Rishudh Thakur is delivering a seminar talk today, please add any follow up questions you might have here, thanks!
ZC
This talk presents a field-scale application of Fiber Optic Distributed Temperature Sensing (FO-DTS) as a practical tool to improve interconnected surface water (ISW) identification. FO-DTS enables continuous temperature measurement at sub-meter spatial resolution along a fiber-optic cable placed on the streambed, over distances of several kilometers.
Recording (Passcode: W@dEd1GH)
What limits the length of cable that can be laid out? I presume there is some limit to the max length of a single cable which is what? and is the limit just because it it too hard to handle a longer cable, or is it related to attenuation/accuracy of the measurement?
what is the cost of buying or renting a cable?
How long did it take to lay out the one km of cable?
Since the connectivity is rapidly changing, what does this say about how much WTD is changing over space? is the maximum hydraulic gradient very high (what would it be approximately)?
Just reiterating my question from the talk: How can you estimate the seepage rate from this FODAS data?
@SashaMcLarty I know you wanted to catch this recording! ^^
For laying the cable the first time it took us about 4.5 - 5 hours, but the second time it took us only about 3 hours. We were more familiar with the process by then. We were 3–4 people at any given time while laying the cables. The river stage and streamflow can definitely add or reduce the time it takes to lay the cables.
The fiber optic cables we used in our study cost about $5 per meter. They can also be rented from C-TEMPS for $ 1.25 per meter. Although, if the cable has been used a lot then it becomes difficult to read the meter marks on the cable. These marks are necessary for georeferencing the cable correctly.
Basically, to estimate seepage rate we need to solve the 1D heat transport equation, and this can be solved with the help of data acquired with DTS in specific configurations or DTS + other sensors.
If we want to measure the seepage rate with the passive DTS cables (ones we deployed along the Tuolumne) then we need to deploy two or more cables vertically separated by each other. One cable would be on the riverbed, and others would be buried beneath this cable. We can then model the vertical temperature gradient to estimate the direction and magnitude of seepage (flux) using heat as a tracer.
You can also avoid the above setup and use vertical temperature profilers at multiple points along the cable to get the temperature boundary conditions and extrapolate them for the entire cable. But this is based on the assumption that temperature boundary conditions do not vary between the profilers. A technique which overcomes this assumption is the active DTS technique which involves heating up the cable. This approach optimizes for each measurement point along the heated section, the values of thermal conductivity and flux that best reproduce the temperature increase measured over the heating period.
If you want to read further, I am providing the link to three papers that gets to your question in detail. The seepage rate estimated from DTS has shown to be in agreement with the seepage rates estimated from other techniques namely differential gauging, dilution gauging, and geochemical techniques.
Paper 1: HESS - Combining passive and active distributed temperature sensing measurements to locate and quantify groundwater discharge variability into a headwater stream
Paper 2: https://ngwa.onlinelibrary.wiley.com/doi/full/10.1111/gwat.13007
Paper 3: https://onlinelibrary.wiley.com/doi/full/10.1002/hyp.8200
Its a combination of both, but from my experience with deploying a cable for just 1 km, I believe it would be very hard to handle a cable of longer length. The 1100 m long reel that we had weighed upwards of 100 pounds. Also, from the papers I have read which deployed the cables as part of the study, almost all of them were about a km or less. A study by John Selker estimated lakebed temperatures across the lake of Geneva, but they leveraged an existing fiber optic cable running through the lake.
In terms of the physics of the measurement, the DTS precision is affected by the signal to noise ratio (SNR). For a longer cable less light will arrive at a measurement point located further along the cable, and the temperature dependent backscatter would attenuate even further. This will result in a high SNR. To overcome this we would need to increase both the temporal and linear integration distance to get more backscatter, costing us both the temporal and spatial resolution of the measurements. For the 1 km cable, the optimal integration intervals for 0.1 degree C precision were 0.5 m and 4 minutes. All these numbers would be larger for a longer cable.