Seems like this thread is as good a place as any to post this. Back when I was a CE major I interned for several summers for a firm that specialized in stormwater management. This was back when FEMA was beginning their updates on the Houston area flood maps and the first was Brays Bayou. That study was the first to integrate the use of LIDAR data for elevation mapping. The firm I was working for did the majority of the work and all of the flood mapping.
There are two principal pieces of software used; both created and supported the the USACoE. One determines the volume and timing of water that will pass a given location within the watershed following a given rainfall event. The second takes those volumes and determines the depth that volume will reach as it passes that reach along the stream.
For those that might be curious... It's a good deal of reading, but here are the NOAA papers for rainfall duration, intensity, and probably frequency that are used by the smart people to model 100yr floods.
Hydro-35 5 to 60 Min Precip Freq for Eastern & Central USTech Paper 40 -- Rainfall Freq Atlas for Duration of 30min-24hrs for 1-100yr eventsTech Paper 49 -- Rainfall Intensity 2-10Day for Return Periods of 2 to 100 YearsThese contain the graphs for hypothetical returns that generate a given flood event. I believe most modeling is done from the 24 hr events. The thing is there are actually 5 min 100yr rains, 10 min 100 yr rains, 30 min, 60 min, 2 hr, 12 hr, all the way out to 10 days. Obviously the shorter duration are higher intensity. Each of these however (for a "100 Year flood') has a 1.0% chance of occurring in any given year. Same goes for the 500 year event, it's a 0.2% chance in any given year event. Furthermore, it's for that given area.
It just kinda grinds on me when I hear folks say, "this is the second 500 year flood in X years." Well that's not entirely true. Take TS Allison for example. Allison dumped most of its rain inside Beltway8 on the west side between there an downtown. The vast majority of that was inside Loop 610. Yesterday's rain was more focused outside Beltway 8 towards Katy, Hockley, Tomball, and around to the Woodlands. So I'd argue that, No, you have not had 2 500 year floods in X years. Two different areas have each seen 1 500 Year (0.2% chance) floods.
Can someone remind me where the Memorial Day floods in Houston were the worst?
I realize there has been widespread localized flooding and a lot of street flooding. Most streets are designed to handle a 2 year (50% chance) rainfall event. Neighborhood and individual business retention/detention ponds are designed to hold back 100 yr events to slow their release into the watershed at pre-development rates. Once the 100 Yr design capacity is reached the excess spills unchecked over a weir into the bayou or creek. Back-flow preventers should keep creeks from backing up into neighborhoods where they are installed and functioning properly.