Tc and IDF Curves

[CPENG78]

Modeling question: The ahj has a minimum Tc = 5 minutes requirement. The available rainfall data for the project site has IDF curves for a series of storms with a minimum duration of 10 minutes and higher. Have you encountered such a scenario? It would appear to me that I don't necessarily need to have an IDF for a 5 minute duration. I am working on confirming the storm duration requirement from the ahj. Your thoughts and input is greatly appreciated (My apologies if this seems as hydrology and hydraulics 101 questions but its a few years since I touched on this topic).

 

[gbam]

If you are in the States, NOAA Atlas 14 includes a 5 minute duration.

http://hdsc.nws.noaa.gov/hdsc/pfds/index.html

 

[CPENG78]

Thanks gbam.

I am in the states but my project is in the middle east.

Hope others can chime in on this to get additional input.

 

[SMIAH]

I'd then try to find out how were the IDF curves derived (measured data).
If you can get to the data (e.g. 5 min maximum annual daily precipitation or else), you could (re)perform the statistical analysis with it and make that IF curve for 5 min.

Less precise, but quick :
Plot the x (minutes) vs y (intensity) available and add a trend curve (power) and with the equation, estimate the value for the 5 minutes duration. I'd not design a dam with this procedure though.

 

[gbam]

One can use Statistics to extrapolate the 5-minute value, If you want to evaluate that.

 

[Twinkie]

On the same lines as SMIAH's idea but plot in Excel, adding a point for 0,0 and have excel plot the trendline, then just pluck the intensity of your new IDF graph.

Does it even rain in the Middle East anyway?

 

[SMIAH]

Does it even rain in the Middle East anyway?

What if all the points are 0,0 !?

Seriously if you use a duration-intensity series then don't plot the 0,0 in Excel as it won't match the rest and you won't be able to use the (power) trendline.
Just input the known values and with the equation y = Ax^-B calculate the y with a x of 5.

 

[gbam]

I don't think Excel has the appropriate plotting. Man its been a lifetime ago (25+years)but Gumble or Gimble distribution is what's floating in my head. I recall plotting depths on some type of graph paper to determine the return periods. I don't have my statistical hydrology book handy right now but I can check later.

 

[psmart]

Since IDF data generally follows a log-log relationship, just plot your data using log-log axes and extrapolate from the two lowest values down to 5 minutes.

Peter Smart
HydroCAD Software

http://www.hydrocad.net

 

[cvg]

I'd not design a dam with this procedure though

luckily, dams are not designed for small watersheds with a Tc of 5 minutes...

seriously though, I have generally just extrapolated these, by hand with a pencil. Plot the curve at a large scale, sketch in the line and measure with ruler. statistics is probably unnecessary here as a Tc of 5 minutes indicates a very small watershed.