Discussion: View Thread

go to NOAA Atlas 14.  If you use of the data is for design or something that may require governmental approval then you might need to check with your local government or state DEQ to ensure they are accepting atlas 14 for return frequencies.  They may only accept TP40 or some other return frequency analysis.

Good luck!

NOAA Altas 14

It looks like your original post and subsequent post are somewhat completely different (in concept). If you want actual rainfall data then you can go to the following website:

http://www.ncdc.noaa.gov/cdo-web/datasets

this website will allow you to query NOAA rain gauges closest to your site. The available sites might be too far from your study area and/or data not recent enough. Typically, local government or county government have a more robust rain gauge network where you could request data from them. If possible, try to obtain the lowest increment of rain they record (like 5-minute intervals). Once you have that data you can sum each duration interval to determine the max 5-minute, max 10-minute, max 15-minute…(and so on to max 24-hours or max 10-days…even up to max 60-days) in order to determine your largest return frequency compared against the first link that I sent. And when I say “sum each duration”…if, for example, you have 5-minute data, then you would have 12 possible values for a peak 1-hour increment as you would sum from 0:00 to 0:55, then 0:05 to 1:00, then from 0:10 to 1:05 and so on to give max peak hour. You would need to do this for every duration interval to determine max. This would allow you to say that the rain event on a particular date was an X-year return frequency.

Having said all of that…it sounds like you might need to develop your hydrology first though. If your culverts are in a FEMA designated floodplain, then you can look up the area in a Flood Insurance Study (FIS) to get flow rates in 10-year, 50-year, 100-year and 500-year intervals. (also, if in fema floodplain then you will likely have to submit a LOMR if changing the floodplain).  If you are not in a FEMA floodplain then you have to determine areas tributary to each culvert and then use something like TR55 to develop your peak Q for your design event (check with your state DEQ though because, for example, in Michigan we fill out a drainage request online and the MDEQ provides us the discharge rates from 2-year to 500-year return frequencies). As for your design event…that will depend on your state and local government…might be 100-year.

I could be wrong, but I think you’d only want recent rainfall data if you had a dynamic model of your drainage system with stream gauges and you wanted to calibrate your model using rainfall data compared to stream gauge data (or some metered data). With a calibrated model you would simulate the design event (like 100-year, 24-hour) and see how your system performs.  I’m not sure what else you would do with recent rainfall data in relation to the redesign of your culverts.  Actually, the only other use would be to prove that the recent rain events that caused flooding were beyond the design event.

Hope this helps.

------------------------------
Jennifer Morreale P.E., CFM, M.ASCE
Project Engineer
Hubbell, Roth & Clark
Detroit MI
(248) 454-6828

I am not aware of any sites that give you the frequency of precipitation event on the fly.

What I do is determine the precip frequency for any location using this site - this should give you the statistics of rainfall over any precip duration. I set this to MS - so hope it comes through.  Just center your cursor over the centroid of your watershed.

PFDS: Contiguous US

I then go to the AHPS cite and look for the precipitation rain event that you are interested.

AHPS Precipitation Analysis

Go to the location your are interested in, then go to the bottom to specify the date you are interested in, and read the precipitation color bar on the to determine the precip over the 24 hr period.

Here is a screen shot of the precip on March 19 over New Orleans, showing about 5 inches for that day. From the PFDS site, find the 24 hour duration row, go right until you see the column that has 5 inches.  Then go up the column to find the frequency in years.  Good for only 24 hr duration storms - but better that waiting!  Good luck

As Geoff Bonnin and Trevor Timberlake mentioned, we (MetStat) are continuously translating near real-time 1-, 2-, 3-, 6- and 24-hour gauge-adjusted radar-estimated precipitation into its equivalent Average Recurrence Interval (ARI).  We base our ARI results off the NWS MPE as well as QPE derived from MetStormLIVE (a high-resolution gauge-adjusted Dual-Pol radar-estimated precipitation product); these results can be seen here. We have an extensive archive, so if you need an ARI analysis for a specific duration/location, let me know. Similar to HDSC, we highlight extreme storms in our blog that often feature ARI analyses/maps.

We also create 6- and 24-hour ARI forecast maps out 5 days based on the WRF model four times a day. These are premium products, but are excellent for situational awareness of future extreme events. Below is a graphic of the maximum 6-hour ARI over the next 5 days.

Hope this helps.  Please reach out if you have any further questions.

(David -- we're practically neighbors in Fort Collins -- let's get together sometime!)