In this work, a time domain reflectometry (TDR)-based system for continuous and diffused monitoring of soil water content in agriculture is presented. The proposed TDR-based system employs elongate sensing elements (SEs). In practical application, each wire-like SE is buried along the cultivation row to be monitored, and through a single TDR measurement it is possible to retrieve the water content profile of the cultivation along the length of the SE. By connecting the TDR-based monitoring system to the irrigation machines, it would be possible to automatically start/stop irrigation based on the actual water requirement of the cultivations, thus favoring precision agriculture and enhancing irrigation efficiency. To demonstrate the feasibility of the proposed monitoring solution, a dedicated hardware+software platform was developed and the TDR-based system was experimented in open-field cultivations.
Irrigation is crucial in the economic and productive management of agricultural systems; as a result, the efficiency of water use has become a priority also due to the increasing limitations on this natural resource [
Appropriately designed and built irrigation systems are essential for optimizing the irrigation process, increasing cultivation profitability, rationalizing available resources, and reducing waste. In such a scenario, the automated and real-time monitoring of the actual water requirement is strategic for the optimal management of agricultural systems.
The first step for assessing the cultivation’s water requirement is to assess soil moisture content. However, at the state of the art, soil moisture content sensors are usually point sensors; therefore, a large number of point sensors would be required to monitor extended cultivation areas. This ends up limiting the possibility of real-time monitoring, and automated irrigation control can only rely (at best) on preestablished irrigation scheduling or on weather stations.
In such a context, the objective of this work was to develop and validate an innovative system based on the use of time domain reflectometry (TDR) technique and of elongate, low-cost sensing elements (SEs) for continuous and real-time, diffused monitoring of soil water content. The SEs are buried in the soil, close to the cultivations to be monitored. One single SE could be tens of meters long and would allow assessing the water content profile of the cultivation.
The developed monitoring system would communicate with the irrigation systems, and it would be able to control the valves automatically, depending on the water content detected in the soil, possibly allowing limiting irrigation only to the zones in need of water. In practical applications, when the measured soil moisture content falls below an alert threshold, the system would open the automatic control valves, thus starting irrigation. Similarly, in the case of excessive water content in the soil (for example, because of rain), the system would prevent the start of the pumping system, thus delaying irrigation. The proposed system can encompass the three major components of a sustainable agricultural management as identified in [
In the following, first, the background is briefly outlined. Successively, the materials, the methods, and the developed hardware/software platforms are described in detail. Then, the results of the experimental validation of the system in two practical application scenarios are reported (one set of experiment relates to a cultivation of trees and the other relates to plant cultivation). Finally, results are discussed and conclusions are drawn.
TDR is a well-established measurement technique used in several fields. This technique has low implementation costs and offers the possibility of remote control and continuous monitoring. Additional features such as real-time response and adaptability make TDR particularly attractive for countless applications (e.g., localization of wire faults [
The literature related to TDR-based soil moisture content measurements is extensive [
TDR relies on the analysis of the signal that is reflected when an electromagnetic signal (typically, a voltage step signal with very fast rise time or a pulse signal) is propagated along a probe or a SE inserted into the system under test. In particular, the propagating TDR signal is reflected at the electrical impedance variations encountered along its path [
The typical output of a TDR measurement is a reflectogram, which displays the reflection coefficient (
As a general rule, the higher the relative dielectric permittivity of the medium in which the SE is inserted, the slower will the TDR signal propagate along the SE [
With regard to
On such basis, the proposed TDR-based system relies on sensing the variations of the dielectric permittivity that occur in the soil as a result of irrigation. In fact, the relative dielectric permittivity of water is approximately 78 [
Two major sets of experiments were carried out for validating the proposed system: soil water content measurements in a cultivation of trees; soil water content measurements in plant cultivations.
In the following sections, the materials and methods used in these experiments are described in detail.
For the diffused soil moisture measurements, the TDR-307usb instrument was used. This instrument was selected because of the optimal trade-off between measurement accuracy and low cost (approximately 1’000 Euros); hence, it is ideal for agricultural applications which are generally characterized by stringent cost requirements. The TDR-307usb is a portable instrument that generates a pulse signal, whose duration can be varied from 10 ns to 50
As for the elongate SE, the sensing section consists of an RG59 coaxial cable and of a wire (W1), which are mutually insulated and run parallel to each other. Figure
Axonometric cross section of the elongate SE. A copper wire runs parallel to a coaxial cable.
The coaxial cable serves only the purpose of calibrating the distance. In fact, first the TDR signal is propagated along the coaxial cable to assess the actual length of the sensing section (in this way, the SE can be cut according to the cultivation length, without preliminarily establishing the length of the SE).
Instead, the SE itself consists of the bifilar transmission line that is formed by two elongate conductors: the outer conductor of the RG59 and the wire W1.
For the practical use of the proposed system, also a dedicated hardware platform for controlling the irrigation system was implemented. The core of the hardware platform consists of the TDR RI-307usb, a Raspberry Pi, a relay module (SMTRELAY08), and electrovalves. The hardware platform also includes the necessary peripherals, namely, a touchscreen monitor display, a 2A 5V power supply, and a memory card.
As for the developed software, in addition to acquiring and storing the measurement data, the developed software can control the valves of the irrigation system, according to the measured value of soil water content. The operating system chosen for the software is Raspbian Stretch with desktop version: the programming language chosen to write the algorithm was Python 3. The software allows selecting a distance range in which to check soil water content; setting soil water content thresholds, in correspondence of which the valves are to be opened/closed.
With regard to the output of the monitoring system, specific data are provided in the experimental section. However, it is useful to describe the behaviour of the typical output. By default, the abscissa axis reports the apparent distance (
As for the ordinate axis, in the previous section it was mentioned that typically the reflection coefficient (
In the first set of experiments, a tree cultivation was considered. The separation between the trees was approximately 20 m. A SE with length 318 m was placed along the row of trees, as depicted in Figure
Experiment #1: schematization of the disposition of the SE along the row of trees. The trees were approximately 20 m apart. Dimensions not to scale.
Also this experiment was carried out in an open-field cultivation. In particular, 12 SEs were positioned along as many rows of plants; the length of each elongate SE was
Experiment #2: (a) top-view schematization of the experimental setup.
For the TDR measurements, the TDR-307usb was connected to the beginning of each SE, thus obtaining one reflectogram for each plant row.
In addition to the elongate SEs, also eight two-rod probes (
Figure
Figures
Experiment #2: (a) picture of the arrangement of the elongate SEs. (b) Picture of one of the vertical probes.
It is important to point out that the TDR measurements with the vertical probes were carried out through the TDR200 reflectometer. This portable instrument generates a step-like signal with fast rise time. This TDR instrument includes a more sophisticated electronic; hence, the cost of the TDR200 is approximately 5’000 Euros. Although the cost of this piece of equipment is higher, it has the advantage that it can be connected to multiplexers and can control up to 512 SEs, thus reducing considerably the implementation costs.
After positioning the elongate SE as depicted in Figure
Experiment #1: (a) reflectogram acquired in dry condition; (b) reflectogram acquired after drip irrigation. The threshold curves for
Successively, the trees were drip-irrigated for several hours. After irrigation, another reflectogram was acquired (Figure
It can be observed that although the same amount of water (approximately 50 litres) was provided to each tree, the peaks are less prominent as the distance increases. This is due to the attenuation of the propagating TDR electromagnetic signal along the 318 m long SE.
Because of electromagnetic losses, minima at difference distance
As mentioned in Section
The thresholds
On the other hand, if the dip is lower that
When the dip rises up to and (again) higher than
This cycle repeats throughout the cultivation lifetime. This procedure allows maintaining the desired amount of soil water content, thus satisfying the cultivation water requirement.
In the second set of experiments, the arrangement of the SEs was as depicted in Figure Day #1: approximately one week after seeding to make sure that the soil was settled; Day #14: one hour after 3.5 h long irrigation; Day #20: before irrigation; Day #21: approx. 10 hours after a 3.5 h long irrigation; Day #24: before irrigation.
For the sake of brevity only the results related to one of the elongate SEs are reported. In particular, Figure Day #1: in this condition, the soil was dry. From the reflectogram, at approximately Day #14: the plants were irrigated for 3.5 h. The TDR measurement was carried out after waiting approximately one hour, in order to allow water to diffuse and settle. It can be seen that while the Day #20: one day before the scheduled irrigation, another TDR measurement was carried out. From the reflectogram, it can be seen that Day #21: the irrigation went on for 3.5 h, and after waiting an additional ten hours, another reflectogram was acquired. As expected, because of the irrigation, the apparent length of the SE has increased, as Day #24: one day after irrigation, another TDR measurement was performed. Once again, the apparent length of the SE has decreased (
Experiment #2: comparison of the reflectogram acquired from SE3 during the observation period: (a) day #1; (b) day #14; (c) day #20; (d) day #21; (e) day #24. The shift of the abscissa corresponding to the end of the SE indicates a different amount of soil water content.
The observation lasted for approximately three months, for the whole lifetime of the cultivations. The trend of the lengthening/shortening of the apparent length of the SE was confirmed throughout the observation period. The behaviour was consistent with all the 12 SEs.
It should be mentioned that, from these measurements, it was not possible to discriminate zones with higher moisture content. This is due to the fact that the mutual distance between the plants was approximately 30 cm, and as a result of the irrigation the water diffused also in between the plants. However, the lengthening of the apparent length of SE was a consistent indicator of the variation of soil moisture content.
For the sake of example, Figure
Experiment #2: reflectograms acquired from the vertical two-rod probe, v1.
In view of possible application of the proposed system, it is important to make some considerations regarding the implementation costs and modalities.
With regard to the cost of the SE, it may vary from approx. 0.20 €/m to 7 €/m, depending on the quality of the SE (in terms of internal losses, reusability, robustness, and durability in time) as required for the specific application. For example, for seasonal cultivations, it may be advisable to employ low-cost disposable SEs, which are used for a season and are then removed from the soil and disposed of. For monitoring cultivations of trees, on the other hand, more robust and durable (and costly) SEs could be used. In fact, in this case, the lifetime of the SE is supposed to be comparable with that of the tree.
With regard to the TDR measurement instrument, as aforementioned, with some TDR instruments (as the TDR 200), it is possible to employ multiplexing devices. These allow controlling up to 512 SEs with a single measurement instrument, thus considerably reducing the implementation costs. Figure
Sketch of the possible configuration of the system for monitoring, with a single measurement instrument, up to 512 independent SEs.
As for the installation of the SE, it consists only in burying the SE approximately 10-50 cm underground (depending on the cultivation). The procedure is very simple and only requires letting one end of the SE emerge from the soil, in order to allow the connection to the measuring instrument.
Finally, with regard to the “density” of installation of SEs (i.e., number of SEs per ha), thanks to the modularity of the proposed system, it follows that a high degree of freedom is available. Based on the experimental validation carried out so far, large farms could be entirely monitored with hundreds of SEs; however, the farmer could choose to monitor the moisture content profile on selected zones in the cultivations. The operator may as well choose not to install a SE along each row, but, for example, one SE every 0.5 ha. In the experience of the authors so far, for a given cultivation, an optimal trade-off for the density of installation may be one SE every 10-15 rows. And, if multiple cultivations are present, then the aforementioned procedure should be followed for each cultivation.
In this work, a TDR-based system for continuous and diffused monitoring of soil water content in agriculture was presented. Differently from traditional point sensors, the proposed system allows monitoring the soil water content profile along elongate SEs buried along the cultivation rows. The suitability of the developed TDR-based system for the intended application was tested in actual open-field cultivations.
The proposed system holds considerable potential as a cost-effective solution for real-time monitoring of soil water content in agriculture. In practical applications, by connecting the proposed measurement system to the irrigation machines, it would be possible to automatically start/stop irrigation based on the actual water requirement of the cultivations. This integration of the proposed system with irrigation systems could favor precision agriculture and enhancing irrigation efficiency.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors declare that there are no conflicts of interest regarding the publication of this paper.
This research was funded by Regione Puglia-Area Politiche per lo Sviluppo Rurale-Servizio Agricoltura within the Public Notice “Call for Project Proposals for Research and Experimentation in Agriculture,” as part of the activities of the research project “SCOPRI - Sistema per il Controllo ed Ottimizzazione dei PRocessi di Irrigazione” (Eng.: “System for Control and Optimization of PRocesses of Irrigation”, project ID number: PRS 100). The paper reflects the authors’ point of view and not Regione Puglia’s.