Hi James, yes there is the challenge. My take way from the report was that providing its exceeding 1cfs or 0.028m3/sec it can be done fairly accurately with traditionally USGS methods – which sounds like you are well in that range. For environmentally significant flows in drought California we are often looking at less than 1cfs/0.028m3/sec
The CTD-12 scale is 0 to 10,000mm with 0.5% accuracy across depth range, with the specification stating “no sharp changes in temperatures”. Resolution is different than accuracy. Specifically they are saying no temperature compensation. So the type of raw sensor (typically piezo electric, resistive) and how it responds to temperature change is critical – I’ve put out a lot of data on instruments I’ve used on how the apparent reported water depth changes with temperature. In some cases with lower cost instruments a small temperature change causes the reading to wildly swing across the range. So for the CTD-12 the range is +50mm to -50mm. So if you can find a measurement situation where which has pretty stable temperature change in water temperature (ice melt?) and can calibrate your CTD-12 across that temperature change, its likely to be provide useful readings. The places that have stable temperature’s are typically in ground water, if pumping they are also likely to have a large range of water movement – ideal perhaps for the CTD-12. When I checked with the manufactures over a year ago they where thinking about doing temperature compensation, but no statements yet on that.
So then I would think you are looking for the water column depth to vary significantly over the range of your instrument. Or for locations you have, what is the range of depth change for the flows. Can you place a sensor on the upstream side of the culvert
For environmental flows the hydrologists I’m working find that the Insitu LT500 is their instrument of choice with a 0.1% total accuracy (including temperature compensation) over 0-3m. It is however a top-end instrument when it comes to the sensor, the cable, and the stripped cable to be able to feed it in to a Mayfly modbus.
I’m personally trying to do more calibration an trialing with the Keller Nanolevel which is 0-3m and is a large capacitive based sensor. However the first Mayfly I deployed with it, got stolen after 9mnths, fortunately the sensor cable pulled away so still have the sensor. I’m looking to connect it over WiFi, so still working on that.