Stream gauge design

.

Below are some links for understanding and implementing tracer dilutions methods to develop stage-flow relationships. I suggest use of this in most small (1st order to 3rd order) streams where weirs or culverts without tailwater influences are not available. I've used table salt (generally along the coast) and rhodamine dye. Slug tests are the easiest to do in the field, but you need to make sure you have very good mixing and flow-through.










...also strongly suggested as a basic reference for anyone measuring stream flow is the US Bureau of Reclamation's "Water Measurement Manual". Below are download links and the table of contents...



If you need more information, post again and I'll try to help.

.


tsgrue: site engineering, stormwater
management, landscape design, ecosystem
rehabilitation, mathematical simulation

Hmmm... thanks for the info. But perhaps I should clarify that we're making improvements to a *permanent* flow measurement station, logging 24/7/365. Another consultant is developing a stage-flow curve. I just need to ensure that the level measurement is accurate.

My concern is mounting the level sensor (most likely a submersible transducer) in such a way that it is accurate, repeatable, and protected from sediment and impact. Although this is an instrumentation question, I thought Civils might have input since it's really about installation methods. I can think of benefits and drawbacks to stilling wells, or u-bolting the sensor to the upstream side of the weir where it meets the stream bed.

A few years ago I found the ITRC report below that specifically addresses use of water level sensors that you might want to check out. Unfortunately I found this information *after* purchasing several pressure transducers for a low-budget monitoring effort.

Pressure transducers can be buggy - you will probably get what you pay for in terms of accuracy. I had some cheapies (less than $500) that went out of spec within months of installation, showed unexplicable diurnal fluctuations, etc; I will never look at pressure transducers as an "inexpensive" option again considering the amount of data we missed. For your scenario, if you use a transducer, I think that not using a stilling well may very well cause problems with the data quality. For example, how will you really know there wasn't something stuck on the end of the sensor (even bubbles can through off some of the models).

Maybe a non-recording simple crest-stage gage (the kind with cork dust in a tube... cost about $20USD to make) would be sufficient to verify the data...




pg77

I would contact the USGS. They probably have the most experience with stream guages. If I assume from your screename that you are in CA, here is the website for USGS surface water data (which includes 449 real time guages and over 2000 daily guages):


One can follow links from that site to find contacts.

Terry - thanks for the suggestion, and yes I'm in Ca. I may follow up on this.

bltseattle - That ITRC report is very informative, mainly because it compared level sensors in real world conditions. It helped to quantify a suspicion I've held that ultrasonic sensors could have more variability due to waves and path issues. One mistake in the report though - even when it was written in 1999, PLCs were commonly available with 4-20mA inputs (not just 0-5V). Allen-Bradley products come to mind.

We got further clarification from our client that an existing sensor has done well for a few years. It's strapped to angle iron, and the angle is concreted into the stream bed. The sensor (a submersible transducer) is contained in a perforated PVC stilling well wrapped with geotextile fabric in order to allow water level equalization without sediment entry. I suspect we'll mimic their installation. Nice to see their "Stevens" brand level sensors were well received by ITRC...