Let’s Keep Our Most Precious Land Uses Out of the 1,000-Year Floodplain
I am writing this as a father and as a river engineer. As a father of three daughters, my heart goes out to the families impacted by the Texas flood disaster, and I pray for them as well as the first responders and local officials. As a river engineer, after viewing this video clip showing the arrival of a flash flood wave on the Guadalupe River near Comfort Texas, I wanted to know more.
The video was taken at the Lane Valley Road Bridge about six miles upstream from the USGS stream gage at Comfort, TX (08167000). Figure 1 below shows the change in flood stage and discharge from July 2-7 at this stream gage. On July 4th the river stage rose 32.5-feet (10-meters) in 2-hours to a peak flood discharge of 177,000 cubic feet per second (cfs)…that’s incredible!
But historically, what magnitude of peak flood discharges have occurred in this region? Research by Asquith and Slade (1995) explored the extreme flood potential for Texas rivers by evaluating peak discharges from 832 sites in natural basins in Texas for over 100 years. They developed extreme peak flood discharge envelope curves for hydrologic regions in Texas by plotting peak discharges and associated contributing drainage areas (see Figure 2 below). This curve defines the upper limit of past peak discharges for a given drainage area. The stream gage on the Guadalupe River at Comfort, TX has a contributing drainage area of 839 square miles and in this hydrologic region the envelope curve equation presents a peak discharge of about 677,000 cfs. In other words, there is the potential for a flood to occur at this location on the Guadalupe River nearly four times higher than the peak flow from the recent July 4th flood event.
These peak flood events seem massive—and they are—but when viewed in the context of floods that have occurred at this location over the past 125 years, other stories emerge. For instance, Figure 3 below shows annual peak streamflows at this stream gage plotted on a water year basis (Oct to Sept) with their relationship to statistical Annual Exceedance Probability (AEP) flood events. The highest annual peak discharges have consistently occurred in the June to August timeframe — during the summer months — and the July 2025 flood event was the third highest peak discharge at this stream gage since 1900 and 1978, and it was less than a 1% AEP event (100-year flood event). However, in many places along the Guadalupe River, the flooding was observed to exceed the mapped 1% AEP FEMA floodplains (Inal, Charalambous, and Vinick 2025). Having mapped river and coastal floodplains for FEMA flood insurance studies, I realize that these flood hazard boundaries represent a “snapshot in time” and can often be invalidated quickly as floods continue over time.
The 677,000 cfs peak discharge from the peak flood discharge envelope curve described above is “off the chart” below and assigning an AEP to this event by extrapolating the AEPs shown on Figure 3 is highly uncertain. A recommended optimal limit of credible extrapolation for flood discharges estimated from regional streamflow data (such as the extreme peak flood discharge envelope curves) is a 0.1% AEP event or a 1,000 year flood (Utah State University and U.S. Bureau of Reclamation 1999) and based on an extrapolation of the 4.0% to 0.2% AEP peak discharges this is estimated to be 450,000 cfs. But can this extreme flood ever happen? Statistically, there is an 8% chance that this 1,000-year flood could occur sometime over the 78-year life expectancy of a person in the U.S. (CDC 2025) (National Weather Service 2025).
From Data to Decisions
Now that we have looked at the data, what are some recommendations? Early warning systems? Yes, just do it. Here in Oregon, we have sirens to warn coastal residents in populated areas of potential tsunamis. Also, a recommendation I made back in 2006 is pertinent here. I was involved in the first evaluation of the FEMA National Flood Insurance Program and authored Chapter 11 in the study report “Assessing the Adequacy of the National Flood Insurance Program’s 1 Percent Flood Standard” (Water Policy Collaborative 2006). One observation I made in Section 11.4.1 was that our use of a national uniform 1% AEP flood event “ignores the significant regional variation of floodplain…processes” and “additional considerations should be given to the climatic and geographic characteristics of a flood source” and “a higher standard may be suitable where intense…precipitation events can occur”. I used a graphic showing a plot of peak discharge versus drainage area for the entire U.S. derived from 22,063 streamflow stations, similar to the peak discharge envelope curve for Texas described above, to encourage the use of data more conservative than the 1% AEP flood standard.
So, what should we do? Let’s learn from the past and move forward tailoring our flood hazard mapping to better manage flood risk to people and property. And for our most precious land uses — such as summer camp bunk houses — let’s keep them out of the 1,000-year floodplain, knowing that floods even greater than this one could potentially happen in one’s lifetime.
Kevin Coulton, PE, CFM is a consulting engineer, writer, and photographer living in Portland, Oregon. He can be reached via email at kevin.coulton@gmail.com.
