a small area, or multi-county regional coverage.
• Rainfall depth:  How much does it rain? The amount of rain can be as little as precipitation trace (less than 0.1 inch), to as much as 60.6 inches of rainfall reported at the Nederland gage near Beaumont during Harvey.

Within these rainfall parameters there are an infinite number of combinations that could cause flooding. It could be a moderate event lasting one or two days, a heavy downpour in a small area for three hours, a not-so-intense rainfall event over the headwater of a watershed, a rainfall in a place that causes substantial overland flow, and so on.

2.  The definition of the 100-year storm needs to be understood.

So, given an infinite number of combinations that could lead to flooding, how do we choose a rainfall event to guide the design for flood protection?  Hydrologists and engineers typically use data from past rainfall to create models that illustrate how likely it is for rainfall events with certain durations and amounts of rain to occur. The most recognized rainfall event from these calculations is the so-called 100-year storm – but that is really only one possible event, and is not inclusive of all events that could cause flooding.

Let’s clarify the definition of the “100-year” storm.  News articles have defined this as a rainfall event so severe that it has a one percent chance of happening in any given year. However, this is misleading. The more accurate definition is: “The 100-year storm is a rainfall event that has one percent chance of being equaled or exceeded in any given year.”

The added words “equaled or exceeded” are important because it means it is possible that amount of rainfall will be surpassed and could be exceeded by a lot. This also means that areas outside the 100-year floodplain boundaries may be subject to flooding.

3.  The 100-year floodplain was adopted primarily for insurance purposes.

The 100-year storm was defined as a national standard to delineate floodplains for insurance purposes. But it could really be a storm with another frequency. We could delineate floodplains for the 50-year, the 200-year or the 500-year storm and adopt any of these other floodplains for regulatory and insurance purposes. Somewhere we need to draw the line. For that reason, I propose that we start using the term “regulatory floodplain” instead of “100-year floodplain.”

4.  Floodplain maps do not reflect all sources of flooding.

Floodplain maps reflect inundation levels during the 100-year storm, but they only take into consideration riverine flooding in studied streams. Not all streams are included in the floodplain studies and therefore, there are areas subject to flooding that are not shown in the maps. Also, other sources of flooding, such as overwhelmed local drainage systems or overland flow, are not always illustrated.

The floodplain maps are not necessarily wrong. They just provide a limited source of information rather than the total picture of flood risk. A flood risk assessment for a property should consider not only its relative location to the regulatory floodplain, but also flow patterns around the property, capacity of drainage systems, topography, and history of past flooding, among other factors.

5.  Flood risk management is not only an engineering issue.  It involves public policy and setting an acceptable level of risk for a community, which is a subjective matter.

Imagine that you are part of the design team of a new neighborhood with 1,000 homes and need to protect the subdivision from flooding. It is not possible or practical to protect for all events that could cause flooding. It is not possible either to model or account for all rainfall events that could come in the future. Funds are most likely limited. Thus, some judgement is required to achieve a balance between actual flood risk and funding availability.

Consider the following questions:

• What storm will you consider to design a flood protection system?
• How much will you be willing to pay to protect against that storm?
• Can you accept the fact that the storm that you select above may be exceeded in the future?
• What parameters would you use to define flooding?  Would it be water in the streets, the number of homes potentially flooded, the amount of property damage and repair cost, or something else?
• Would you rather spend money in maintaining a flood protection facility or repairing homes? What if the cost of building and maintenance of the infrastructure is much higher than the cost or repairs?

If you ask these questions to twenty different people from planners, engineers, city officials, and homeowners, you will probably get twenty different answers. Economic analysis could be helpful in assessing monetary damages versus the cost of flood protection. But there is always an emotional and subjective element. We can’t put dollars to the distress in someone’s life when it comes to recovery from a flood.

Moving forward

It is imperative that we expand our definition of flood risk. Some recommendations as we recover from Harvey:

• Accept that rainfall is unpredictable and uncertain. This is a compelling reason to buy flood insurance regardless of where you live. Remember that flood policy does not provide protection, it only defines the minimums that the community has agreed are acceptable. There will always be a storm that exceeds the policy. Some thought needs to be invested in how to deal with that reality.
• Engage in discussion with your peers to gather opinions and more information. Remember this is not only a technical issue but also a public policy issue. The more information and opinions you can obtain from others, the better your judgment.
• Stop designing to and preparing for past storms only. Future storms will be different (as Harvey has shown us). We need to build for uncertainty in the future and be ready to make adjustments along the way.
• Regularly update flood risk maps to reflect changing conditions and ensure decisions are being made based on the best available data. Some communities around the Houston-Galveston region are using flood maps that are over 10 years old, including Galveston, Friendswood, and Texas City.

Most importantly, those of us who are communicating with the public about flood risks should put transparency at the forefront of our conversations, and should urge FEMA and other public-facing entities to do the same. It is our responsibility to help the region’s residents make informed decisions by being open and direct about the limitations and risks of our current flood mapping processes.

---

 
Andres Salazar, Ph.D. PE, D.WRE. serves as the Director of Water Resources Engineering for Walter P Moore and Associates. His areas of experience and expertise include water resources planning, flood risk management, hydraulic modeling, and water supply design.