ADCP measurements in regulated flow conditions

[Rana Bengül]

How do you do ADCP moving boat measurements in regulated flow conditions? I’m especially interested in the measurement length or amount of transects during a measurement.  I was told to measure more transects, which makes sense but how do I know I measured enough transects. In a discussion the measurement interval of 18 min came up but I was not able to find any literature on this. Any input, ideas, experiences or literature tips are appreciated. Thanks!

Topic

[Kristoffer Florvaag-Dybvik]

Hi. I am not familiar with the expression “regulated flow conditions”. Does it mean that discharge is changing rapidly due to opening/closing of upstream hatches/locks/turbines in an upstream dam? If so, I really don’t know. Perhaps, if you have 4-6 (or more) transects and you see a linear trend, you could calculate Deviation from trendline (and not from mean) and use the MPRR condition on that deviation?

For a steady flow NVE (Norway) have 3 conditions

– Even number of transects

– 12 minutes at least of total time (this comes from USGS + Qrev/QrevInt)

– No “red transects” in WinRiver, that is, no transect more different from the mean than the MPRR condition in WinRiver (https://www.researchgate.net/publication/305650623_Estimation_of_Type_A_uncertainty_of_moving-boat_ADCP_streamflow_measurements)

The last condition is encouraged also when using RSL (with Sontek ADCPs), even if the operator need to bring a table to check MPRR versus number of transects. For a very bad site, we may allow “red transects” after collecting 8-12 transects, if it seems obvious that we will not manage to avoid the “red transects” within a sensible amount of time.

In addition, if we QrevInt to process data on site, and we use the Oursin uncertainty model in QrevInt. If we see that COV (covariation, difference between transects) is not the biggest source of unsertainty, then we know that we have enough transects, and that to reduce uncertainty we need to improve somtehing else.

Attachments:

You must be logged in to view attached files.

[Rana Bengül]

Yes I mean flow measurements effected due to upstream hatches/locks/turbines etc. But this can also include really low flow velocities and the “bath-tub” effect (not sure how it is called in English, the situation were the water, due to the effects of the locks alternates the flow velocity and can lead in some cases even to a changes in flow direction)

[dave@genesishydrotech.com]

If the flow is oscillating, the best approach is to collect enough transects that you capture the full range of the change one or more times depending on the duration of the change. You can then decide in post processing if you want to average transects that cover the full range for one cycle to get an overall average, or average fewer (maybe use individual transects) to define the full range and clearly show the cyclic nature of the flow. Collecting many transects and then using individual transects or an average of just a few at a time is commonly done to characterize flow in tidal environments. In your situation, the key is to collect enough transects to fully characterize the flow changes.

 

Dave

[Rana Bengül]

Hello Dave,
I will have a look in only using individual transects as we have never done that before. Normally we average all usable transects from min to max of one cycle to get an overall average.

Thanks,
Rana

• This reply was modified 1 year, 1 month ago by Rana Bengül.

[Kristoffer Florvaag-Dybvik]

One more possible problem: Will these discharge measurements be used to make a rating curve for a gauging station? If so, you need to collect stage just as often as discharge, and you need to check if there is a time delay between the location where you measure discharge and the site of the stage measurements.

[Rana Bengül]

Thanks Kristoffer I will keep that in mind.

[Author]

Replies