After May 7th and September 19th storms, reports started pouring in about homes that flooded that had not flooded during Harvey. How could that be? Harvey dumped more rain than any storm in history of the continental U.S. The short answer: When you look closely at the rainfall statistics, its comes down to totals versus intensities. Even though Harvey brought more rain, its peak intensity never came close to the other storms’. Those higher peaks can create street flooding miles from any river. And they did.
Those looking for a more detailed explanation can read below.
Let’s start by looking at three local rainfall gages on the West and East Forks. Then we’ll compare statistics.
Where To Find Statistics
Using the Harris County Flood Warning System, you can find historical rainfall data for any date or gage in the region. Here’s what I found for three local gages and three storms.
Comparison of Storm Totals, Peak Intensities and Durations
During Harvey, we received:
- 33.04 inches at the West Fork/US59, with a peak 2.36 inches in one hour.
- 27.44 inches at the East Fork/FM1485, with a peak 1.4 inches in one hour.
- East Fork/FM2090 – no data reported.
On May 7th, we received:
- 6.76 inches at the West Fork/US59 with a peak 3.64 inches in one hour
- 9.84 inches at the East Fork/FM1485 with a peak 3.16 inches in one hour
- 6.76 inches at the East Fork/FM2090 with a peak 1.6 inches in one hour
On September 19th, we received:
- 11.56 inches at the West Fork/US59 with a peak 4.56 inches in one hour
- 18.88 inches at the East Fork/FM1485, with a peak 9.4 inches in three hours.
- 19.68 inches at the East Fork/FM2090, with the peak of 3.24 inches in one hour.
Comparing Totals and Intensities
From this data, several things become immediately apparent.
- Harvey dropped the most rain, but had the lowest peak intensity.
- Compared to May 7th, the September 19th storm dropped roughly 2-3X more rain at each gage.
- Comparing gage locations, you can also see tremendous variability over just a few miles for any given storm. For instance, on September 19th, 63% more rain fell on New Caney than the Humble/Kingwood area.
Comparing All 3 Storms to Precipitation Frequency Estimates
The next question: How do these numbers compare to the hypothetical 100- and 500-year rainfall events? Here are the updated precipitation frequency estimates for this area from NOAA Atlas 14.
Take the totals and peak intensities from the gage information above. Then locate them in the table below. Here’s what you find.
Frequency Estimates for Harvey – August 25-30, 2017
US59 at West Fork during Harvey
- 500-year event based on 5-day total
- 2- to 5- year event based on peak hour.
FM1485 At East Fork during Harvey
- 200-year event based on 5-day total
- 1-year event based on peak hour.
FM2090 at East Fork during Harvey
- No data available
Frequency Estimates for May 7 Storm
US59 Gage on May 7
- 5-year storm based on 24-hour totals
- 25-year storm based on the 1-hour peak
Street and minor yard flooding in an event like this is normal. But it should not have entered your house if the house is above the 100-year flood plain. (Most homes are at least two feet above.)
If you’re near this gage and your home flooded on May 7th, it probably wasn’t because of the peak rainfall intensity or total. You should be looking for clogged or broken storm drains, fallen trees/sediment blocking ditches, or upstream development, like Woodridge Village, that overtaxed the capacity of drainage systems.
FM1485 Gage on May 7
- 25-year storm based on rainfall total
- 10-year storm based on peak intensity (1-hour rate).
Again, this should be well within the capacity of infrastructure to handle. If your home flooded, look for other causes like those above.
FM2090 Gage on May 7
- 5-year storm based on rainfall total
- 1-year storm based on peak intensity.
Infrastructure should have handled that easily. If your home flooded, look for one of the possible causes listed above.
Frequency Estimates for September 19 Storm
US59 Gage on September 19
- 25-year storm based on 24-hour total
- 100-year storm based on peak intensity.
Storm drains are designed to handle about 2 inches per hour, but when we got 4.56 inches between 9 and 10 a.m., we clearly pushed infrastructure capacity to the limit. It’s no accident that Jeff Miller’s security camera witnessed a huge surge of muddy water from Woodridge Village coming down his street at 10:10 a.m.
Most infrastructure, homes, and businesses are built to handle a hundred-year rain event. Even homes surrounded by 100-year flood plains are usually raised above them.
If you flooded in this event, it may have been because of the extreme rainfall or because your slab wasn’t elevated two feet above the 100 year flood plain like most building regs require.
Other causes may have factored in also. When there’s little margin for error, a partially clogged storm drain or ditch, or upstream development could have made the difference between flooding and not flooding.
A massive 268-acre, clear cut area with only 25% of the detention installed immediately upstream from you would quickly turn a marginal situation hopeless.
FM1485 Gage on September 19
- 100- to 200-year storm based on 24-hour totals.
- 100- to 200-year storm based on 3-hour peak intensity.
It easily surpassed the design capacity of infrastructure. Events like this make a good case for flood insurance if you don’t have it.
FM2090 Gage on September 19
- 200-year event based on 24 hour total
- 10-year event based on peak intensity.
Again, think flood insurance. You’re way past the design capacity of infrastructure.
River Flooding Vs. Street Flooding
Peak intensity usually affects streets first. Storm totals usually affect rivers later.
From the frequency estimates above we can see that Harvey was extreme in its totality, but did not reach the peak intensity that either the May 7th or September 19th storms did. In fact…
The highest one-hour total for the May 7th and September 19th storms at all three gages exceeded the highest one-hour measurement during Harvey.
The real story of Harvey: how much rain KEPT falling for days and days. And how much water was released from the Lake Conroe Dam at the peak of the storm. Harvey was not about street flooding. It was about river flooding.
River Flooding Versus Street Flooding
Rainfall totals, intensities and durations affect flooding differently.
Intense bursts of rain like we saw on September 19th create street flooding. Short bursts quickly exceed storm drain capacity, and the storage capacity of sewers and streets. Water comes up in the streets quickly and goes down quickly. This could be miles from a river, long before river flooding. Rainfall just can’t get out of the neighborhood quickly enough.
It can take days for water to migrate to a river and for large rainfall totals to force the river out of its banks. River flooding usually happens long AFTER street flooding, when the storm totals exceed the conveyance capacity of the river.
Importance of Location
During Imelda, New Caney received 20 inches of rain, while Lake Conroe received only 2. So there was no need to release water from Lake Conroe. For the most part, West Fork flooding was minor. But it was a totally different event on the East Fork where heavy rain piled up for two days.
Harvey – Classic River Flooding
During Harvey, the rainfall intensity only briefly surpassed the carrying capacity of infrastructure. But when the rain piled up in rivers for days and the SJRA released 80,000 cubic feet per second from Lake Conroe, tens of thousands of people flooded.
- As the river rose, the water in ditches had nowhere to go.
- As the ditches rose, the water in storm drains had nowhere to go.
- As the water in storm drains backed up, the water in streets had nowhere to go.
- And when the rain kept coming…you know the rest.
I hope this helps explain why some people flooded on May 7th and/or September 19th and not during Harvey. Different intensities and different durations produced different types of flooding.
Posted by Bob Rehak on 9/27/19 and dedicated to my old friend David Lyday
759 Days since Hurricane Harvey