Real-Time Monitoring and Simultaneous Verification of Water Percolation Using Electrical Resistivity Tomography and Photography Techniques

Abstract

Water percolation usually occurs in soil, making it highly challenging to simultaneously monitor and verify the water percolation process in real-time. We propose employing electrical resistivity tomography and the photography technique, respectively, to visualize and validate water percolation. An experiment was conducted to monitor water percolation in sand within a transparent glass trough using electrical resistivity tomography (ERT) and photography techniques. The experimental results show that the relatively low-resistivity anomalies in the electrical resistivity profiles and correction electrical resistivity profiles, derived from the experimental data, have a half-ellipse shape, while the wetting areas captured in the photographs exhibit a bulb-shaped pattern. The relatively low-electrical-resistivity anomaly areas delineated by the 2000 Ω·m contour line in the electrical resistivity profiles and the 120 Ω·m contour line in the correction electrical resistivity profiles exhibit a remarkable correspondence with the wetting areas captured in the simultaneous photographs. Our findings show that the ERT and photography techniques are suitable for the real-time monitoring and simultaneous verification of water percolation in sand within a narrow glass trough.

1. Introduction

Water seepage phenomena are closely tied to human activities, such as agricultural irrigation, the construction of water-related infrastructure, and the prevention and management of water inrush in coal mines or tunnels. A lucid visual comprehension of water percolation is of paramount importance. Water percolation generally occurs in soil and can be likened to a process occurring within a black box, making it highly challenging to monitor in real time and simultaneously verify.

Traditionally, numerous methods have been employed to discern moisture content and monitor water percolation in soil. Noteworthy methods include the neutron probe method [1,2,3] and the time domain reflectometer method [4,5,6]. However, these methods have two inherent limitations. Firstly, the pursuit of water percolation monitoring through these means often incurs destruction. Installing sensors into the soil disrupts its natural structure and alters its water percolation pathways. Secondly, these techniques only capture point-localized information due to their point measurements, failing to capture a comprehensive profile and volumetric water percolation data.

To solve these limitations, geophysical exploration methods, such as ERT [7,8,9,10,11,12,13] and ground-penetrating radar [14,15,16,17], are commonly employed. These methods have a minimal impact on the soil and allow for profile measurements. In the ground-penetrating radar method, scanning must be conducted to obtain a soil profile during monitoring, requiring continuous movement and scanning. However, manual movement may result in variations in the data collection points due to inconsistent routes. In contrast, for the resistivity tomography method, one fixes the electrode positions in advance, enabling automated data collection. Therefore, resistivity tomography is more suitable for monitoring water percolation in soil. To verify monitoring accuracy, it is often necessary to employ other geophysical methods for mutual verification [18,19] or to perform sampling and drilling [20,21,22]. However, the verification provided using other geophysical methods is only indirect and lacks intuitive confirmation. Sampling or drilling provides point-based direct verification but does not enable simultaneous verification across the profile.

Artificially synthesized transparent soil consists of two components: transparent soil particles and a pore fluid with corresponding refractive indices. Because these two components have similar or nearly identical refractive indices, light can pass through them without refraction, making this “soil” transparent. Transparent soil particles are typically crafted from amorphous silica powder, amorphous silica gel, or fused quartz. Meanwhile, the pore fluid is commonly formulated by blending saturated hydrocarbons and mineral oil in specific proportions or by dissolving solid calcium bromide in water to create a solution of a predetermined concentration [23,24]. In recent years, some scholars have utilized transparent soil in combination with imaging techniques to study water percolation [25,26,27,28]. The use of transparent soil combined with photography techniques allows for the simulation and simultaneous verification of percolation processes [29,30,31]. However, the components that comprise transparent soil differ significantly from those of actual soil. Transparent soil experiments typically involve small volumes, making the performance of equal or representative proportional experiments challenging.

To address the abovementioned problems, we propose the utilization of ERT and photography techniques to monitor in real time and simultaneously verify water percolation in sand filling a narrow transparent glass trough. In this study, our objectives were to utilize ERT to visualize the water percolation process in the sand and simultaneously verify it through photographs. To assess the effectiveness of visualization and validation, our primary focus was to compare the coherence between the relatively low-resistivity anomaly areas in the electrical resistivity profile or correction electrical resistivity profile and the wetting areas captured in the photographs.

2. Method and Experiment

2.1. Method

The ERT method entails the injection of a low-frequency electrical current into a geological body through a pair of current source electrodes (designated as A and B) and the measurement of the potential between two electrodes (referred to as M and N) or multiple different potential electrodes. Electrical resistivity can be calculated using parameters such as the current, voltage, electrode spacing, and electrode configuration parameter. The equation for electrical resistivity is as follows:

$$\mathsf{\rho }=\mathrm{K}\frac{∆\mathrm{U}}{\mathrm{I}}$$

(1)

$$\mathrm{K}=2\mathsf{\pi }\frac{1}{\left(\mathrm{A}\mathrm{M}-\mathrm{A}\mathrm{N}\right)-\left(\mathrm{B}\mathrm{M}-\mathrm{B}\mathrm{N}\right)}$$

(2)

where ρ is the electrical resistivity, and K is the geometrical coefficient that depends on the position of the four electrodes A, B, M, and N. ΔU is the potential measured between the electrodes M and N. I is the current injected through electrodes A and B. AM, BM, AN, and BN represent the geometrical distance between electrodes A and M, B and M, A and N, and B and N, respectively.

Common electrode arrays in ERT include pole–pole, pole–dipole, Wenner, dipole–dipole, and others, which are determined by the arrangement of electrodes. When AM = MN = NB, the device coefficient k = π/AM, and this electrode array is referred to as the Wenner array.ERT achieves profiling and depth detection by moving the minimum electrode array and altering the electrode spacing within the electrode array. When the electrode spacing within the electrode array is fixed, with each movement of the electrode array, Equation (1) is used to determine the electrical resistivity value at a certain depth point. By continuously moving, we can obtain an electrical resistivity profile at the same depth. By changing the electrode spacing within the electrode array once and following the profiling method, you can acquire another electrical resistivity profile at a different depth. Continuing to adjust the electrode spacing within the electrode array generates a series of depth profiles. Collecting all these depth profile electrical resistivity data allows one to construct ERT.

The foundation of ERT in studying soil moisture and water percolation in soil lies in establishing the relationship between electrical resistivity and water content. Commonly employed relationships include that in the equation below(Equation (3), which is specifically applicable to coarse-grain soils [32].

$$\mathsf{\rho }={\mathsf{\rho }}_{\mathrm{w}}{\mathsf{\Phi }}^{\left({\mathrm{s}}_{\mathrm{a}}-{\mathrm{c}}_{\mathrm{a}}\right)}{\mathsf{\theta }}^{{\mathrm{s}}_{\mathrm{a}}}$$

(3)

where $\mathsf{\rho }$ and ${\mathsf{\rho }}_{\mathrm{w}}$ are theelectricalresistivityof the soiland the electrical resistivity of the pore water (Ω·m), $\mathsf{\Phi }$ is the soil’s porosity (%), θ is the water content, and ${\mathrm{s}}_{\mathrm{a}}$ and ${\mathrm{c}}_{\mathrm{a}}$ are empirical parameters called Archie’s cementation and saturation exponents, respectively.

Using the relationship between soil resistivity and moisture content in Equation (3), we can qualitatively assess the percolation of water. The greater the electrical resistivity of the soil is, the lower the soil’s moisture content, and, conversely, the higher the moisture content. By analyzing the spatiotemporal changes in electrical resistivity, we can assess variations in the percolation process and the moisture content during percolation.

2.2. Experiment

2.2.1. Materials

The experiment involving the real-time monitoring and simultaneous verification of water percolation in sand within a narrow glass trough was conducted at the geophysical laboratory of the China University of Mining and Technology. Sandy soils obtained from a riverside near Xuzhou City, Jiangsu Province, were utilized in the experiment. The sandy soils were sieved through screens with sizes of 5 mm, 2 mm, 0.5 mm, 0.45 mm, and 0.074 mm. The sandy soils collected from each screen were carefully gathered and weighed.

Table 1. Sandy soil gradation in the experiment.

Media Diameter (mm)	Mass Percent (%)
>5	0.4
2	7.8
0.5	51.2
0.45	32.1
0.074	7.6
<0.074	0.9

shows that the sandy soils with diameters ranging from 0.074 mm to 2 mm constituted approximately 98.7% of the total mass.

The narrow transparent glass trough depicted in Figure 1 was filled with sand soil. It had dimensions of 150 cm in length, 10 cm in width, and 50 cm in depth (Figure 1) and had six viewpoints, five of which were closed, leaving only the top view open without glass.

2.2.2. Design of Experiment

The water supply system (Figure 1) comprises three components: a water tank, a hose, and a control valve. The water tank was a 20 L plastic bucket. The hose had two sides. One side of the hose was connected to a plastic bucket, while the other was used as the water outlet. The control valve was in the middle of the hose. To enable water dynamics, the bucket was placed on a desk with a height of 0.5 m.

The acquisition system comprised electrodes, an electrical resistivity meter, and connecting cables. The electrodes were made of pencil lead and had a length of 4 cm and a diameter of 0.07 cm. In our experiment, 29 electrodes were placed at intervals of 5 cm along the measuring line (Figure 1). The electrodes were inserted to a depth of 3 cm, ensuring a consistent electrode position and contact resistance throughout the experiment. Additionally, a reference electrode was installed near the first electrode, and an infinity electrode was placed in the northwest corner of the glass trough. The instrument for acquiring ERT data in the experiment was a parallel electrical resistivity network system (WBD-1).

The working principle of the Network Parallel Electrical Resistivity Method (WBD-1) entails the use of the AM or ABM method with single-pole or dipole power supply. All remaining electrodes function as simultaneous receivers. Each electrode takes turns acting as the power supply electrode, facilitating the swift acquisition of a significant volume of data within a brief timeframe. The collected data arethen processed and deciphered using WBDPro, leading to apparent resistivity computations. The resistivity data are exported in the conventional electrode array format and subsequently employed for mapping with suitable software.

For the data-acquiring mode, we followed the AM method, with each power supply excitation lasting for 0.5 ms and a sampling interval of 50 ms. The power supply mode was single-positive. The environmental temperature was room temperature, and the water source was tap water from the geophysical laboratory at China University of Mining and Technology. The electrical resistivity of the tap water at room temperature was quantified at 10 Ω·m, determined through the Wenner method with an electrode spacing of 1 cm.

The procedure for acquiring data was as follows: firstly, one data point was acquired using WBD-1 before supplying water as the background data; secondly, we opened the valve and started the timer, and the water outlet was positioned between the 15# and 16# electrodes along the measuring line. The electrical resistivity instrument acquired one data point every ten minutes, with a total acquisition time of one hour. Seven groups of data were obtained in this experiment.

The data collected using the WBD-1 parallel electrical instrument were processed using specialized software known as WBDpro. Initially, the potential distribution patterns measured using the other measurement electrodes were sequentially examined when each current electrode provided a current. Any anomalies in the data were either removed or corrected. Then, the data were decoded using averaging methods with a selected cutoff time of 50 ms to minimize interference. Finally, based on the data from the right side of the current electrode and utilizing Equation (1),the apparent resistivity was calculated. We extracted and exported data according to the Wenner array configuration, and created an apparent electrical resistivity profile using Surfer8 software.

3. Data Processing

Due to the effect of the narrow glass trough, the electrical resistivity values and the contour lines in the electrical resistivity profiles based on the experimental data differed from the true value and distribution. To address this issue, a correction equation was applied to the experimental data. To evaluate the effectiveness of the correction, a three-dimensional body model for monitoring water percolation was constructed to perform forward simulations, ensuring that no boundaries influenced the simulation results. Then, the simulation results were compared with the corrected experimental data.

3.1. Effect of Glass Trough on Electrical Resistivity and Correction Equation

The sand in the glass trough in our laboratory experiment was macroscopically homogeneous. However, upon observing the 0 min (Figure 2) and 10 min to 60 min electrical resistivity profiles (Figure 5), two discrepancies from the actual conditions were evident. Firstly, there was a value significantly higher than the actual resistivity values. Secondly, the distribution of the electrical resistivity contour lines exhibited a predominantly parallel pattern. With increasing depth, the electrical resistivity gradually increased. These results can be attributed to the influence of the glass boundaries.

The correction equation applied to mitigate the influence of the narrow glass trough is as follows:

$${\mathsf{\rho }}_{\mathrm{n}}^{\ast }=\frac{{\mathsf{\rho }}_{\mathrm{n}}}{{\mathsf{\rho }}_0}\times {\mathsf{\rho }}_{\mathrm{b}}$$

(4)

where ${\mathsf{\rho }}_{\mathrm{n}}^{\ast }$ is the correction electrical resistivity after water percolation for n minutes; ${\mathsf{\rho }}_{\mathrm{n}}$ is the electrical resistivity after water percolation for n minutes; ${\mathsf{\rho }}_0$ is the electrical resistivity before water percolation; and ${\mathsf{\rho }}_{\mathrm{b}}$ is the background soil electrical resistivity.

Considering that the electrical resistivity value of the sand used in our experiment, measured under three-dimensional conditions, was 400–450 Ω·m, the background resistivity was assumed to be 400 Ω·m.

3.2. Numerical Simulation

In order to achieve a more realistic representation of the experiment and eliminate boundary effects, a numerical simulation was performed using three-dimensional forward modeling. The simulated detection area was a cubic volume measuring 1.55 m × 1.55 m × 1.55 m, with a minimum grid size of 0.01 m × 0.01 m × 0.01 m. The boundary conditions were implemented using a mixed boundary condition approach.

Based on the characteristics of water percolation in sand, four different wetting bulbs of varying sizes were selected to represent different stages of the percolation process. All models are shown above in Figure 2. In model 1, the wetting bulb exhibits a surface circle with a diameter Φ = 0.05 m, and the height from the maximum depth to the surface is h = 0.04 m. In model 2, the wetting bulb exhibits a surface circle with the diameter Φ = 0.1 m, and the height from the maximum depth to the surface is h = 0.09 m. In model 3, the wetting bulb exhibits a surface circle with the diameter Φ = 0.15 m, and the height from the maximum depth to the surface is h = 0.14 m. In model 4, the wetting bulb exhibits a surface circle with the diameter Φ = 0.22 m, and the height from the maximum depth to the surface is h = 0.22 m.

Considering that the sand used in the experiment had a resistivity value of 400–450 Ω·m, the background resistivity was assumed to be 400 Ω·m (Figure 3). The resistivity within the wetting area was set to 1 Ω·m (Figure 3).

Within the detection area, there was a measurement line (Figure 3). The measurement line consisted of 30 electrodes with a spacing of 0.05 m between adjacent electrodes. The measurement lines of the starting and ending points of L1 were (0.05 m, 0.75 m, 0 m) and (1.5 m, 0.70 m, 0 m).

Figure 4 displays the electrical resistivity profiles along the measurement line obtained through three-dimensional forward modeling using models 1 to 4. The resistivity profiles in the sand for models 1 to 4 exhibit areas of relatively low resistivity that correspond to the wetting zones specified in the models. There is some disparity in depth between models 1 and 2 compared with their intended depths. The depths determined by the 100 Ω·m contour line in the resistivity profiles of models 3 and 4 correspond to the actual depths of these models.

3.3. Comparison and Analysis between Correction Laboratory Experimental Results and Numerical Simulation Results

A comparison of the electrical resistivity profiles (Figure 4) obtained through forward modeling of models 1 to 4 (Figure 3) and the corrected electrical resistivity profiles from C-10 min to C-60 min, based on experimental data (Figure 5), shows that the contours of the profiles are very similar. In the correction experimental and numerical simulation electrical resistivity profiles, there is an area of relatively low resistivity in the center, surrounded by higher resistivity regions in every profile. The relatively low-resistivity anomaly in the center appears as a semi-ellipse shape, and the 400 Ω·m contour lines in the higher area exhibit similar patterns. Since it is not possible to perfectly replicate the model in the actual setup, the results cannot be identical. However, based on the comparison, it can be inferred that the method used to eliminate boundary effects has some effectiveness. The corrected resistivity profile can reflect the actual resistivity situation.

4. Results

4.1. Analysis of Electrical Resistivity Profiles and Photos

The 0 min profile (Figure 5a) served as the baseline for the electrical resistivity measurements. The distribution of electrical resistivity contour lines exhibits a predominantly parallel pattern. However, there are scattered regions with lower electrical resistivity values near the surface, primarily attributed to the presence of electrode coupling agents. With increasing depth, the electrical resistivity gradually increases. Despite the uniformity of the sand within the narrow glass trough measuring 10 cm, the measured electrical resistivity values vary. This variability can be attributed to the influence of the glass boundaries. As the detection depth increases, the impact of the boundaries on both sides of the glass trough becomes more significant, leading to a gradual increase in the electrical resistivity values. At the same depth, the measured electrical resistivity remains constant. The 0 min photograph depicts a transparent glass trough filled with uniformly distributed soil on a macroscopic scale. Since no percolation has occurred yet, there are no discernible wetting areas visible in the soil within the transparent glass.

In the 10 min to 40 min electrical resistivity profiles (Figure 5b–e), a distinct half-ellipse-shaped low-resistivity anomaly can be observed, extending horizontally from (0.65 m, 0.85 m) to (0.62 m, 0.92 m) and in a vertical depth range from (−0.05 m, −0.2 m) to (−0.05 m, −0.32 m). The center of the half ellipse exhibits the lowest electrical resistivity value. Moving away from the percolation point, the electrical resistivity gradually increases on both sides of the center. The 10 min and 30 min profiles show a symmetrical pattern on both sides. However, in the 20 min profile, the contour lines representing resistivity values ranging from 200 Ω·m to 750 Ω·m deviate to the left, and some curves come close to intersecting with others. Similarly, in the 40 min profile, the contour line representing resistivity values at 750 Ω·m also deviates to the left.

The p-10 min to p-40 min photographs (Figure 5b–e) depict a bulb-shaped wetting area extending horizontally from (0.65 m, 0.85 m) to (0.55 m, 0.92 m) and vertically from (−0.05 m, −0.2 m) to (−0.05 m, −0.35 m). The percolation point serves as the horizontal center of this area, while a vertical line passing through the percolation point divides the wetting area into two symmetrical halves. The wetting areas on the left and right sides exhibit symmetry in the p-10 min and p-30 min photos. However, in the p-20 min and p-40 min photographs, the wetting area on the left side is larger than that on the right side. The relatively low-resistivity area corresponding to the 2000 Ω·m contour line in the resistivity profile aligns with the wetting area captured in the photographs.

In the 50 min and 60 min electrical resistivity profiles (Figure 5f,g), a distinct half-ellipse-shaped low-resistivity anomaly is observed, spanning horizontally from 0.6 m to 0.92 m and vertically from −0.05 m to −0.35 m. As the distance from the center increases, the electrical resistivity gradually increases on both sides. The p-50 min and p-60 min photographs (Figure 5f,g) display a bulb-shaped wetting area, extending horizontally from 0.55 m to 0.92 m and vertically from 0 m to 0.5 m. The percolation point serves as the horizontal center of this area, while a vertical line passing through the percolation point divides the moist region into two symmetrical halves. In the resistivity profile, the relatively low-resistivity area corresponding to the 8000 Ω·m contour line aligns with the wetting area captured in the photograph. The resistivity values of 2000 Ω·m or 8000 Ω·m for relatively low-resistivity anomalies are significantly higher than the actual resistivity values of the soil due to the influence of boundaries. However, the low-resistivity anomaly areas determined by these values are located just below the percolation point and change with the percolation. These correspond to the wetting areas in the photographs, indicating that even though the resistivity value is high, it can still reflect the seepage process of water in a narrow glass trough.

4.2. Analysis of the Correction of Electrical Resistivity Profiles and Photographs

In Figure 5, the electrical resistivity profiles for C-10 min to C-60 min (Figure 5b–g) show half-elliptical-shaped areas of relatively low resistivity, while the corresponding photographs for P-10 min to P-60 min (Figure 5b–g) depict the wetting areas. The relatively low-resistivity areas are attributed to percolation effects. This speculation aligns perfectly with the wetting areas visible in the photographs. The low-resistivity anomalies enclosed by the 160 Ω·m contour line in the C-10 min resistivity profile correspond to the bulb-shaped wetting region in the P-10 min photograph. Similarly, the low-resistivity anomalies surrounded by the 120 Ω·m contour line in the C-20 min to C-50 min electrical resistivity profiles correspond to the bulb-shaped wetting areas in the respective photographs (P-20 min to P-50 min). However, in the C-60 min electrical resistivity profile, the 120 Ω·m contour line does not correspond to the bulb-shaped wetting area in the photograph (P-60 min) due to the influence of detection depth. The low-resistivity anomalies of 160 Ω·m or 120 Ω·m, generated due to water percolation, correspond to the locations of the wetting areas in the photo. This indicates that the photographs (P-10 min to P-50 min) can indeed validate the resistivity changes in the profiles (C-10 min to C-50 min) caused by water percolation.

5. Discussion

The electrical resistivity profiles used to monitor water percolation in sand within a narrow glass trough reveal a half-ellipse-shaped low-electrical-resistivity anomaly. Our experimental findings are similar to the results obtained in field studies on drip irrigation using ERT [8,33]. In both our experiment and the study conducted by Daniela, 2D electrical tomography was employed to monitor water percolation. The main difference between the two studies lies in the fact that the sand in our narrow glass trough represents an approximate two-dimensional body, whereas the other authors’ field study investigated a three-dimensional sand body. Nevertheless, considering the principle of similarity, our findings for the sand bodies within the narrow glass trough could be used to simulate relevant properties of the three-dimensional sand bodies found in the field.

The photograph captured from the front view of the narrow glass trough during water percolation confirms that the wetting pattern exhibits a bulb shape. This observation aligns with the simulation results reported by Al-Ogaidi [33]. In homogeneous soil in point source conditions, the wetting pattern typically takes on a bulb shape, appearing partially elliptical or partially circular when displayed in two dimensions. A comparison between the simulated data and the observed bulb shape in the photograph reveals that the simulated bulb shape is more regular, while the bulb shape in the photograph appears less regular. This disparity can be attributed to the fact that the simulated homogenous sand is in an ideal state, whereas the experimental sand demonstrates macroscopic homogeneity but microscopic non-homogeneity.

Based on the characteristics of seepage sources, seepage sources can be categorized into point sources, line sources, surface sources, and volumetric sources. Among these types, point sources are the simplest and serve as the foundation for investigating other seepage sources. In this study, to validate the proposed real-time monitoring and simultaneous verification of water percolation using ERT and photography techniques, we employed the most fundamental seepage source: the point source. Surface drip irrigation, widely applied in agricultural irrigation, is an example of point source utilization [34,35]. The insights gained through this study could be extended to subsurface drip irrigation and even line source drip irrigation [36,37]. In the future, the scope of this research could be expanded to investigate the widespread occurrence of surface or subsurface seepage in the natural environment. Examples include evaluating the seepage and stability of dams and embankments, studying contaminant transport, assessing slope stability, and various other applications. It can even provide inspiration for monitoring air seepage [38,39,40,41].

The 2000 Ω·m (Figure 5, 10 min to 40 min) and 8000 Ω·m (Figure 5, 50 min to 60 min) contour lines identifying the low-resistivity anomaly areas are notably higher than the actual resistivity values of the wetting sand. However, these results remain reliable. On one hand, the identified relatively low-resistivity areas correspond to the wetting regions in the photographs resulting from water percolation, indicating that the temporal and spatial variations in these low-resistivity areas can indeed reflect the percolation of water through the sand in the narrow glass trough. On the other hand, the corrected resistivity values (Figure 5, P-10 min to P-50 min) are within the range of resistivity for wetting sand. The shape of the low-resistivity anomaly area (Figure 5, P-10 min to P-60 min) closely aligns with the shape of the simulated relatively low-resistivity anomaly areas (Figure 4a–d) and corresponds to the photographs, confirming their capacity to accurately depict water percolation through the sand in the narrow glass trough. Our study has three limitations. Firstly, the wetting area observed in the photograph of the 10 cm narrow glass trough corresponds to the region identified based on the low-resistivity anomaly. However, we did not conduct research to determine the optimal width of the glass trough for the best alignment between the wetting areas, as determined through photography and using the low-resistivity anomaly area in the resistivity profile.

Secondly, the narrow glass trough represents a two-dimensional body when filled with sand, whereas most geological bodies are three-dimensional. We did not investigate the relationship between two-dimensional and three-dimensional bodies during water percolation in terms of ERT. Such research would enhance our understanding of water percolation in ERT and contribute to the development of relevant theories and practices through similarity numerical simulation experiments.

Finally, our study has demonstrated the real-time monitoring and simultaneous verification of water percolation employing electrical resistivity tomography and photography techniques. It is important to note that we have not considered certain hydrological parameters like flow velocity and the sand permeability coefficient in this study. Future research should delve into the influence of these hydrological parameters on the effectiveness of this method.

6. Conclusions

In order to visualize and verify the water percolation process simultaneously, research on water percolation in sandy soil within a narrow glass trough was conducted using ERT and photography techniques. The main conclusions are as follows:

(1)

The correction equation is suitable for mitigating the impact of the glass trough on the electrical resistivity value. The corrected electrical resistivity values exhibit a stronger alignment with reality, allowing for a more realistic representation of the sand soil’s electrical resistivity value.

(2)

An electrical resistivity contour line value can be used to delineate the water percolation area in the electrical resistivity profile and correct the electrical profile. The areas of relatively low electrical resistivity, delineated by the 2000 Ω·m or the 8000 Ω·m contour lines in the electrical resistivity profiles and the 160 Ω·m or the 120 Ω·m contour lines in the correction electrical resistivity profiles, exhibit a remarkable correspondence with the wetting area captured in the simultaneous photographs. While the values of 2000 Ω·m and 8000 Ω·m are influenced by the boundaries, the identified low-resistivity anomaly area, which changes over time, can reflect the process of water percolation.

(3)

The wetting pattern of point source water percolation in the sand was determined. The wetting area captured in the photograph exhibits a bulb-shaped pattern, while the low-resistivity anomaly detected using ERT has a half-ellipse shape.

(4)

The combined use of ERT and photography techniques allowed us to visualize and verify the water percolation process simultaneously. ERT enables one to visualize the water percolation process and identify both the temporal and spatial distribution of water percolation, while photography serves as a means of verification. This approach can be applied in order to physical model water percolation in agricultural irrigation, pollution studies, landslide assessments, etc.