the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Oct 2023
A Distributed Temperature Sensing based soil temperature profiler
Abstract. Storage of heat in the soil is one of the main components of the energy balance, and is essential in studying the land-atmosphere heat exchange. However, its measurement proves to be difficult, due to (vertical) soil heterogeneity and sensors easily disturbing the soil.
Improvements in precision and resolution of Distributed Temperature Sensing (DTS) equipment has resulted in widespread use in geoscientific studies. Multiple studies have shown the added value of spatially distributed measurements of soil temperature and soil heat flux. However, due to the spatial resolution of DTS measurements (~30 cm), soil temperature measurements with DTS have generally been restricted to (horizontal) spatially distributed measurements. In this paper a device is presented which allows high resolution measurements of (vertical) soil temperature profiles, by making use of a 3D printed screw-like structure.
A 50 cm tall probe is created from segments manufactured with fused filament 3D printing, and has a helical groove to guide and protect a fiber optic cable. This configuration increases the effective DTS measurement resolution, and will inhibit preferential flow along the probe. The probe was tested in the field, where the results were in agreement with the reference sensors. The high vertical resolution of the DTS-measured soil temperature allowed determination of the thermal diffusivity of the soil at a resolution of 2.5 cm, many times better than feasible with discrete probes.
Future improvements in the design could be integrated reference temperature probes, which would remove the need for DTS calibration baths. This could, in turn, support making the probes `plug and play' of the shelf instruments, without the need to splice cables or experience in DTS-setup design. The design can also support integrating an electrical conductor into the probe, and allow heat tracer experiments to derive both the heat capacity and thermal conductivity over depth at high resolution.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(11090 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11090 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2292', Bartosz Zawilski, 30 Oct 2023
In this paper, the authors describe a soil temperature probe construction that provides an almost continuous and accurate temperature profile. This probe is built using a 3D printer making it affordable. The concept is welcome as the instrument duplication cost is often a blocking point. However, following scientific deontology, existing probe design should be signaled and some raised or crucial points developed.
Indeed, a commercial probe for soil water content and temperature has a very similar design: Campbell SCI “SoilVue10” https://www.campbellsci.com/soilvue10. It is a discrete-level measurement probe; however, the entire sensor is screw-like. Also, a reference to “Soivue10” should be added as this probe exists already even if the employed technique is not the same.
A raised point but not developed is the inhomogeneity of the soil or the soil surface exposition to the sunlight. How these inhomogeneities can affect DTS probe measurements?
Another crucial point needs to be explicitly aborded even if it was not checked more than for 3-month measurements. The soil may be a very aggressive medium, especially for plastic used for 3D impressions. The 3D-printed DTS probe should resist soil humidity, eventual soil acidity, shrinkage, and so on. It may make it difficult, if not impossible, to use 3D-printed probes in some soils. I understand that it would be difficult to develop a point that was not extensively checked but I suggest including a warning for interested readers about it to push them to check their soil compatibility with the planned plastic filament before they invest in a similar project.
Finally, It would be very interesting to see authors continue their works and develop a probe as given in the Outlook paragraph and conclusion.
There are some remarks:
L-1. Formally speaking, it is not the heat storage that is the component of the energy balance but the heat flow through the soil surface, as the balance is for the energy flow, not the energy storage.
I think there is a little confusion about the energy flux balance and the soil temperature measurements, as the heat flux measurement is possible using several techniques (usually soil heat flux plates) but is also possible using a soil temperature profile. In the last case, it requires a temperature profile measurement in the soil versus the depth and time. Only surface measurement cannot allow soil heat flux measurement. The goal of the soil temperature measurements needs to be clarified.
L-26 “is strongly heterogeneous due to larger organic matter content” The soil inhomogeneity and organic matter presence are a possibility, not a general characteristic.
L-70 Unfortunately, this design does not guarantee good soil contact with the sensors in the case of vertisol, such as clayey soil. The same problem arises for the described sensor, and the statement of line 70 must be qualified.
L-96 PLA does not afford prolonged wet soil contact; is PTG resistant enough to soil humidity and, eventually, acid conditions? Was the probe checked for aging once it was installed in the soil?
L-171 The formula 1 supposes not only that the medium is homogeneous but also that the heat exchange is uniform on its surface (1D heat flow; there is no lateral heat exchange).
Figure 6. If I understand well, the temperatures of the reference probe were compared with the temperatures measured by the DTS probe all along the profile measured by the DTS probe. In this case, we have always had the best temperature correspondence, not for the same depth but about 1.5° lower (on DTS) for 10 and 30 cm depth, about 4cm lower on DTS for 0cm depth, and relatively good correspondence for the top of the liter. Is this shift resulting only from the liter thickness? I guess this point was particularly well-checked during reference probe installation.
L-230 The dampening is related to the soil density, not only to the organic matter content. The liter density is much lower than the soil density.
Figure 8-a “mean” means: averaged on the depth? If not, which depth is compared?
L-253 Please do not mix the diffusivity and the conductivity. It is tied, of course, but not the same. The liter is “poorly conductive” but also has a low heat capacity so poor conductivity is not enough to explain the low diffusivity.
L-259. To have an error of 1cm depth measurement with non-null angle insertion on the end of the 50 cm probe (the resulting error is most important on the end of the probe), the angle of insertion should be greater than 11° which is rather unlikely.
L-264 It would be interesting to have an estimation of the described probe cost. This is certainly one of its advantages.
Citation: https://doi.org/10.5194/egusphere-2023-2292-RC1 - AC1: 'Reply on RC1', Miriam Coenders-Gerrits, 10 Jan 2024
-
RC2: 'Comment on egusphere-2023-2292', Anonymous Referee #2, 05 Dec 2023
In this study, the authors introduce a DTS-based soil temperature probe which allows for the estimation of thermal diffusivity at high spatial resolution. I found the paper to be interesting and well written. I enjoyed the fact that the design of the probe is open-access. I found that the results of the DTS system compared to reference sensors are consistent and promising.
I only have a few comments and questions :
- The authors claim that “ It was possible to determine the thermal diffusivity of the soil in resolutions down to 2.5 cm”. Could the authors explain where this “2.5 cm” interval comes from ? What is the spatial sampling of the DTS measurements ? and how do you determine the position/location of the temperature point measurements around the probe ?
- A crucial consideration is the spatial resolution of DTS measurements.
First, I find that the manuscript does not clearly take into account the difference between the sampling and the spatial resolution (10.3390/s20020570; 10.1029/2008WR007052) (what is the performance of the DTS unit here ?)
Then, I wonder how the spatial resolution of measurements affects the results ? The collected data at sample spacing is not truly independent of their adjacent samples. Here, considering the size of the probe, the issue should be addressed. - L. 237 “For the DTS probe data we chose to estimate the diffusivity over increasingly large intervals, from a 2.5 cm wide interval near the surface, to a 10 cm wide interval near the deeper measurement points”. Could you show the results for each interval ? It would be interesting to see the differences. Why did you decide to present the results with a 5 cm interval (Figure 8) ?
- It seems that results are not consistent in the first cm of soils. Could it be due to the spatial resolution of DTS measurements ? The temperature measured near the surface also depends of temperature measurements outside the soil.
Minor comments :
- In streams, some studies already proposed to wrap the FO cable (1016/j.jhydrol.2009.10.033 ; 10.1029/2011WR011227)
- In completement of Bakker and des Tombe, you should cite 1029/2020WR028078, as the study includes the estimation of thermal conductivity.
- Concerning references, I have the feeling the most references are works of teams from The Netherlands. It would worst strengthen the past literature (https://doi.org/10.1016/bs.agron.2017.11.003)
Citation: https://doi.org/10.5194/egusphere-2023-2292-RC2 - AC2: 'Reply on RC2', Miriam Coenders-Gerrits, 10 Jan 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2292', Bartosz Zawilski, 30 Oct 2023
In this paper, the authors describe a soil temperature probe construction that provides an almost continuous and accurate temperature profile. This probe is built using a 3D printer making it affordable. The concept is welcome as the instrument duplication cost is often a blocking point. However, following scientific deontology, existing probe design should be signaled and some raised or crucial points developed.
Indeed, a commercial probe for soil water content and temperature has a very similar design: Campbell SCI “SoilVue10” https://www.campbellsci.com/soilvue10. It is a discrete-level measurement probe; however, the entire sensor is screw-like. Also, a reference to “Soivue10” should be added as this probe exists already even if the employed technique is not the same.
A raised point but not developed is the inhomogeneity of the soil or the soil surface exposition to the sunlight. How these inhomogeneities can affect DTS probe measurements?
Another crucial point needs to be explicitly aborded even if it was not checked more than for 3-month measurements. The soil may be a very aggressive medium, especially for plastic used for 3D impressions. The 3D-printed DTS probe should resist soil humidity, eventual soil acidity, shrinkage, and so on. It may make it difficult, if not impossible, to use 3D-printed probes in some soils. I understand that it would be difficult to develop a point that was not extensively checked but I suggest including a warning for interested readers about it to push them to check their soil compatibility with the planned plastic filament before they invest in a similar project.
Finally, It would be very interesting to see authors continue their works and develop a probe as given in the Outlook paragraph and conclusion.
There are some remarks:
L-1. Formally speaking, it is not the heat storage that is the component of the energy balance but the heat flow through the soil surface, as the balance is for the energy flow, not the energy storage.
I think there is a little confusion about the energy flux balance and the soil temperature measurements, as the heat flux measurement is possible using several techniques (usually soil heat flux plates) but is also possible using a soil temperature profile. In the last case, it requires a temperature profile measurement in the soil versus the depth and time. Only surface measurement cannot allow soil heat flux measurement. The goal of the soil temperature measurements needs to be clarified.
L-26 “is strongly heterogeneous due to larger organic matter content” The soil inhomogeneity and organic matter presence are a possibility, not a general characteristic.
L-70 Unfortunately, this design does not guarantee good soil contact with the sensors in the case of vertisol, such as clayey soil. The same problem arises for the described sensor, and the statement of line 70 must be qualified.
L-96 PLA does not afford prolonged wet soil contact; is PTG resistant enough to soil humidity and, eventually, acid conditions? Was the probe checked for aging once it was installed in the soil?
L-171 The formula 1 supposes not only that the medium is homogeneous but also that the heat exchange is uniform on its surface (1D heat flow; there is no lateral heat exchange).
Figure 6. If I understand well, the temperatures of the reference probe were compared with the temperatures measured by the DTS probe all along the profile measured by the DTS probe. In this case, we have always had the best temperature correspondence, not for the same depth but about 1.5° lower (on DTS) for 10 and 30 cm depth, about 4cm lower on DTS for 0cm depth, and relatively good correspondence for the top of the liter. Is this shift resulting only from the liter thickness? I guess this point was particularly well-checked during reference probe installation.
L-230 The dampening is related to the soil density, not only to the organic matter content. The liter density is much lower than the soil density.
Figure 8-a “mean” means: averaged on the depth? If not, which depth is compared?
L-253 Please do not mix the diffusivity and the conductivity. It is tied, of course, but not the same. The liter is “poorly conductive” but also has a low heat capacity so poor conductivity is not enough to explain the low diffusivity.
L-259. To have an error of 1cm depth measurement with non-null angle insertion on the end of the 50 cm probe (the resulting error is most important on the end of the probe), the angle of insertion should be greater than 11° which is rather unlikely.
L-264 It would be interesting to have an estimation of the described probe cost. This is certainly one of its advantages.
Citation: https://doi.org/10.5194/egusphere-2023-2292-RC1 - AC1: 'Reply on RC1', Miriam Coenders-Gerrits, 10 Jan 2024
-
RC2: 'Comment on egusphere-2023-2292', Anonymous Referee #2, 05 Dec 2023
In this study, the authors introduce a DTS-based soil temperature probe which allows for the estimation of thermal diffusivity at high spatial resolution. I found the paper to be interesting and well written. I enjoyed the fact that the design of the probe is open-access. I found that the results of the DTS system compared to reference sensors are consistent and promising.
I only have a few comments and questions :
- The authors claim that “ It was possible to determine the thermal diffusivity of the soil in resolutions down to 2.5 cm”. Could the authors explain where this “2.5 cm” interval comes from ? What is the spatial sampling of the DTS measurements ? and how do you determine the position/location of the temperature point measurements around the probe ?
- A crucial consideration is the spatial resolution of DTS measurements.
First, I find that the manuscript does not clearly take into account the difference between the sampling and the spatial resolution (10.3390/s20020570; 10.1029/2008WR007052) (what is the performance of the DTS unit here ?)
Then, I wonder how the spatial resolution of measurements affects the results ? The collected data at sample spacing is not truly independent of their adjacent samples. Here, considering the size of the probe, the issue should be addressed. - L. 237 “For the DTS probe data we chose to estimate the diffusivity over increasingly large intervals, from a 2.5 cm wide interval near the surface, to a 10 cm wide interval near the deeper measurement points”. Could you show the results for each interval ? It would be interesting to see the differences. Why did you decide to present the results with a 5 cm interval (Figure 8) ?
- It seems that results are not consistent in the first cm of soils. Could it be due to the spatial resolution of DTS measurements ? The temperature measured near the surface also depends of temperature measurements outside the soil.
Minor comments :
- In streams, some studies already proposed to wrap the FO cable (1016/j.jhydrol.2009.10.033 ; 10.1029/2011WR011227)
- In completement of Bakker and des Tombe, you should cite 1029/2020WR028078, as the study includes the estimation of thermal conductivity.
- Concerning references, I have the feeling the most references are works of teams from The Netherlands. It would worst strengthen the past literature (https://doi.org/10.1016/bs.agron.2017.11.003)
Citation: https://doi.org/10.5194/egusphere-2023-2292-RC2 - AC2: 'Reply on RC2', Miriam Coenders-Gerrits, 10 Jan 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 236 | 90 | 24 | 350 | 9 | 11 |
- HTML: 236
- PDF: 90
- XML: 24
- Total: 350
- BibTeX: 9
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 109 | 31 |
| Germany | 2 | 40 | 11 |
| Netherlands | 3 | 35 | 10 |
| China | 4 | 22 | 6 |
| France | 5 | 14 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 109
Bart Schilperoort
César Jiménez Rodríguez
Bas Van de Wiel
Miriam Coenders-Gerrits
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(11090 KB) - Metadata XML