ArticlePDF Available

Seasonal Groundwater Recharge Characterization using Time-Lapse Electrical Resistivity Tomography in Thepkasattri Watershed on Phuket Island, Thailand

Authors:

Abstract and Figures

Understanding the recharge mechanisms in the vadose zone is crucial to groundwater management and artificial recharge development. In this study, a systematic characterization of seasonal groundwater recharge was done using time-lapse electrical resistivity tomography (time-lapse ERT). The objective of this study was to characterize the seasonal groundwater recharge through the vadose zone and streams. A total of six electrical resistivity surveys in two locations were taken during the dry and rainy seasons using an advanced geosciences incorporated (AGI) SuperSting R2 resistivity meter in 2018. Then, time-lapse inversion was calculated using the dry season ERT as the base model and the rainy season ERTs as the monitoring datasets. The results showed a significant decrease in inverted resistivity from the dry season to the rainy season, which suggests rainwater infiltration through the vadose zone. Similarly, significant water level rise was observed in wells monitored during the survey indicating groundwater recharge. The time-lapse ERT showed, in one case, the Nang Dak stream and the unsaturated zones are the preferential groundwater recharge zones throughout the year; in another case, the Rieng stream is the groundwater discharge zone and the vadose zone is the preferential recharge zone. Finally, a simplified conceptual hydrogeological model representing the study area is presented to visualize the recharge mechanisms in the study area.
Content may be subject to copyright.
hydrology
Article
Seasonal Groundwater Recharge Characterization
Using Time-Lapse Electrical Resistivity Tomography
in the Thepkasattri Watershed on Phuket
Island, Thailand
Yacob T. Tesfaldet 1, 2, * and Avirut Puttiwongrak 1,3
1Interdisciplinary Graduate School of Earth System Science and Andaman Natural Disaster Management,
Prince of Songkla University, Phuket 83120, Thailand; avirut.p@phuket.psu.ac.th
2Department of Earth Sciences, Eritrea Institute of Technology, Asmara 12676, Eritrea
3Andaman Environment and Natural Disaster Research Center, Prince of Songkla University, Phuket 83120,
Thailand
*Correspondence: yacob.t.tesfaldet@phuket.psu.ac.th; Tel.: +66-0943-297-976
Received: 25 February 2019; Accepted: 1 May 2019; Published: 5 May 2019


Abstract:
Understanding the recharge mechanisms in the vadose zone is crucial to groundwater
management and artificial recharge development. In this study, a systematic characterization
of seasonal groundwater recharge was done using time-lapse electrical resistivity tomography
(time-lapse ERT). The objective of this study was to characterize the seasonal groundwater recharge
through the vadose zone and streams. A total of six electrical resistivity surveys in two locations were
taken during the dry and rainy seasons using an advanced geosciences incorporated (AGI) SuperSting
R2 resistivity meter in 2018. Then, time-lapse inversion was calculated using the dry season ERT as
the base model and the rainy season ERTs as the monitoring datasets. The results showed a significant
decrease in inverted resistivity from the dry season to the rainy season, which suggests rainwater
infiltration through the vadose zone. Similarly, significant water level rise was observed in wells
monitored during the survey indicating groundwater recharge. The time-lapse ERT showed, in one
case, the Nang Dak stream and the unsaturated zones are the preferential groundwater recharge zones
throughout the year; in another case, the Rieng stream is the groundwater discharge zone and the
vadose zone is the preferential recharge zone. Finally, a simplified conceptual hydrogeological model
representing the study area is presented to visualize the recharge mechanisms in the study area.
Keywords:
Time-lapse electrical resistivity tomography; vadose zone; preferential groundwater
recharge zones; inverted resistivity; Thepkasattri watershed
1. Introduction
Groundwater is the main source of fresh water all over the world and accounts for 30% of all
freshwater bodies [
1
]. To meet the increasing demand for fresh water, groundwater needs to be
replenished. Groundwater recharge is a hydrogeological phenomenon where water from rainfall, lakes,
snow melts, or streams moves downward to an aquifer and is stored as groundwater. There are two
types of groundwater recharge: diuse and focused. Diuse recharge comes from widely distributed
sources, for example a rainfall, while focused recharge comes from localized sources, for example
lakes, streams, or depressions [
2
]. The vadose zone is the soil layer that separates the ground surface
from the water saturated zone. Any water infiltration attributed to the subsurface passes through the
vadose zone before it reaches the water table (saturated zone) [
3
]. Thus, understanding the recharge
mechanisms in the vadose zone is crucial for developing and protecting groundwater resources.
Hydrology 2019,6, 36; doi:10.3390/hydrology6020036 www.mdpi.com/journal/hydrology
Hydrology 2019,6, 36 2 of 15
Time-lapse electrical resistivity tomography (time-lapse ERT) or four-dimensional (4D) geophysics
is a technique used in geoelectrical surveys to measure variations in resistivity (electrical properties of
the subsurface) over time at fixed locations [
4
]. It provides perception of the water movement of a
location by analyzing the sequence of resistivity distributions in the subsurface. The time-lapse ERT
survey requires periodic measurements at a fixed location with the same command file. Time-lapse ERT
utilizes the variations in inverted resistivity caused by water movement in the vadose zone. During the
dry season when rainfall is scarce, the vadose zone has high electrical resistance. On the other hand,
during the wet season when rainfall is abundant, the vadose zone has low electrical resistance due
to the infiltration of water (electrolytic conduction). It is a non-invasive, practical, and cost-eective
method for characterizing and delineating preferential recharge zones.
Time-lapse ERT has been applied in various studies involving arid and humid climates, and
yielded good results in delineating and characterizing seasonal diuse and focused recharges [
3
,
5
7
].
Descloitres et al. [
5
] used time-lapse ERT and magnetic resonance sounding to characterize seasonal
local recharge in the sub-humid tropical climate of western Ghats, India. The authors selected an
experimental watershed along a stream, and they succeeded in delineating preferential groundwater
recharge zones of the watershed and identifying the stream bed as the recharge zone for the aquifer
beneath. Mojica et al. [
7
] applied time-lapse ERT to study seasonal rainfall infiltration in the tropical
climate of the Gamboa area, Panama. Their study showed strong variation in resistivity between dry
and wet seasons. Based on the change in resistivity, the researchers delineated the superficial extension
of water content as well as the contribution of the stream near the study site. Time-lapse ERT coupled
with soil moisture measurement was also used to characterize the Dry Creek watershed, Idaho USA.
This research identified the direct association between conductivity and subsurface moisture content
which may be used to identify infiltration pathways in basement rocks [6].
Although many researchers have used time-lapse ERT for seasonal groundwater recharge
characterization on the same vadose zone under dierent climates, we have not encountered anyone
who applied time-lapse ERT on two geologically contrasting vadose zones under the same climate.
In this study, we used the later approach to study the vadose zone in the Thepkasattri watershed,
Phuket Island, Thailand. Our objective was to characterize seasonal groundwater recharge variation in
the vadose zone consisting of alluvium and weathered/fractured granite, and to evaluate the roles
of the nearby streams in groundwater recharge. Besides, we aimed to evaluate the applicability and
resolution power of time-lapse ERT in such a setting where the zones have the same climate but
dierent geology.
2. Materials and Methods
2.1. Site Description and Hydrogeology
The Thepkasattri watershed is located in the Thalang district of northeast Phuket Island, Thailand.
The watershed has an area of 54.82 km
2
and a catchment area of around 38.79 km
2
. Tributaries
originating from the surrounding mountain slopes converge on the plains of Thepkasattri with a
watershed in the middle of the floodplain, as shown in Figure 1b. The tributaries from the watershed
merge in the floodplain and form two main streams that merge at Choeng Thale estuary, finally
draining to the Andaman Sea.
Geologically, the watershed is bounded by Mesozoic intrusive rocks of Khao Prathiu Suite, and
by Kata Beach Suite granitic mountains in both the east and west. The flat plains are covered by
alluvial sediments underlined by the granitic bedrock [
8
]. The hydrogeology of the study area is
represented by an unconfined aquifer, which is further divided into two types: consolidated and
unconsolidated aquifers. The unconsolidated aquifer consists of the alluvial deposit (Quaternary)
called the Rayong-Satoon aquifer and is mainly found in the floodplains (central part) of the study area
(see Figure 12). The consolidated aquifer is composed of weathered and fractured granite basement
(Mesozoic). The stratigraphic section of the watershed shows that the thickness of the alluvial deposit
Hydrology 2019,6, 36 3 of 15
ranges from 0 m to 23 m and that of the weathered/fractured granite ranges from 10 m to 27 m.
The geological section of the area drawn from well log data is shown in Figure 2.
Hydrology 2019, 6, 36 3 of 15
Figure 1. Maps of the study area and groundwater flow directions. (a) Thailand map, the highlighted
red box shows Phuket Island; (b) Phuket Island, the red box indicates the location of the study area,
weather station and piezometer (P1: Wat Phra Thong, P2: College of Disaster Prevention and
Mitigation (CDPM); (c) Electrical resistivity tomography (ERT) survey lines (Line-1: alongside Nang
Dak stream and Line-2: alongside Rieng stream), boreholes (BH) and recharge model sections.
Geologically, the watershed is bounded by Mesozoic intrusive rocks of Khao Prathiu Suite, and
by Kata Beach Suite granitic mountains in both the east and west. The flat plains are covered by
alluvial sediments underlined by the granitic bedrock [8]. The hydrogeology of the study area is
represented by an unconfined aquifer, which is further divided into two types: consolidated and
unconsolidated aquifers. The unconsolidated aquifer consists of the alluvial deposit (Quaternary)
called the Rayong-Satoon aquifer and is mainly found in the floodplains (central part) of the study
area (see Figure 12). The consolidated aquifer is composed of weathered and fractured granite
basement (Mesozoic). The stratigraphic section of the watershed shows that the thickness of the
alluvial deposit ranges from 0 m to 23 m and that of the weathered/fractured granite ranges from 10
m to 27 m. The geological section of the area drawn from well log data is shown in Figure 2.
Based on groundwater specific capacity (SPC), data were computed from 504 wells; the highest
groundwater potential in Phuket is found in the alluvium aquifer of the Thepkasattri watershed [9].
The groundwater level in Thepkasattri ranges from 1 m to 12 m below ground level (mbgl). Well
hydrographs of the study area suggest that water table fluctuation follows the monthly variation of
rainfall. The water table rises to its peak during the wet season (May to October) when rainfall is
abundant, and it drops during the dry season (November to March) when rainfall is absent or scarce.
In Thepkasattri, groundwater flow directions follow the topography of the area as shown in Figure
1b. The overall groundwater flow is toward the valley and then finally drains to the Andaman Sea.
On Phuket Island, Thailand, freshwater demand is supplied mainly by groundwater wells (68% of
the total water supply) [10]. The proper utilization and protection of this resource requires
characterizing the recharge zones that depend on seasonal variations in rainfall. Furthermore,
Figure 1.
Maps of the study area and groundwater flow directions. (
a
) Thailand map, the highlighted
red box shows Phuket Island; (
b
) Phuket Island, the red box indicates the location of the study area,
weather station and piezometer (P1: Wat Phra Thong, P2: College of Disaster Prevention and Mitigation
(CDPM); (c) Electrical resistivity tomography (ERT) survey lines (Line-1: alongside Nang Dak stream
and Line-2: alongside Rieng stream), boreholes (BH) and recharge model sections.
Based on groundwater specific capacity (SPC), data were computed from 504 wells; the highest
groundwater potential in Phuket is found in the alluvium aquifer of the Thepkasattri watershed [
9
].
The groundwater level in Thepkasattri ranges from 1 m to 12 m below ground level (mbgl).
Well hydrographs of the study area suggest that water table fluctuation follows the monthly variation
of rainfall. The water table rises to its peak during the wet season (May to October) when rainfall is
abundant, and it drops during the dry season (November to March) when rainfall is absent or scarce.
In Thepkasattri, groundwater flow directions follow the topography of the area as shown in Figure 1b.
The overall groundwater flow is toward the valley and then finally drains to the Andaman Sea.
On Phuket Island, Thailand, freshwater demand is supplied mainly by groundwater wells (68% of the
total water supply) [
10
]. The proper utilization and protection of this resource requires characterizing
the recharge zones that depend on seasonal variations in rainfall. Furthermore, understanding the
recharge phenomena helps improve artificial recharge for groundwater sustainability.
Based on data from the years 1981–2010, The mean temperature in this area ranges from 26.7
C
to 28.6
C with maximum temperatures recorded in March and April. The mean total monthly amount
of rainfall, primarily caused by the Southeast monsoon, reaches up to 361 mm in September and
falls down to 24 mm in February as shown in the bar chart in Figure 4. The line plot in Figure 4
shows the total monthly rainfall for the study year (2018), which shows a similar rainfall pattern to the
Hydrology 2019,6, 36 4 of 15
average of the years 1981–2010 except for the high amount of rainfall in January (271 mm) and June
(459 mm) of the study year. Overall, in comparison to the mean of the years 1981–2010, the study
year was relatively wet. However, the seasonal variation of rainfall over the study year was within the
mean. Total mean annual rainfalls for the years 1981–2010 and the study year 2018 were 2219 mm and
2557 mm, respectively.
Hydrology 2019, 6, 36 4 of 15
understanding the recharge phenomena helps improve artificial recharge for groundwater
sustainability.
Figure 2. Geology of the subsurface layers from well log data. a.s.l. = above sea level.
Based on data from the years 1981–2010, The mean temperature in this area ranges from 26.7 °C
to 28.6 °C with maximum temperatures recorded in March and April. The mean total monthly
amount of rainfall, primarily caused by the Southeast monsoon, reaches up to 361 mm in September
and falls down to 24 mm in February as shown in the bar chart in Figure 4. The line plot in Figure 4
shows the total monthly rainfall for the study year (2018), which shows a similar rainfall pattern to
the average of the years 1981–2010 except for the high amount of rainfall in January (271 mm) and
June (459 mm) of the study year. Overall, in comparison to the mean of the years 1981–2010, the study
year was relatively wet. However, the seasonal variation of rainfall over the study year was within
the mean. Total mean annual rainfalls for the years 1981–2010 and the study year 2018 were 2219 mm
and 2557 mm, respectively.
2.2. Electrical Resistivity Survey and Data Inversion
The location of the geophysical survey lines is southwest of the study area which was chosen
based on the paths of the two streams crossing the study area. Each line is aligned at an angle with
its respective stream’s drainage direction to investigate recharge from the stream (Figure 1c).
The electrical resistivity tomography was carried out using an advanced geosciences
incorporated (AGI) SuperSting R2 with 56 electrodes arranged in a Wenner array. The Wenner array
is a symmetrical electrode configuration where the spacing between the four electrodes is equidistant
(Figure 3). The electrode spacing was 2.5 m in line one and 3 m in line two. The total profile lengths
were 137.5 m and 165 m respectively, for the two lines. The spacing was chosen based on the
availability of space and the desired depth of penetration. The details of the survey plan are given in
Table 1.
Figure 2. Geology of the subsurface layers from well log data. a.s.l. =above sea level.
2.2. Electrical Resistivity Survey and Data Inversion
The location of the geophysical survey lines is southwest of the study area which was chosen
based on the paths of the two streams crossing the study area. Each line is aligned at an angle with its
respective stream’s drainage direction to investigate recharge from the stream (Figure 1c).
The electrical resistivity tomography was carried out using an advanced geosciences incorporated
(AGI) SuperSting R2 with 56 electrodes arranged in a Wenner array. The Wenner array is a symmetrical
electrode configuration where the spacing between the four electrodes is equidistant (Figure 3).
The electrode spacing was 2.5 m in line one and 3 m in line two. The total profile lengths were 137.5 m
and 165 m respectively, for the two lines. The spacing was chosen based on the availability of space
and the desired depth of penetration. The details of the survey plan are given in Table 1.
Hydrology 2019, 6, 36 5 of 15
Figure 3. A schematic diagram of the Wenner array electrode configuration. ‘A’ and ‘B’ = current
electrodes (I = current), ‘M’ and ‘N’ = potential electrodes (V = potential difference) and ‘a’ = spacing
between electrodes.
Table 1. Electrical resistivity survey plan.
Date
(dd/mm/yyyy) Array Type Electrode Spacing (m) Total Length (m) Measurement Time
(min)
Line 1 Line 2 Line 1 Line 2 Line 1 Line 2
30/03/2018 Wenner 2
.
5 3 137
.
5 165 78 78
14/06/2018 Wenner 2
.
5 3 137
.
5 165 78 78
05/09/2018 Wenner 2
.
5 3 137
.
5 165 78 78
The timing of this electrical resistivity survey was chosen according to meteorological
information of the area (Figure 4). The dry season extends from October to March and the average
(1981–2010) amount of rainfall during these months was 116 mm. The wet season is from May to
September and the average rainfall was 254 mm. Considering the seasonal rainfall variabilities, the
first ERT was done during the dry season (in March), and the second and third ERTs were done
during the wet season in June and September, respectively.
Figure 4. Mean total monthly rainfall. The bar chart shows mean total monthly rainfall from 1981 to
2010, line plot shows total monthly rainfall for 2018 from a weather station in the study area. The ERT
survey months are indicated on the top of the relevant bars.
The data from the electrical resistivity survey were inverted using EarthImager 2D inversion
software developed by Advanced Geosciences Incorporated (AGI) (see Supplementary Materials). In
this study, the inversion model chosen for the measured resistivity value was the least squares
inversion method (L2). This model applies the L2 norm and minimizes the sum of squares of the
spatial changes in the model resistivity and the data misfits. Then, the inversion model minimizes
the root mean square error between the measured and inverted resistivity values [11]. Finally, the
time-lapse ERT inversion was carried out using the dry season survey datasets as the base model
Figure 3.
A schematic diagram of the Wenner array electrode configuration. ‘A’ and ‘B’ =current
electrodes (I =current), ‘M’ and ‘N’ =potential electrodes (V =potential dierence) and ‘a’ =spacing
between electrodes.
Hydrology 2019,6, 36 5 of 15
Table 1. Electrical resistivity survey plan.
Date
(dd/mm/yyyy)
Array
Type
Electrode Spacing (m) Total Length (m) Measurement Time (min)
Line 1 Line 2 Line 1 Line 2 Line 1 Line 2
30/03/2018 Wenner 2.5 3 137.5 165 78 78
14/06/2018 Wenner 2.5 3 137.5 165 78 78
05/09/2018 Wenner 2.5 3 137.5 165 78 78
The timing of this electrical resistivity survey was chosen according to meteorological information
of the area (Figure 4). The dry season extends from October to March and the average (1981–2010)
amount of rainfall during these months was 116 mm. The wet season is from May to September and
the average rainfall was 254 mm. Considering the seasonal rainfall variabilities, the first ERT was done
during the dry season (in March), and the second and third ERTs were done during the wet season in
June and September, respectively.
Figure 4.
Mean total monthly rainfall. The bar chart shows mean total monthly rainfall from 1981 to
2010, line plot shows total monthly rainfall for 2018 from a weather station in the study area. The ERT
survey months are indicated on the top of the relevant bars.
The data from the electrical resistivity survey were inverted using EarthImager 2D inversion
software developed by Advanced Geosciences Incorporated (AGI) (see Supplementary Materials).
In this study, the inversion model chosen for the measured resistivity value was the least squares
inversion method (L2). This model applies the L2 norm and minimizes the sum of squares of the
spatial changes in the model resistivity and the data misfits. Then, the inversion model minimizes
the root mean square error between the measured and inverted resistivity values [
11
]. Finally, the
time-lapse ERT inversion was carried out using the dry season survey datasets as the base model
(background data) for the wet season datasets recorded later. The output is presented as percentage
changes in resistivities from the base model.
In addition, water table fluctuation data from two piezometers installed in the study area were
used. The piezometers were installed in dierent locations, depths, and years. The first piezometer was
installed by the Department of Groundwater Resources (DGR) at Wat Phra Thong in 2007 but stopped
functioning in 2016, while the second piezometer was installed by Prince of Songkla University at
College of Disaster Prevention and Mitigation (CDPM) in May 2018.
Hydrology 2019,6, 36 6 of 15
2.3. A Conceptual Hydrogeological Model Visualization
A conceptual hydrogeological model was prepared to visualize the recharge mechanism in
the study area. The cross-section was drawn in an NW–SE direction across the two ERT profiles
(Figure 1c). The elevation data were taken from the Advanced Spaceborne Thermal Emission and
Reflection Radiometer (ASTER) Global Digital Elevation Model (GDEM) captured on 11 October 2011.
The hydrogeological information was analyzed from well log data and ERT results. The cross-section
was drawn using AutoCAD 2017 software.
3. Results and Discussion
3.1. Thepkasattri Line One alongside the Nang Dak Stream
The water table fluctuation measures the groundwater level in a monitoring well caused by
groundwater recharge or discharge. During the presence of rainfall, water level rise in an unconfined
aquifer is a response to change in storage due to water infiltration (recharge) [
12
]. On the other hand,
water level decline indicates discharge from an unconfined aquifer through pumping, base flow, etc.
Monitoring of water level fluctuation provides information on the dynamics of recharge on a local
scale that represents a spatial area of thousands of square meters [
2
]. In Thepkasattri, a piezometer
installed by DGR was used to observe the water table fluctuation pattern. The well hydrograph
was plotted against average monthly rainfall for 2015 (Figure 5). The data may not be up-to-date,
but the climatological information is very similar. In the study area, mean monthly temperature
and humidity do not deviate much. The total annual rainfall in 2015 and 2018 was 2472 mm and
2557 mm respectively.
Hydrology 2019, 6, 36 6 of 15
(background data) for the wet season datasets recorded later. The output is presented as percentage
changes in resistivities from the base model.
In addition, water table fluctuation data from two piezometers installed in the study area were
used. The piezometers were installed in different locations, depths, and years. The first piezometer
was installed by the Department of Groundwater Resources (DGR) at Wat Phra Thong in 2007 but
stopped functioning in 2016, while the second piezometer was installed by Prince of Songkla
University at College of Disaster Prevention and Mitigation (CDPM) in May 2018.
2.3. A Conceptual Hydrogeological Model Visualization
A conceptual hydrogeological model was prepared to visualize the recharge mechanism in the
study area. The cross-section was drawn in an NW–SE direction across the two ERT profiles (Figure
1c). The elevation data were taken from the Advanced Spaceborne Thermal Emission and Reflection
Radiometer (ASTER) Global Digital Elevation Model (GDEM) captured on 11 October 2011. The
hydrogeological information was analyzed from well log data and ERT results. The cross-section was
drawn using AutoCAD 2017 software.
3. Results and Discussion
3.1. Thepkasattri Line One alongside the Nang Dak Stream
The water table fluctuation measures the groundwater level in a monitoring well caused by
groundwater recharge or discharge. During the presence of rainfall, water level rise in an unconfined
aquifer is a response to change in storage due to water infiltration (recharge) [12]. On the other hand,
water level decline indicates discharge from an unconfined aquifer through pumping, base flow, etc.
Monitoring of water level fluctuation provides information on the dynamics of recharge on a local
scale that represents a spatial area of thousands of square meters [2]. In Thepkasattri, a piezometer
installed by DGR was used to observe the water table fluctuation pattern. The well hydrograph was
plotted against average monthly rainfall for 2015 (Figure 5). The data may not be up-to-date, but the
climatological information is very similar. In the study area, mean monthly temperature and
humidity do not deviate much. The total annual rainfall in 2015 and 2018 was 2472 mm and 2557 mm
respectively.
The water table fluctuation dynamics at Wat Phra Thong station follows the rainfall variation of
the study area (Figure 5). The hydraulic head reached the deepest level of 4.15 m between March and
April (dry season), then started to rise during the wet season and reached the peak (~1.65 m) between
July and September. The relationship between water level fluctuation and rainfall suggests the
dependency of recharge on the rainfall pattern of the study area. This relationship further implies
significant variations in resistivity of the subsurface during the onset of wet season as anticipated.
Figure 5.
A chart showing water table fluctuation versus the total monthly rainfall in 2015 from a
station at Wat Phra Thong. mbgl =meters below ground level.
The water table fluctuation dynamics at Wat Phra Thong station follows the rainfall variation
of the study area (Figure 5). The hydraulic head reached the deepest level of 4.15 m between March
and April (dry season), then started to rise during the wet season and reached the peak (~1.65 m)
between July and September. The relationship between water level fluctuation and rainfall suggests
the dependency of recharge on the rainfall pattern of the study area. This relationship further implies
significant variations in resistivity of the subsurface during the onset of wet season as anticipated.
The inversion of line one ERT (alongside Nang Dak stream) is characterized by three horizons
as shown in Figure 6. The first horizon represented by blue color has a resistivity range from
4.8
m to 13
m (low resistivity). It is associated with water flux (demarcated by black dotted
Hydrology 2019,6, 36 7 of 15
lines). In the tomography for March (Figure 6a), the low resistivity zone does not extend up to the
surface but vertically extends from 3.2 m to the bottom of the profile (demarcated by black dotted
lines). The absence of this horizon up to the surface is due to the lack of rain, but the horizon can
be distinguished at depth (marked by a horizontal black arrow) extending from the beginning of the
profile (0 m) where the Nang Dak stream passes to the distance of 54 m. The presence of this horizon
suggests that water infiltration comes from the Nang Dak stream which is located southeast of the
survey line (Figure 1c). In (b) and (c) of Figure 6, the low resistivity horizon extends up to the surface
(marked by a black arrow) which is due to superficial infiltration in response to rainfall from April
to September. The eect of the Nang Dak stream water infiltration is pronounced at this time, with
apparent inverted resistivity decrease (demarcated by a red dotted line). The decreasing resistivity
value implies that more water flux is coming from the Nang Dak stream to the aquifer.
Hydrology 2019, 6, 36 7 of 15
Figure 5. A chart showing water table fluctuation versus the total monthly rainfall in 2015 from a
station at Wat Phra Thong. mbgl = meters below ground level.
The inversion of line one ERT (alongside Nang Dak stream) is characterized by three horizons
as shown in Figure 6. The first horizon represented by blue color has a resistivity range from 4.8 Ωm
to 13 Ωm (low resistivity). It is associated with water flux (demarcated by black dotted lines). In the
tomography for March (Figure 6a), the low resistivity zone does not extend up to the surface but
vertically extends from 3.2 m to the bottom of the profile (demarcated by black dotted lines). The
absence of this horizon up to the surface is due to the lack of rain, but the horizon can be distinguished
at depth (marked by a horizontal black arrow) extending from the beginning of the profile (0 m)
where the Nang Dak stream passes to the distance of 54 m. The presence of this horizon suggests that
water infiltration comes from the Nang Dak stream which is located southeast of the survey line
(Figure 1c). In (b) and (c) of Figure 6, the low resistivity horizon extends up to the surface (marked
by a black arrow) which is due to superficial infiltration in response to rainfall from April to
September. The effect of the Nang Dak stream water infiltration is pronounced at this time, with
apparent inverted resistivity decrease (demarcated by a red dotted line). The decreasing resistivity
value implies that more water flux is coming from the Nang Dak stream to the aquifer.
The low resistivity zone located at a distance between 45 m and 75 m, and a depth from 6.4 m to
the 25.7 m (demarcated by yellow dotted lines in Figure 6b) was visible in the March profile,
disappeared in June and then re-emerged in September (Figure 6). The inconsistency of the June
profile at depth is associated with poorly fit datum points exterminated during ERT inversion (see
Supplementary Materials). The bad datum points are due to high contact resistance between the
electrode and ground or bad electrodes. The number of iterations used to fit measured and predicted
resistivity was eight. Fitting the measured resistivity with predicted resistivity was somewhat
difficult (Table 2).
Figure 6.
Electrical resistivity tomography of line one. The red arrow indicates the location of resistivity
log at 81 m distance: (
a
) 30 March 2018; (
b
) 14 June 2018; (
c
) 05 September 2018. RMS =root mean
squared error; L2-norm =the sum of the squared weighted data errors.
The low resistivity zone located at a distance between 45 m and 75 m, and a depth from 6.4 m to the
25.7 m (demarcated by yellow dotted lines in Figure 6b) was visible in the March profile, disappeared
in June and then re-emerged in September (Figure 6). The inconsistency of the June profile at depth
is associated with poorly fit datum points exterminated during ERT inversion (see Supplementary
Materials). The bad datum points are due to high contact resistance between the electrode and ground
or bad electrodes. The number of iterations used to fit measured and predicted resistivity was eight.
Fitting the measured resistivity with predicted resistivity was somewhat dicult (Table 2).
Hydrology 2019,6, 36 8 of 15
Table 2.
Electrical resistivity inversion statistics: number of iterations and root mean squared
(RMS) error.
Date
(dd/mm/yyyy) Array Iterations RMS (%)
Line 1 Line 2 Line 1 Line 2
30/03/2018 Wenner 6 3 3.00 2.50
14/06/2018 Wenner 8 3 3.67 2.51
05/09/2018 Wenner 7 4 3.20 2.61
The second major horizon represented by green and yellow colors has resistivity ranging from
20
m to 114
m (medium resistivity). This layer associated with alluvial deposits consists of
clayey sand and sandy gravel. The top part of this horizon has an undulating shape due to surficial
water infiltration.
The third horizon is the most resistive layer (demarcated by a blue dotted line) located at 97 m to
the end of the profile distance, and vertically from 6.2 m to 19.3 m (Figure 6). The resistivity of this layer
ranges from 200
m to 550
m. The resistive body located at distances from 105 m onwards might be
a 3D artifact associated with minor elevation change [
3
], or it might be due to dry or impermeable rock.
3.2. Thepkasattri Line Two Alongside the Rieng Stream
A piezometer coupled with rain gauge installed in May 2018 at CDPM station was used to observe
water table fluctuation pattern during the ERT survey time. The piezometer recorded water level
and rainfall on an hourly basis throughout the day. The well hydrograph along with the variation of
rainfall in the study area is shown in Figure 7. It is apparent from the figure that there is an association
between water level rise and the amount of rainfall. Water level started to rise in May, corresponding
to the onset of the rainy season reaching peak value of 3.6 m (bgl) at the end of the season. The peak
water level corresponds to the highest rainfall recorded in September and October. This association
further indicates that recharge is episodic and it depends on the amount of rainfall and the hydraulic
property of the vadose zone.
Hydrology 2019, 6, 36 8 of 15
Figure 6. Electrical resistivity tomography of line one. The red arrow indicates the location of
resistivity log at 81 m distance: (a) 30 March 2018; (b) 14 June 2018; (c) 05 September 2018. RMS = root
mean squared error; L2-norm = the sum of the squared weighted data errors.
The second major horizon represented by green and yellow colors has resistivity ranging from
20 Ωm to 114 Ωm (medium resistivity). This layer associated with alluvial deposits consists of clayey
sand and sandy gravel. The top part of this horizon has an undulating shape due to surficial water
infiltration.
The third horizon is the most resistive layer (demarcated by a blue dotted line) located at 97 m
to the end of the profile distance, and vertically from 6.2 m to 19.3 m (Figure 6). The resistivity of this
layer ranges from 200 Ωm to 550 Ωm. The resistive body located at distances from 105 m onwards
might be a 3D artifact associated with minor elevation change [3], or it might be due to dry or
impermeable rock.
3.2. Thepkasattri Line Two Alongside the Rieng Stream
A piezometer coupled with rain gauge installed in May 2018 at CDPM station was used to
observe water table fluctuation pattern during the ERT survey time. The piezometer recorded water
level and rainfall on an hourly basis throughout the day. The well hydrograph along with the
variation of rainfall in the study area is shown in Figure 7. It is apparent from the figure that there is
an association between water level rise and the amount of rainfall. Water level started to rise in May,
corresponding to the onset of the rainy season reaching peak value of 3.6 m (bgl) at the end of the
season. The peak water level corresponds to the highest rainfall recorded in September and October.
This association further indicates that recharge is episodic and it depends on the amount of rainfall
and the hydraulic property of the vadose zone.
Figure 7. A chart showing water table fluctuation versus average monthly rainfall in 2018 from a
station at CDPM.
The inversion result of line two alongside the Rieng stream shows that the tomography is
dominated by two horizons as shown in Figure 8. The low resistivity zone is represented by green
and irregularly distributed blue colors (demarcated by black dotted lines). The resistivity in this zone
ranges from 9.5 Ωm to 100 Ωm. The zone extends to a depth of 23.1 m at a 36 m distance, and from
that distance the layer pinches out. At distances of 139 m to 157 m, the layer was found only to a
depth of 7.7 m in all profiles. This zone is associated with the weathered part of granite bedrock
functioning as an aquifer. However, the zone slightly increased vertically in the June to September
profiles, and this change is associated with moisture movement in the vadose zone.
Figure 7.
A chart showing water table fluctuation versus average monthly rainfall in 2018 from a
station at CDPM.
The inversion result of line two alongside the Rieng stream shows that the tomography is
dominated by two horizons as shown in Figure 8. The low resistivity zone is represented by green
and irregularly distributed blue colors (demarcated by black dotted lines). The resistivity in this zone
ranges from 9.5
m to 100
m. The zone extends to a depth of 23.1 m at a 36 m distance, and from
Hydrology 2019,6, 36 9 of 15
that distance the layer pinches out. At distances of 139 m to 157 m, the layer was found only to a depth
of 7.7 m in all profiles. This zone is associated with the weathered part of granite bedrock functioning
as an aquifer. However, the zone slightly increased vertically in the June to September profiles, and
this change is associated with moisture movement in the vadose zone.
Hydrology 2019, 6, 36 9 of 15
Figure 8. Electrical resistivity tomography of line two. The red arrow indicates location of resistivity
log at 81m distance: (a) 30 March 2018; (b) 14 June 2018; (c) 05 September 2018.
The most conductive zone (dark to light blue color) with resistivities from 9.5 Ωm to 40 Ωm
shows variations in shape across the three dates (March, June, and September). The variation is
prominent in the June and September profiles as shown in the three tomographies (Figure 8,
demarcated by black dotted lines). The overall variation of the zone indicates water flux movement.
At the distance of 81 m (marked by red arrow), the zone extends from 3 m depth in March to around
12 m depth in September. The extension of the low resistivity indicates the vertical movement of
water in the vadose zone. Similarly, the low resistivity horizon shows variation between 18 m and 68
m distance (marked by inclined black arrows), indicating water infiltration both vertically and
horizontally. In this profile, there is a crescent shape conductive zone associated with preferential
flow from the vadose zone at 18 m marked by vertical black arrows in (b) and (c) of Figure 8. The
resistivity changes in the profiles indicate groundwater flow into the Rieng stream located on the left
parts of the profiles. In this area, the average groundwater level is 4.0 m (bgl), varying by about 1 m
below or above the Rieng stream bed, depending on the season. This implies that groundwater is
discharging as a base flow through the Rieng stream.
The second dominant zone represented by yellow/red color and demarcated by a blue dotted
line has resistivity range from 200 Ωm to 825 Ωm. This zone dominates the left side of the profile. The
horizon regularly extends from 40 m to 165 m. This zone is interpreted to represent granite bedrock
and can be easily identified in the well log data and resistivity log (Figure 2 and Figure 9b).
Figure 8.
Electrical resistivity tomography of line two. The red arrow indicates location of resistivity
log at 81m distance: (a) 30 March 2018; (b) 14 June 2018; (c) 05 September 2018.
The most conductive zone (dark to light blue color) with resistivities from 9.5
m to 40
m shows
variations in shape across the three dates (March, June, and September). The variation is prominent in
the June and September profiles as shown in the three tomographies (Figure 8, demarcated by black
dotted lines). The overall variation of the zone indicates water flux movement. At the distance of
81 m (marked by red arrow), the zone extends from 3 m depth in March to around 12 m depth in
September. The extension of the low resistivity indicates the vertical movement of water in the vadose
zone. Similarly, the low resistivity horizon shows variation between 18 m and 68 m distance (marked
by inclined black arrows), indicating water infiltration both vertically and horizontally. In this profile,
there is a crescent shape conductive zone associated with preferential flow from the vadose zone at
18 m marked by vertical black arrows in (b) and (c) of Figure 8. The resistivity changes in the profiles
indicate groundwater flow into the Rieng stream located on the left parts of the profiles. In this area,
the average groundwater level is 4.0 m (bgl), varying by about 1 m below or above the Rieng stream
bed, depending on the season. This implies that groundwater is discharging as a base flow through the
Rieng stream.
The second dominant zone represented by yellow/red color and demarcated by a blue dotted
line has resistivity range from 200
m to 825
m. This zone dominates the left side of the profile.
The horizon regularly extends from 40 m to 165 m. This zone is interpreted to represent granite
bedrock and can be easily identified in the well log data and resistivity log (Figures 2and 9b).
Hydrology 2019,6, 36 10 of 15
Hydrology 2019, 6, 36 10 of 15
Figure 9. Showing the resistivity log extracted at an 81 m distance (Figure 6 and Figure 8). Top soil is
characterized by root and soil interaction and the active zone is characterized by uniform resistivity
change. (a) Resistivity log of line one; (b) Resistivity log of line two.
The use of resistivity log is helpful in outlining the geologic layers and active zones characterized
by major resistivity change in monitoring datasets [13]. As Alamry et al. [13] reported the top soil is
characterized by uneven resistivity distribution indicating vegetation soil interaction, while the active
zone represents uniform resistivity change throughout the survey dates, indicating water movement.
Therefore, a resistivity log at a distance of 81 m (marked by the red arrow in Figure 6 and Figure 8)
was extracted to confirm the geological layers and demarcate the active zone represented by
significant variations in resistivity. The resistivity log of line one is shown in Figure 9a; the top soil is
characterized by variable resistivity associated with the variability of water saturation due to
evapotranspiration. The thickness of this layer is around 5 m. The second layer is the active zone
characterized by major resistivity change among the three dates. This layer has a thickness of 17 m,
and beyond that depth, the resistivity log does not show a significant difference among the three
dates.
The resistivity log for line two at a distance of 81 m is shown in Figure 9b. The top soil
represented by variable resistivity value has a thickness of around 3 m. The active zone is much
thinner in line two than line one, which is due to the high permeability, and the porosity zone is
thicker in line one. In both resistivity logs, the layer below the active zone is characterized by almost
no resistivity change, indicating the bedrock.
Table 2. Electrical resistivity inversion statistics: number of iterations and root mean squared (RMS)
error.
Date
(dd/mm/yyyy) Array Iterations RMS (%)
Line 1 Line 2 Line 1 Line 2
30/03/2018 Wenner 6 3 3.00 2.50
14/06/2018 Wenner 8 3 3.67 2.51
05/09/2018 Wenner 7 4 3.20 2.61
3.3. Line One Time-Lapse ERT
Time-lapse inversion was done using EarthImager 2D 2.4.4 following the standard procedures
in the manual [14]. The Smooth model inversion was used for all time-lapse ERT inversions. The
time-lapse ERT data were inverted by setting the dry season profile as the base model for the
Figure 9.
Showing the resistivity log extracted at an 81 m distance (Figures 6and 8). Top soil is
characterized by root and soil interaction and the active zone is characterized by uniform resistivity
change. (a) Resistivity log of line one; (b) Resistivity log of line two.
The use of resistivity log is helpful in outlining the geologic layers and active zones characterized
by major resistivity change in monitoring datasets [
13
]. As Alamry et al. [
13
] reported the top soil is
characterized by uneven resistivity distribution indicating vegetation soil interaction, while the active
zone represents uniform resistivity change throughout the survey dates, indicating water movement.
Therefore, a resistivity log at a distance of 81 m (marked by the red arrow in Figures 6and 8) was
extracted to confirm the geological layers and demarcate the active zone represented by significant
variations in resistivity. The resistivity log of line one is shown in Figure 9a; the top soil is characterized
by variable resistivity associated with the variability of water saturation due to evapotranspiration.
The thickness of this layer is around 5 m. The second layer is the active zone characterized by major
resistivity change among the three dates. This layer has a thickness of 17 m, and beyond that depth,
the resistivity log does not show a significant dierence among the three dates.
The resistivity log for line two at a distance of 81 m is shown in Figure 9b. The top soil represented
by variable resistivity value has a thickness of around 3 m. The active zone is much thinner in line
two than line one, which is due to the high permeability, and the porosity zone is thicker in line one.
In both resistivity logs, the layer below the active zone is characterized by almost no resistivity change,
indicating the bedrock.
3.3. Line One Time-Lapse ERT
Time-lapse inversion was done using EarthImager 2D 2.4.4 following the standard procedures in the
manual [
14
]. The Smooth model inversion was used for all time-lapse ERT inversions. The time-lapse
ERT data were inverted by setting the dry season profile as the base model for the following June and
September profiles, as recommended by Miller et al. [
6
]. Finally, the inversion produces a percentage
dierence in resistivity between the base and monitoring datasets.
Time-lapse ERT inversion of line one from March to June and March to September are shown
in Figure 10. On the right side of both profiles, up to a distance of 135 m, there are strong resistivity
decreases from
50% to
100% (demarcated by black dotted lines). The strong resistivity change
extends from the surface to the bottom of the profile. The change implies water infiltration through the
vadose zone during the rainy season; the percentage variation in resistivity indirectly indicates the
volume of water occupying the pore spaces [
6
]. The consistency of the time-lapse ERT in both dates
indicates infiltration from the surface, and the Nang Dak stream is the preferential recharge zone for the
Hydrology 2019,6, 36 11 of 15
aquifer. The preferential recharge zones are regularly connected with the bottom aquifer throughout
the profile, forming funneled flow. Superficially continuous resistivity decrease throughout the profile
in the sedimentary vadose zone is also reported by Mojica et al. and Zhang et al. [
7
,
15
]. The nature
of the preferential flow suggests the hydraulic heterogeneity of the alluvial deposit. However, in
(Figure 10b) the preferential flow occurs as patches to a depth of 9.6 m. This irregularity is associated
with artifacts caused by the relocation of electrodes and bad datum points.
Hydrology 2019, 6, 36 11 of 15
following June and September profiles, as recommended by Miller et al. [6]. Finally, the inversion
produces a percentage difference in resistivity between the base and monitoring datasets.
Time-lapse ERT inversion of line one from March to June and March to September are shown in
Figure 10. On the right side of both profiles, up to a distance of 135 m, there are strong resistivity
decreases from 50% to 100% (demarcated by black dotted lines). The strong resistivity change
extends from the surface to the bottom of the profile. The change implies water infiltration through
the vadose zone during the rainy season; the percentage variation in resistivity indirectly indicates
the volume of water occupying the pore spaces [6]. The consistency of the time-lapse ERT in both
dates indicates infiltration from the surface, and the Nang Dak stream is the preferential recharge
zone for the aquifer. The preferential recharge zones are regularly connected with the bottom aquifer
throughout the profile, forming funneled flow. Superficially continuous resistivity decrease
throughout the profile in the sedimentary vadose zone is also reported by Mojica et al. and Zhang et
al. [7,15]. The nature of the preferential flow suggests the hydraulic heterogeneity of the alluvial
deposit. However, in (Figure 10b) the preferential flow occurs as patches to a depth of 9.6 m. This
irregularity is associated with artifacts caused by the relocation of electrodes and bad datum points.
Figure 10. Time-lapse inversion of line one with root mean squared (RMS) error: (a) 30 March 2018 to
14 June 2018; (b) 30 March 2018 to 05 September 2018.
The resistive layer represented by the yellow to red color and demarcated by the brown dotted
line shows a strong resistivity increase by 50% to 100%. In Figure 10a, a resistive body is located at
distances from 48 m to 72 m and vertically from 12.8 m to the bottom of the profile (demarcated by
brown dotted lines). This layer later disappeared in the March–September time-lapse inversion
(Figure 10b). This implies that the resistive layer is caused by bad datasets in the June profile as
described in Section 3.1. In addition, in Figure 10a, an increase in resistivity is shown at a 75 m
distance and between 1 m to 6.4 m deep, and, in Figure 10b, irregularly distributed from 22 m to 97
m distances and 9.6 m deep. The increase in resistivity is associated with a time-lapse inversion
artifact caused by moisture contrast between the dry and wet season. The inversion artifacts are false
results caused by the inversion model’s incapability to fit the contrasting resistivity values during
time-lapse inversion [4]. In addition, there is a resistivity decrease at depth in both time-lapse ERT
that might be due to enrichment in the ionic concentration of groundwater [16]. On the other hand,
it might be due to the effect of the overlying material electrical property or poor model resolution at
depth [15–18].
3.4. Line Two Time-Lapse ERT
Figure 10.
Time-lapse inversion of line one with root mean squared (RMS) error: (
a
) 30 March 2018 to
14 June 2018; (b) 30 March 2018 to 05 September 2018.
The resistive layer represented by the yellow to red color and demarcated by the brown dotted
line shows a strong resistivity increase by 50% to 100%. In Figure 10a, a resistive body is located
at distances from 48 m to 72 m and vertically from 12.8 m to the bottom of the profile (demarcated
by brown dotted lines). This layer later disappeared in the March–September time-lapse inversion
(Figure 10b). This implies that the resistive layer is caused by bad datasets in the June profile as
described in Section 3.1. In addition, in Figure 10a, an increase in resistivity is shown at a 75 m
distance and between 1 m to 6.4 m deep, and, in Figure 10b, irregularly distributed from 22 m to 97 m
distances and 9.6 m deep. The increase in resistivity is associated with a time-lapse inversion artifact
caused by moisture contrast between the dry and wet season. The inversion artifacts are false results
caused by the inversion model’s incapability to fit the contrasting resistivity values during time-lapse
inversion [
4
]. In addition, there is a resistivity decrease at depth in both time-lapse ERT that might be
due to enrichment in the ionic concentration of groundwater [
16
]. On the other hand, it might be due
to the eect of the overlying material electrical property or poor model resolution at depth [1518].
3.4. Line Two Time-Lapse ERT
Time-lapse ERT inversions of line two data for March to June and March to September comparisons
are shown in Figure 11. These profiles are dominated by less than 75% resistivity decreases demarcated
by black dotted lines. These strong resistivity changes are associated with the first rainfall in the
rainy season. In Figure 11a, the resistivity decrease is attributed to preferential recharge zones located
between 27 m and 122 m distance. Similarly, the consistency of those preferential recharge zones is
observed in the March to September time-lapse inversion as well. The presence of consistent preferential
recharge zones supports the concept that recharge in weathered/fractured granite is mainly controlled
by slowly moving wetting fronts and well-connected fractures [
19
]. The top part of the tomography
is characterized by unstable flow concentrated into conductive fingers in March to June time-lapse
ERT (Figure 11a). Then some conductive fingers become preferential recharge zones (marked by black
Hydrology 2019,6, 36 12 of 15
arrow at X =27 m, 54 m, 63 m, 81 m, 104 m, 135 m and 157 m) in the March to September time-lapse
ERT (Figure 11b). At an 81 m distance, the wetting front is penetrating vertically downward to a depth
of around 29 m, indicating a well-connected fracture zone. Similarly, in all distances marked by a
black arrow, the infiltrated water follows uniform pathways (preferential recharge zones) to reach the
other parts of the profile. This implies structurally controlled infiltration pathways typical of fractured
hard rocks [
18
20
]. Robert et al. [
21
] also found structurally controlled preferential recharge zones
in an investigation done in fractured/karstified limestones (southern Belgium). The study was done
by injecting salt tracer to map flow paths using time-lapse ERT. The complicated preferential flow in
the vadose zone is associated with textural heterogeneity and hydraulic properties of the weathered
granite layer [22].
Hydrology 2019, 6, 36 12 of 15
Time-lapse ERT inversions of line two data for March to June and March to September
comparisons are shown in Figure 11. These profiles are dominated by less than 75% resistivity
decreases demarcated by black dotted lines. These strong resistivity changes are associated with the
first rainfall in the rainy season. In Figure 11a, the resistivity decrease is attributed to preferential
recharge zones located between 27 m and 122 m distance. Similarly, the consistency of those
preferential recharge zones is observed in the March to September time-lapse inversion as well. The
presence of consistent preferential recharge zones supports the concept that recharge in
weathered/fractured granite is mainly controlled by slowly moving wetting fronts and well-
connected fractures [19]. The top part of the tomography is characterized by unstable flow
concentrated into conductive fingers in March to June time-lapse ERT (Figure 11a). Then some
conductive fingers become preferential recharge zones (marked by black arrow at X = 27 m, 54 m, 63
m, 81 m, 104 m, 135 m and 157 m) in the March to September time-lapse ERT (Figure 11b). At an 81
m distance, the wetting front is penetrating vertically downward to a depth of around 29 m,
indicating a well-connected fracture zone. Similarly, in all distances marked by a black arrow, the
infiltrated water follows uniform pathways (preferential recharge zones) to reach the other parts of
the profile. This implies structurally controlled infiltration pathways typical of fractured hard rocks
[18–20]. Robert et al. [21] also found structurally controlled preferential recharge zones in an
investigation done in fractured/karstified limestones (southern Belgium). The study was done by
injecting salt tracer to map flow paths using time-lapse ERT. The complicated preferential flow in the
vadose zone is associated with textural heterogeneity and hydraulic properties of the weathered
granite layer [22].
Figure 11. Time-lapse inversion of line two with root mean squared (RMS) error: (a) 30 March 2018 to
14 June 2018; (b) 30 March 2018 to 05 September 2018.
The resistivity increased from 5% to 50% in a layer that mainly occurs at the top part of the
profile and the left side of the profile from a distance of 9 m to 40 m and at depths from 3 m to 21 m.
This layer is associated with a low permeability body that is part of the granite bedrock. Then, later
on, the time-lapse ERT of March to September shows that the layer seems to decrease vertically by 4
m due to deep percolation of water. Furthermore, there is a resistive body characterized by 70% to
100% resistivity increase found between the top surface and 11 m depth (demarcated by a brown
dotted lines). This resistive layer is an inversion artifact associated with changes in moisture content
or with the relocation of electrodes during those dates [4,6,7].
The resistivity variations between dry and wet seasons are much higher for line one than for line
two. There was around 100% resistivity decrease in line one, while in line two there was a 75%
decrease. Based on petrophysical transformation, the change in resistivity (electrical conductivity)
Figure 11.
Time-lapse inversion of line two with root mean squared (RMS) error: (
a
) 30 March 2018 to
14 June 2018; (b) 30 March 2018 to 05 September 2018.
The resistivity increased from 5% to 50% in a layer that mainly occurs at the top part of the profile
and the left side of the profile from a distance of 9 m to 40 m and at depths from 3 m to 21 m. This layer
is associated with a low permeability body that is part of the granite bedrock. Then, later on, the
time-lapse ERT of March to September shows that the layer seems to decrease vertically by 4 m due
to deep percolation of water. Furthermore, there is a resistive body characterized by 70% to 100%
resistivity increase found between the top surface and 11 m depth (demarcated by a brown dotted
lines). This resistive layer is an inversion artifact associated with changes in moisture content or with
the relocation of electrodes during those dates [4,6,7].
The resistivity variations between dry and wet seasons are much higher for line one than for
line two. There was around 100% resistivity decrease in line one, while in line two there was a 75%
decrease. Based on petrophysical transformation, the change in resistivity (electrical conductivity) can
be related to the degree of water saturation. As the Miller et al. [
6
] study in Dry Creek watershed
has shown, an increase in electrical conductivity corresponds to an increase in water saturation.
Therefore, it can be inferred that the degrees of saturation in both locations are dierent. Geologically,
the vadose zone in line one is an alluvial deposit and is weathered/fractured granite in line two
(Figure 2). This indicates that in weathered and fractured granite, the recharge is controlled by slowly
draining wetting fronts (flow paths) as revealed in the three tomograms and time-lapse ERT (see
Figures 8and 11). Structurally controlled wetting fronts are also identified in hard rocks by previous
researchers [
3
,
19
]. As
Arora et al. [3]
reported, in vadose zones consisting of fractured/weathered
Hydrology 2019,6, 36 13 of 15
granite, water movement is both horizontal and vertical. The result was further confirmed by
self-potential survey and in situ soil moisture profiling.
Furthermore, the interannual change in the well hydrograph correlates considerably with
interannual rainfall pattern of the study area. In addition, time-lapse ERT shows strong seasonal
resistivity variation in the vadose zone, indicating water infiltration during the rainy season.
The relationship suggests that groundwater recharge is dependent on the annual rainfall pattern of
the study area [
23
,
24
]. As Wu et al. [
23
] investigated, the relationship between annual recharge and
rainfall is significant for shallow groundwater depths, which is the case in the Thepkasattri watershed.
3.5. A Conceptual Hydrogeological Model
Based on the hydrogeological information of the study area and the time-lapse ERT result, a
conceptual hydrogeological model was drawn along an NW–SE cross-section to visualize the recharge
mechanism in the study area (Figure 1c). The hydrogeological information was taken from groundwater
data provided by DGR, ERT results, and field observation. The thickness of the geological layers is not
to scale; it was inferred based on well log and field observation. Then the available information was
used to make the conceptual hydrogeological model using AutoCAD 2017 software (Figure 12).
Hydrology 2019, 6, 36 13 of 15
can be related to the degree of water saturation. As the Miller et al. [6] study in Dry Creek watershed
has shown, an increase in electrical conductivity corresponds to an increase in water saturation.
Therefore, it can be inferred that the degrees of saturation in both locations are different. Geologically,
the vadose zone in line one is an alluvial deposit and is weathered/fractured granite in line two
(Figure 2) . This indicates that in weathered and fractured granite, the recharge is controlled by slowly
draining wetting fronts (flow paths) as revealed in the three tomograms and time-lapse ERT (see
Figures 8 and 11). Structurally controlled wetting fronts are also identified in hard rocks by previous
researchers [3,19]. As Arora et al. [3] reported, in vadose zones consisting of fractured/weathered
granite, water movement is both horizontal and vertical. The result was further confirmed by self-
potential survey and in situ soil moisture profiling.
Furthermore, the interannual change in the well hydrograph correlates considerably with
interannual rainfall pattern of the study area. In addition, time-lapse ERT shows strong seasonal
resistivity variation in the vadose zone, indicating water infiltration during the rainy season. The
relationship suggests that groundwater recharge is dependent on the annual rainfall pattern of the
study area [23,24]. As Wu et al. [23] investigated, the relationship between annual recharge and
rainfall is significant for shallow groundwater depths, which is the case in the Thepkasattri
watershed.
3.5. A Conceptual Hydrogeological Model
Based on the hydrogeological information of the study area and the time-lapse ERT result, a
conceptual hydrogeological model was drawn along an NW–SE cross-section to visualize the
recharge mechanism in the study area (Figure 1c). The hydrogeological information was taken from
groundwater data provided by DGR, ERT results, and field observation. The thickness of the
geological layers is not to scale; it was inferred based on well log and field observation. Then the
available information was used to make the conceptual hydrogeological model using AutoCAD 2017
software (Figure 12).
Figure 12. A simplified conceptual hydrogeological model of the study area.
A conceptual hydrogeological model of the study area shows the topographic highs including
the stream passing nearby line one are groundwater recharge areas. Generally, the topographic highs
in the study have a gentle slope favorable for diffuse recharge [2]. The topographic lows near line
two are groundwater discharge zones as base flow.
Seasonal groundwater recharge in the Thepkasatri watershed varies based on the geology of the
vadose zone and rainfall pattern. The time-lapse ERT in both locations further suggests that water
movement is mainly controlled by hydraulic heterogeneity and fracture pattern of the vadose zone.
4. Conclusions
Time-lapse ERT coupled with well hydrograph was applied to investigate the seasonal
groundwater recharge variations in the Thepkasattri watershed. Time-lapse ERT is a non-destructive
method and provides information on spatial and temporal variations of water infiltration in the
vadose zone. Water table fluctuation observation and well log data were used to determine recharge
variation throughout the seasons and provide subsurface geological information, respectively. The
Figure 12. A simplified conceptual hydrogeological model of the study area.
A conceptual hydrogeological model of the study area shows the topographic highs including the
stream passing nearby line one are groundwater recharge areas. Generally, the topographic highs in
the study have a gentle slope favorable for diuse recharge [
2
]. The topographic lows near line two are
groundwater discharge zones as base flow.
Seasonal groundwater recharge in the Thepkasatri watershed varies based on the geology of the
vadose zone and rainfall pattern. The time-lapse ERT in both locations further suggests that water
movement is mainly controlled by hydraulic heterogeneity and fracture pattern of the vadose zone.
4. Conclusions
Time-lapse ERT coupled with well hydrograph was applied to investigate the seasonal groundwater
recharge variations in the Thepkasattri watershed. Time-lapse ERT is a non-destructive method and
provides information on spatial and temporal variations of water infiltration in the vadose zone.
Water table fluctuation observation and well log data were used to determine recharge variation
throughout the seasons and provide subsurface geological information, respectively. The integration
of hydrogeological and geophysical data in the study area was useful for ERT interpretation.
The time-lapse ERT method also has limitations that require care during data interpretations. These are
artifact generation during inversion and relocation of electrodes [
4
,
6
]. The well hydrograph of the area
indicates recharge is seasonal and dependent on the rainfall pattern. Similarly, the seasonal variations
in groundwater recharge (moisture content) in the time-lapse ERT are expressed as dierences in
resistivity between the base model and the monitoring datasets. In line one and line two of the
tomography, strong seasonal resistivity variations were observed in the vadose zone. The maximum
resistivity decrease in line one reaches around 100%, while in line two it was 75% or less. The decrease
in resistivity is strongly related to infiltration from rainfall during the rainy season. Based on the
Hydrology 2019,6, 36 14 of 15
water table fluctuation datasets and time-lapse ERT, the aquifer replenishment in the study area is
mainly caused by superficial water infiltration that depends on the rainfall pattern of the study area.
This study demonstrated applicability and utilization of time-lapse ERT method to visualize moisture
variation in soil and to identify preferential recharge zones. The well-resolved time-lapse images in
two contrasting unsaturated zones further indicate the potential of the method to provide noninvasive
spatiotemporal moisture variation in the vadose zone.
Supplementary Materials:
The Supplementary Materials are available online at http://www.mdpi.com/2306-5338/
6/2/36/s1.
Author Contributions:
Y.T.T.: project administration, methodology, investigation, and writing. A.P. review,
editing, and supervision.
Funding:
The authors would like to thank the Thailand International Cooperation Agency (TICA) for providing
financial support to carry out the study.
Acknowledgments:
The groundwater data and well log data were provided by the Department of Groundwater
Resources (DGR) of Thailand. The authors would like to thank Ratha Men, Vann Sakannan, Naung Naung Htwe
and Mai Khanh Phoung for helping in the ERT survey. Also, we thank Seppo Karrila and Tanwa Arpornthip for
comments and suggestions on the paper.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Williams, M. What Percent of Earth is Water? Available online: https://www.universetoday.com/65588/what-
percent-of-earth-is-water/(accessed on 3 March 2018).
2.
Healy, R. Estimating Groundwater Recharge, 1st ed.; Cambridge University Press: Cambridge, UK, 2012;
pp. 1–13.
3.
Arora, T.; Ahmed, S. Characterization of recharge through complex vadose zone of a granitic aquifer by
time-lapse electrical resistivity tomography. J. Appl. Geophys. 2011,73, 35–44. [CrossRef]
4.
Carey, A.M.; Paige, G.B.; Carr, B.J.; Dogan, M. Forward modeling to investigate inversion artifacts resulting
from time-lapse electrical resistivity tomography during rainfall simulations. J. Appl. Geophys.
2017
,145,
39–49. [CrossRef]
5.
Descloitres, M.; Ruiz, L.; Sekhar, M.; Legchenko, A.; Braun, J.J.; Kumar, M.S.; Subramanian, S. Characterization
of seasonal local recharge using electrical resistivity tomography and magnetic resonance sounding. Hydrol.
Process. 2007,22, 384–394. [CrossRef]
6.
Miller, C.R.; Routh, P.S.; Brosten, T.R.; McNamara, J.P. Application of time-lapse ERT imaging to watershed
characterization. Geophysics 2008,73, G7–G17. [CrossRef]
7.
Mojica, A.; Diaz, I.; Ho, C.A.; Ogden, F.; Pinzon, R.; Fabrega, J.; Vega, D.; Hendrickx, J. Study of seasonal
rainfall infiltration via time-lapse surface electrical resistivity tomography: Case study of Gamboa area,
Panama Canal Watershed. Air Soil Water Res. 2013,6, 131–139. [CrossRef]
8.
Hummel, C.L.; Phawandon, P. Geology and Mineral Deposits of the Phuket Mining District, South Thailand;
Department of Mineral Resources: Bangkok, Thailand, 1967.
9.
Charoenpong, S.; Suwanprasit, P.; Thongchumnum, P. Impacts of interpolation techniques on groundwater
potential modeling using GIS in Phuket Province, Thailand. In Proceedings of the 33rd ACRS, Pattaya,
Thailand, 26–30 November 2012; pp. 732–738.
10.
Kong, S.O. Hydrogeological Characterization of Phuket Aquifer System with Reference to Groundwater
Development of Phuket, Thailand. Master’s Thesis, Prince Songkla University, Phuket, Thailand, 2017.
11. Loke, M.H.; Acworth, I.; Dahlin, T. A comparison of smooth and blocky inversion methods in 2D electrical
imaging method. Explor. Geophys. 2003,34, 182–187. [CrossRef]
12.
Healy, R.; Cook, P. Using groundwater levels to estimate recharge. Hydrogeol. J.
2002
,10, 91–109. [CrossRef]
13.
Alamry, A.S.; van der Meijde, M.; Noomen, M.; Addink, E.A.; van Benthem, R.; de Jong, S.M. Spatial and
temporal monitoring of soil moisture using surface electrical resistivity tomography in Mediterranean soils.
Catena 2017,157, 388–396. [CrossRef]
14.
Advanced Geoscience Inc. Instruction Manual for EarthImager 2D Version 2.4.0; Advanced Geoscience Inc.:
Austin, TX, USA, 2014; Vol. 2.4.0.
Hydrology 2019,6, 36 15 of 15
15.
Zhang, G.; Zhang, G.; Chen, C.; Chang, P.; Wang, T.; Yen, H.; Dong, J.; Ni, C.; Chen, S.; Chen, C.; et al.
Imaging rainfall infiltration process with the time-lapse electrical resistivity imaging method. Pure Appl.
Geophys. 2016,173, 2227–2239. [CrossRef]
16.
Cl
é
ment, R.; Descloitres, M.; Günther, T.; Ribolzi, O.; Legchenko, A. Influence of shallow infiltration
on time-lapse ERT: Experience of advanced interpretation. Comptes Rendus - Geosci.
2009
,341, 886–898.
[CrossRef]
17.
Loke, M.H. Electrical Imaging Surveys for Environmental and Engineering Studies: A Practical Guide to 2-D
and 3-D Surveys. Available online: https://www.semanticscholar.org/paper/Electrical-Imaging-Surveys- for-
Environmental-and-Loke/efa943b44f5a217e2d22255f7e334778932248 (accessed on 2 May 2019).
18.
Pellicer, X.M.; Zarroca, M.; Gibson, P. Time-lapse resistivity analysis of Quaternary sediments in the Midlands
of Ireland. J. Appl. Geophys. 2012,82, 46–58. [CrossRef]
19.
Gleeson, T.; Novakowski, K.; Kurt Kyser, T. Extremely rapid and localized recharge to a fractured rock
aquifer. J. Hydrol. 2009,376, 496–509. [CrossRef]
20.
Ulusoy, ˙
I.; Dahlin, T.; Bergman, B. Time-lapse electrical resistivity tomography of a water infiltration test on
Johannishus Esker, Sweden. Hydrogeol. J. 2015,23, 551–566. [CrossRef]
21.
Robert, T.; Caterina, D.; Deceuster, J.; Kaufmann, O.; Nguyen, F. A salt tracer test monitored with surface
ERT to detect preferential flow and transport paths in fractured/karstified limestones. Geophysics
2012
,77,
B55–B67. [CrossRef]
22. Nimmo, J.R. Vadose water. Encycl. Inl. Waters. 2009,1, 766–777. [CrossRef]
23.
Wu, J.; Zhang, R.; Yang, J. Analysis of rainfall-recharge relationships. J. Hydrol.
1996
,177, 143–160. [CrossRef]
24.
Kotchoni, D.O.V.; Vouillamoz, J.M.; Lawson, F.M.A.; Adjomayi, P.; Boukari, M.; Taylor, R.G. Relationships
between rainfall and groundwater recharge in seasonally humid Benin: A comparative analysis of long-term
hydrographs in sedimentary and crystalline aquifers. Hydrogeol. J. 2018,27, 447–457. [CrossRef]
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
... The relationship further suggests the eiciency of groundwater recharge during the rainy season (Blarasin et al. 2016). Moreover, the dependence of seasonal groundwater recharge on the hydraulic property of the vadose zone has been reported (Tesfaldet and Puttiwongrak, 2019). The seasonal resistivity variation of vadose zone is analyzed by time-lapse resistivity tomography, which proves the relationship. ...
... Well hydrograph shows water table luctuations and the extrapolation of the recession line (a) 2012, (b) 2013, (c) 2014 and (d) 2015(Tesfaldet and Puttiwongrak, 2019) ...
Article
Full-text available
Whether groundwater resources can be sustainably utilized is largely determined and characterized by hydrogeological parameters. Estimating the groundwater recharge is one of the essential parameters for managing water resources and protecting water resources from contamination. This study researched the spatial and temporal variation of groundwater recharge in the Thepkasattri sub-district through integrating chloride mass balance (CMB) and water table fluctuation (WTF) methods. The chloride content of representative rainfall and groundwater samples was analyzed. Besides, WTF method was adopted from groundwater level data from 2012 to 2015. According to the CMB method, the mean recharge was estimated to be 1 172 mm per year, accounting for 47% of the annual rainfall. Moreover, the estimated recharge from the WTF method took 26% of annual rainfall in 2015. The recharge was underestimated according to the WTF method, because of the uncertainty in specific yield estimates and the number of representative wells in the study area. Moreover, the correlation between rainfall and water table fluctuation data indicated the positive linear relationship between two parameters. The spatial recharge prediction indicated that recharge was higher (1 200-1 400 mm/yr) in the eastern and western catchment, while that in the central floodplains was between 800 mm/yr and 1 100 mm/yr. In addition, low recharge value between 450 mm/yr and 800 mm/yr was observed in the southwest part of Thepkasattri. The spatial variation of recharge partly reflects the influences of land use and land cover of the study area.
... Wenner array.Source:Adapted from[57]. ...
Article
This work presents a comparative study of the state-of-the-art Electrical Resistivity Tomography applications in the context of underground imaging and object detection. Parameters such as data acquisition or mapping, array protocol, inversion model, geological site description, and number of profiles are used to compare each application. In addition, a classification of these applications in underground imaging and development of a thematic taxonomy are presented. The thematic taxonomy is based on the type of underground object detected and surveyed. This includes underground utilities, underground cavities, old and ancient structures, fault zones and potential construction sites, hydrological features, and bedrock, concrete, and other objects. Finally, the issues and challenges are identified from each solution.
... Knowledge of groundwater flow and groundwater level variation is crucial to groundwater management and water quality control (Tesfaldet and Puttiwongrak, 2019). A further understanding of the mechanism governing groundwater movement in the vadose zone requires investigations on both groundwater flow characterization and the recharging process (Nielsen et al., 1986). ...
Article
Groundwater flow monitoring is crucial to groundwater management and pollution control. Traditional in-situ hydrological surveys monitor groundwater flow, pressure or temperature changes to infer the actual conditions of the underground, but there are characteristics of stratum destructiveness and difficult to obtain the regional variation characteristics of groundwater. Compared with geophysical methods, it has the advantages of non-destructive detection and regional quantitative interpretation, and indirect understanding of the physical properties of underground media and groundwater. In this paper, we combine the time-lapse electrical resistivity (ERT) and self-potential (SP) methods to monitor the groundwater flow. The ERT result establishes linkages between the resistivity and porewater content of the underground space, and captures the difference in resistivity through the change of underground water content. The SP method provides a direct way to predict groundwater flow according to its movement characteristics. Then, the joint resistivity and potential field are further used to estimate the soil's apparent-permeability. We carry out a real pumping experiment at a hydrological observation site to verify the proposed method. The time-lapse ERT and SP results reveal a significant increase in resistivity but a decrease in permeability during water pumping and groundwater recharge, representing the process of groundwater decline and recovery. The results also matched well with the groundwater level in-situ hydrological data. The study demonstrates that the joint ERT and SP data can provide a direct and reliable way to monitor groundwater flow or other time-lapse hydrogeological surveys in the vadose zone.
... The groundwater flow or level variation is crucial to groundwater management and site soil pollution control [1]. To understand the location and timing of groundwater traveling through the vadose zone, we must investigate groundwater flow, recharge mechanism, and monitoring groundwater flow. ...
Article
Full-text available
Revealing the dynamics of groundwater movement in the vadose zone is crucial to groundwater management and artificial recharge development. In this study, the groundwater flow characterization of the pumping process is monitored by the time-lapse electrical resistivity tomography (ERT) and self-potential (SP)data tomography. The ERT data invert the resistivity distribution. Then, we combine the SP data and resistivity result to invert the apparent current density and estimate the permeability based on the Poisson equation. A total of 24 hours of time-lapse surveys are taken during the pumping and recharge of groundwater. The results show a significant increase in in resistivity and permeability during pumping water, which suggests groundwater drawdown. Similarly, significant decrease resistivity and permeability are observed during the recovery indicating groundwater recharge. These results have a significant agreement with the groundwater table monitoring result. Our experiment test suggests that combine ERT and SP data could provide a reliable way in groundwater or other hydrogeological surveys.
... The groundwater flow or level variation is crucial to groundwater management and site soil pollution control [1]. To understand the location and timing of groundwater traveling through the vadose zone, we must investigate groundwater flow, recharge mechanism, and monitoring groundwater flow. ...
Thesis
Full-text available
Water recharge areas are the main sources of fresh water and play an important role in the hydrological cycle, allowing the conservation, preservation, renewal and use of water. The SSMVC has been considered as a recharge zone towards the deep aquifer in San Luis Potosi Valley; however, currently there are not enough and robust studies to confirm it. The aim of this work was to determine if the SSMVC works or not as a water recharge zone, applying Multivariate Statistical Techniques: correlation analysis, cluster analysis (CA) and principal components analysis (PCA). A data matrix was integrated with 66 study points and 12 variables (altitude, slope, temperature, soil type, vegetation type, precipitation, geology, relative humidity, PET, land use, runoff coefficient and hydraulic conductivity). The variables were determined at each study point through calculations, estimates and layer overlap in a GIS. The correlation analysis showed high positive and negative correlations between the climatic variables: PET, temperature, humidity and precipitation, while geology was the non-climatic variable that was most correlated with others, highlighting its importance in the study area. The CA by K-means classified the study points in three groups, where cluster 2 presented a recharge potential due to geological composition (alluvium) and land use (agriculture); however, factors such as high PET, temperature and low humidity are limiting this fact, while in cluster 1 and 3 recharge is limited by variables such as geology, altitude and slope, and infiltration could be favored by the faults present in the study area. The PCA revealed that variables such as hydraulic conductivity, precipitation and humidity may favor infiltration, while recharge is limited by the geological units present in the study area, high levels of PET and temperature. On the other hand, Electrical Resistivity Tomography’s (ETR) were analyzed and related, indicating that the combination of vegetation and geology in the upper parts of the SSMVC plays an important role in the surface infiltration, reaffirming the existence of a sub-hypodermical water flow. The results obtained in this study demonstrate that the climatic and geological conditions of the SSMVC do not allow it to function optimally as a recharge zone towards the deep aquifer; however, it complies with other important ecological processes such as the existence of a sub-hypodermical water flow, which can contribute to the filling of embankments and dams as well as the support of other ecosystem services. The results will contribute to the optimal protection, conservation and sustainable use of water resources in the San Luis Potosí Valley.
Article
Full-text available
Globally, groundwater resources play a crucial role in supporting livelihoods and sustaining human health. Groundwater recharge is mainly influenced by the spatial distribution of rainfall patterns across groundwater basins and the heterogeneous distribution of geology, soil, and topographical characteristics. This study evaluates the implications of variations in rainfall and land-use on groundwater level fluctuations in the Kathmandu Valley, Nepal. We selected and analyzed data from thirty-five monitoring wells observed by citizen scientists from July 2017 to June 2019. Over two years, groundwater levels varied spatially from -0.11m (negative sign denotes a groundwater level higher than the ground surface) to 11.5m, with a mean of 4.24m and a standard deviation of 2.29m. Our results indicated a strong positive correlation between rainfall and groundwater levels, as the fluctuation was influenced by the rainfall of that area. The seasonal fluctuations in groundwater level showed the direct influence of monsoonal rainfall. In areas with agricultural land use, 80% of the analysed wells showed a strong and statistically significant correlation between rainfall and groundwater levels. In contrast, higher groundwater extraction rates and surface sealing limited groundwater recharge in built land uses; therefore wells across non-agricultural land-uses showed a weak correlation in most of the cases. In the northern groundwater district, we found areas with highly permeable sand and gravel have nearly constant groundwater levels (shallow in agricultural land-use and deep in built land-use) year-round and act as potential recharge zones for the aquifer. The areas with less permeable but highly porous clay and silt in agricultural land use, and show greater seasonal groundwater fluctuations. Our study highlights the ability of citizens to generate meaningful hydrogeologic datasets, and the importance of rainfall and land use planning to groundwater recharge. Understanding these complex relationships must form the basis for the sustainable management of the rapidly declining groundwater resources of the Kathmandu Valley.
Preprint
Full-text available
Groundwater is the dominant source of water supply on Phuket island, Thailand. The water demand on Phuket has been increasing due to rapid urbanization and population growth. A scarcity of freshwater and over-extraction of groundwater may shortly become severe problems for Phuket. Geoelectrical data obtained by Vertical Electric Sounding (VES) were employed in this study to estimate the Dar Zarrouk parameters of the study area. Twenty-four VES stations were set up using Schlumberger configuration with a 1.0 m minimum spacing. The lithology layers from 24 unpublished borehole data indicating each subsurface layer were validated with the resistivity data obtained from VES. The geoelectric profiles gave a maximum of three layers with varying resistivities and thicknesses across each VES station. Two parameters (longitudinal conductance and transverse resistance) of the Dar Zarrouk model were calculated from VES data to generate the thematic maps in a GIS environment, thus the groundwater potential in Phuket Island was represented as a single map by using the weighted overlay technique in ArcGIS, based on both longitudinal conductance and transverse resistance. The groundwater potentials were classified into three potential levels (low, medium, and high). The results reveal that the highest groundwater potential areas are isolated and in specific locations, while the majority of area has medium level groundwater potential. Finally, the low potential zones are in the flank and the top parts of the study area.
Article
Two-dimensional (2D) electrical resistivity tomography (ERT) surveying systems typically use standard protocols that assume straight survey lines and collinear electrodes. However, due to the presence of infrastructures and other surface constraints, it is not always possible to conduct 2D ERT surveys along straight lines. When survey lines are angled, the apparent resistivity data are affected for measurements where electrodes occur on either side of the position of the angle, since the assumed distances between some of the electrodes are larger than the true distances. This study investigates the impact of angled survey lines on 2D ERT data and inverse resistivity models for the Wenner (α) array. Theoretical considerations show that measurements along angled survey line are sensitive to resistivity changes at position laterally displaced from the survey line, while both the calculated apparent resistivities and the depths of investigation are affected when electrodes straddle the angle. The resulting errors are, however, small. Even for a large angle of 45°, the errors are smaller than 6.5% for the affected apparent resistivity measurements, and smaller than 3.8% for the median depths. Since the apparent resistivity data and the pseudo-depths are affected by angles in the survey lines, the inverse resistivity models are also affected. Numerical modelling shows that even small input errors may be propagated (both in magnitude and spatially) during inversion, thereby distorting the resulting resistivity models. The distorted inverse resistivity models could lead to the incorrect interpretation of the locations, geometries and orientations of subsurface structures, which could lead to poorly sited production boreholes. The results of field surveys show that apparent resistivity data sets recorded along angled lines may be partially corrected by calculating the true geometric factors and by adjusting the apparent resistivity data accordingly. Significantly improved inverse resistivity models are found for corrected data sets.
Article
Full-text available
Groundwater is a vital source of freshwater throughout the tropics enabling access to safe water for domestic, agricultural and industrial purposes close to the point of demand. The sustainability of groundwater withdrawals is controlled, in part, by groundwater recharge, yet the conversion of rainfall into recharge remains inadequately understood, particularly in the tropics. This study examines a rare set of 19–25-year records of observed groundwater levels and rainfall under humid conditions (mean rainfall is ~1,200 mm year⁻¹) in three common geological environments of Benin and other parts of West Africa: Quaternary sands, Mio-Pliocene sandstone, and crystalline rocks. Recharge is estimated from groundwater-level fluctuations and employs values of specific yield derived from magnetic resonance soundings. Recharge is observed to occur seasonally and linearly in response to rainfall exceeding an apparent threshold of between 140 and 250 mm year⁻¹. Inter-annual changes in groundwater storage correlate well to inter-annual rainfall variability. However, recharge varies substantially depending upon the geological environment: annual recharge to shallow aquifers of Quaternary sands amounts to as much as 40% of annual rainfall, whereas in deeper aquifers of Mio-Pliocene sandstone and weathered crystalline rocks, annual fractions of rainfall generating recharge are 13 and 4%, respectively. Differences are primarily attributed to the thickness of the unsaturated zone and to the lithological controls on the transmission and storage of rain-fed recharge.
Article
Full-text available
The present investigation was focused on the variations in rainwater infiltration experienced by soils of Gamboa zone (Panama Canal Watershed) during various seasons of the year, employing a time-lapse analysis of electrical resistivity tomography (ERT). In 2009, a total of 3 geoelectrical tests were undertaken during the dry, transition and rainy seasons across a profile 47 m in length, strategically distributed on site. The results obtained in this study showed strong variations in calculated resistivity between these seasons, taking the dry season as a reference with decreases and increases of percent difference of resistivity between −20% and −100%, and between 50% and 100%, respectively. These decreases, when displayed through a sequence of time-lapse images, reveal a superficial extension of the water content variations along the entire profile, as well as strong inversion artifacts showing false increases of calculated electrical resistivity. Decreases are the product of the rainfall increase obtained in this type of tropical environment; permanent conductive anomalies in 3 tests are associated with the streams close to the study site. The results of this work were compared with a simulation resulting from a series of bidimensional models applied to the 3 studies evaluated: dry, transition and rainy seasons.
Article
Full-text available
Electrical Resistivity Imaging (ERI) was carried out continuously for ten days to map the subsurface resistivity distribution along a potentially hazardous hillslope at the Jieshou Junior High School in Taoyuan, Taiwan. The reliability of the inverted resistivity structures down to about 25 m deep was examined with synthetic modeling using the same electrode arrangements installed on land surface as in field surveys, together with a DOI (depth-of-investigation) index calculated from the ERI data. The subsurface resistivity distribution is consistent with results from well logging. These ERI recordings were taken daily and provided highly-resolved imagery of the resistivity distribution underground and illustrated the dynamical fluid-flow behavior due to heavy rainfall infiltration. By using Archie’s law, the resistivity distribution was transformed into a map of relative water saturation (RWS), which is strongly correlated with the rainfall infiltration process. We then found that the averaged RWS is significantly correlated with daily precipitation. Our observations indicate that time-lapse ERI is effective in monitoring subterraneous rainfall infiltration, moreover, the preferential flow paths can be delineated according to the changes in averaged RWS derived from the ERI data.
Article
Time-lapse electrical resistivity tomography (ERT) is commonly used as a minimally invasive tool to study infiltration processes. In 2014, we conducted field studies coupling variable intensity rainfall simulation with high-resolution ERT to study the real-time partitioning of rainfall into surface and subsurface response. The significant contrast in resistivity in the subsurface from large changes in subsurface moisture resulted in artifacts during the inversion process of the time-lapse ERT data collected using a dipole-dipole electrode array. These artifacts, which are not representative of real subsurface moisture dynamics, have been shown to arise during time-lapse inversion of ERT data and may be subject to misinterpretation. Forward modeling of the infiltration process post field experiments using a two-layer system (saprolite overlain by a soil layer) was used to generate synthetic datasets. The synthetic data were used to investigate the influence of both changes in volumetric moisture content and electrode configuration on the development of the artifacts identified in the field datasets. For the dipole-dipole array, we found that a decrease in the resistivity of the bottom layer by 67% resulted in a 50% reduction in artifact development. Artifacts for the seven additional array configurations tested, ranged from a 19% increase in artifact development (using an extended dipole-dipole array) to as much as a 96% decrease in artifact development (using a wenner-alpha array), compared to that of the dipole-dipole array. Moreover, these arrays varied in their ability to accurately delineate the infiltration front. Model results showed that the modified pole-dipole array was able to accurately image the infiltration zone and presented fewer artifacts for our experiments. In this study, we identify an optimal array type for imaging rainfall-infiltration dynamics that reduces artifacts. The influence of moisture contrast between the infiltrating water and the bulk subsurface material was characterized and shown to be a major factor in contributing to artifact development. Through forward modeling, this study highlights the importance of considering array type and subsurface moisture conditions when using time-lapse resistivity to obtain reliable estimates of vadose zone flow processes during rainfall-infiltration events.
Article
Understanding groundwater recharge is essential for successful management of water resources and modeling fluid and contaminant transport within the subsurface. This book provides a critical evaluation of the theory and assumptions that underlie methods for estimating rates of groundwater recharge. Detailed explanations of the methods are provided - allowing readers to apply many of the techniques themselves without needing to consult additional references. Numerous practical examples highlight benefits and limitations of each method. Approximately 900 references allow advanced practitioners to pursue additional information on any method. for the first time, theoretical and practical considerations for selecting and applying methods for estimating groundwater recharge are covered in a single volume with uniform presentation. Hydrogeologists, water-resource specialists, civil and agricultural engineers, earth and environmental scientists and agronomists will benefit from this informative and practical book. it can serve as the primary text for a graduate-level course on groundwater recharge or as an adjunct text for courses on groundwater hydrology or hydrogeology. for the benefit of students and instructors, problem sets of varying difficulty are available at http://wwwbrr.cr.usgs.gov/projects/GW_Unsat/Recharge_Book/.
Article
Managed aquifer recharge (MAR) is an efficient way to remove organic matter from raw water and, at the same time, reduce temperature variation. Two MAR sites were constructed by Karlskrona municipality on Johannishus Esker in Sweden. One of these sites, Vång, was monitored for electrical conductivity and electrical resistivity (using electrical resistivity tomography - ERT) during a 9-week tracer infiltration test. The aim of the monitoring was to map the pathways of the infiltrated water, with the overall goal to increase the efficiency of the MAR. ERT proved useful in determining both the nature of the esker formation and the water migration pathways. In Vång, the esker ridge follows a tectonically controlled paleo-valley. The fault/fracture zone in the bedrock along this paleo-valley was mapped. During the tracer test, the infiltrated water was detected in the area close to the infiltration ponds, whereas far-situated observation wells were less affected. For sequential infiltration and recharge periods in MAR, the timing of the well pumping is another important factor. Natural groundwater flow direction was a determinant in the infiltration process, as expected. ERT measurements provide supplementary data for site selection, for monitoring the functionality of the MAR sites, and for revealing the geological, hydrogeological and structural characteristics of the site.