The Harris County Flood Control District is currently holding public meetings link to present its SAFER Study—short for Solutions for Advancing Floodplain Evaluation and Resilience. At the same time, the Texas Legislature is convening a special session where flooding and flood infrastructure are among the policy issues under consideration. The convergence of local planning and statewide policymaking makes this an important moment for reflection.
As some readers may know, I am a Fellow of the Royal Meteorological Society. My background in meteorology informs my perspective, but so does my deep concern as a Houstonian who understands both the personal and financial toll of flooding. My overriding concern is simple: flood policy should be driven by good science, good data, and a commitment to minimizing harm in the pursuit of protection.
Too often, financial interests supplant the common good. Take, for example, the proliferation of clickbait weather pages on Facebook. If they were named the Dewey, Cheetum, and Howe Clickbait Weather Page, no one would take them seriously. Instead, they adopt credible-sounding names and often use long range meteorological modeling data to project an unwarranted air of authority.
The same phenomenon can occur with consulting groups. To be clear, I’m not suggesting these firms have the same questionable motives as internet grifters – they do not. However, they create a similar air of credibility, which in turn shields flawed assumptions or methodologies from legitimate scrutiny. Well-intentioned policy veers off course when the underlying data is flawed, incomplete, or improperly applied. Unfortunately, what I observed in the recent SAFER Study presentation suggests we may be headed in that direction.
In the sections that follow, I’ll examine three critical areas of concern. First, the flawed use of precipitation data, particularly how named tropical systems have distorted the calculation of flood return periods. Second, the fundamental under-acknowledgment of how localized sheet flow feeds into river and bayou flooding. Third, the lack of coordination among political jurisdictions, which, despite relying on many of the same consultants, continue to operate in silos without a unified regional flood strategy. These issues are not isolated; they compound each other and together risk undermining a cohesive solution for flood control.
Misused Precipitation Data and the Weight of Outliers
One of the foundational flaws in recent flood planning stems from how precipitation data is being applied – particularly when it comes to rare, high-impact tropical systems.
When voters approved Propositions A and B to fund flood control efforts, the planning and modeling behind those proposals did not exclude rainfall totals from extreme named storms. Not all tropical systems produce 100-year flood events. Ike, Nicholas, and last year’s Beryl did not. However, Tropical Storm Allison, Hurricane Harvey, and Tropical Storm Imelda each exceeded standard thresholds: Allison exceeded the 100-year return period and approached 1-in-200-year levels in some locations; Harvey surpassed 500-year rainfall totals in many areas and even approached the 1,000-year mark; Imelda, while less publicized, approached 500-year intensity in some areas.
Including these outliers in baseline calculations distorts the final calculations. It raises the statistical average for what qualifies as a “100-year flood” and, by extension, expands the footprint of mapped floodplains. That affects everything: insurance rates, building regulations, infrastructure planning, and property values. In short, the models are overestimating risk in many areas – not because they’re broken, but because they’re anchored to extraordinary events that should be treated as anomalies.
Since the passage of those bond measures, HCFCD has transitioned to using Atlas 14 for precipitation data. This shift represents a good faith effort to improve accuracy by using the latest available data – but it unintentionally exacerbated the problem. Because Atlas 14 includes more data it better captures the statistical footprint of those rare events. Rather than balancing the models, it effectively locks in the elevated values – anomalies even among tropical system impacts in Harris County.
We absolutely need effective flood control measures – Harris County has seen too much devastation to argue otherwise. But these measures must be implemented at the lowest cost possible, both financially and in terms of collateral damage. Overinflated models lead to overengineered responses, which in turn lead to expensive and disruptive projects. Purchasing large swaths of land, displacing residents, or invoking eminent domain, while sometimes necessary, should never be the first option – especially when flawed data is driving the decision-making.
On the bright side, this is an easily addressed issue. Simply exclude Allison, Harvey, and Imelda and recalculate the 100-year floodplain. Harris County cannot regulate FEMA flood maps or insurance premiums, but it can control its own mapping and use a recalculated, more realistic floodplain for county-funded projects. We owe it to taxpayers, homeowners, and future generations to get this right.
Sheet Flow and River Flooding Are Linked
Flooding in Harris County generally comes in two forms: river flooding, which includes bayous, and sheet flow flooding, more commonly (and misleadingly) referred to as “street flooding.” While often discussed separately, these two forms of flooding are deeply connected. Any serious mitigation strategy must address both not in isolation, but as they are – a unified hydrologic system.
Here’s what happens when it rains: water hits the ground and is either absorbed into the soil or runs off if the precipitation rate exceeds the land’s absorption capacity. That excess water becomes sheet flow, which enters storm drains and ditches, eventually feeding into our rivers and bayous.
The amount of runoff is calculated by the formula Q = CiA, where:
- Q = runoff rate
- C = runoff coefficient (the only controllable variable)
- i = precipitation intensity (often called precipitation rate)
- A = area receiving precipitation
The C value is the only value that is influenced by urban development. The value is often visualized through an S-curve, which reflects how quickly runoff moves into stormwater systems and, in turn, rivers. S flow determines how fast water reaches those systems, and if the systems aren’t engineered with sufficient acceptance and transport capacity, the result is urban flooding.
It’s not difficult to diagnose where the problem lies. If storm drains are full and water is bubbling up or backflow is visible from inlets or manholes, it’s a transport capacity issue—the system can’t move water out fast enough. If no such signs appear, the issue is more likely acceptance capacity—the system can’t take in water fast enough to begin with.
Effective flood control must address both. Flooded intersections serve as functional retention ponds, slowing water’s entry into the drainage network. The current “acceptable time window” for flooded intersections simply picks winners and losers. The current plan is willing to accept urban flooding and the costs and impositions associated with the flooding. That’s not an acceptable solution. However, addressing street flooding without considering how it accelerates flow into bayous impacts down flow conditions.
These systems are tied together, which brings us to a deeper, self-inflicted challenge: governance.
Jurisdictional Fragmentation and the Illusion of Coordination
The biggest takeaway from the SAFER Study meeting wasn’t a technical flaw; it was the glaring lack of coordination between political subdivisions when it comes to formulating a comprehensive flood mitigation plan.
That’s not to say coordination doesn’t happen. It does, but it’s inconsistent and fragmented. Cities, utility districts, and regional agencies often plan within their own silos, even though floodwaters don’t respect those boundaries.
This disjointed approach is often rationalized two ways: first, through political recognition of the jurisdictional limitations; and second, by saying, “we use the same consultants.” Neither excuse holds up.
Sharing consultants may create the illusion of coordination, but it doesn’t ensure that final plans are integrated or aligned. Each jurisdiction still defines its own scope, contracts separately, and executes independently—without a binding regional strategy. There may be meetings, but there is no master plan.
This problem is no longer academic. With $400 million in new federal funding recently released—alongside the existing county bond funds—we’ve entered the implementation phase. What we build now will either fix or cement the fragmentation.
The most disheartening aspect is that better coordination is possible now. County projects could be formulated with “bolt-on” compatibility to nearby urban initiatives. Urban projects could include hydrologic impact calculations and be temporarily delayed 30 days to allow the County to review and propose coordination strategies. This doesn’t require new laws. All it takes is a cooperative mind set.
Some of that mind set already exists. Commissioner Ramsey’s office has expressed interest in discussions around these proposals. Even if only Precinct 3 begins to include bolt-on opportunities in their projects, it demonstrates that a simple mindset shift can bring us closer to real regional coordination.
We will never take politics out of flood control. However, adjusting basic standard operating procedures can move us closer to what should be our shared goal: a comprehensive, efficient, and minimally disruptive flood control system that protects lives, property, and taxpayer dollars alike.