water

Article
Determination of Soil Electrical Conductivity and Moisture on
Different Soil Layers Using Electromagnetic Techniques in
Irrigated Arid Environments in South Africa
Phathutshedzo Eugene Ratshiedana 1,2, * , Mohamed A. M. Abd Elbasit 3 , Elhadi Adam 2 ,
Johannes George Chirima 1,4 , Gang Liu 5 and Eric Benjamin Economon 1

1 Agricultural Research Council-Natural Resources and Engineering-South Africa, 600 Belvedere Street,
Arcadia, Pretoria 0083, South Africa
2 School of Geography, Archaeology and Environmental Studies, University of the Witwatersrand,
Johannesburg Private Bag x3, Wits, Johannesburg 2050, South Africa
3 Department of Physical and Earth Sciences, School of Natural and Applied Sciences, Sol Plaatje University,
Kimberley 8300, South Africa
4 Department of Geography, Geoinformatics and Meteorology, University of Pretoria,
Pretoria 0028, South Africa
5 Institute of Soil and Water Conservation, Chinese Academy of Sciences and Ministry of Water Resources,
Xianyang 712100, China
* Correspondence: [email protected]; Tel.: +27-012-310-2598

Abstract: Precise adjustments of farm management activities, such as irrigation and soil treatment
according to site-specific conditions, are crucial. With advances in smart agriculture and sensors, it is
possible to reduce the cost of water and soil treatment inputs but still realize optimal yields and high-
profit returns. However, achieving precise application requirements cannot be efficiently practiced
with spatially disjointed information. This study assessed the potential of using an electromagnetic
induction device (EM38-MK) to cover this gap. An EM38-MK was used to measure soil apparent
electrical conductivity (ECa) as a covariate to determine soil salinity status and soil water content
Citation: Ratshiedana, P.E.;
θ post irrigation at four depth layers (Hz: 0–0.25 m; Hz: 0–0.75 m; Vz: 0.50–1 m). The inverse
Abd Elbasit, M.A.M.; Adam, E.;
distance weighting method was used to generate the spatial distribution thematic layers of electrical
Chirima, J.G.; Liu, G.; Economon, E.B.
conductivity. The statistical measures showed an R2 = 0.87; r > 0.7 and p ≤ 0.05 on correlation of ECa
Determination of Soil Electrical
and SWC. Based on the South African salinity class of soils, the area was not saline ECa < 200 mS/m.
Conductivity and Moisture on
Different Soil Layers Using
The EM38-MK can be used to estimate soil salinity and SWC variability using ECa as a proxy, allowing
Electromagnetic Techniques in precise estimations with depths and in space. These findings provide key information that can aid in
Irrigated Arid Environments in South irrigation scheduling and soil management.
Africa. Water 2023, 15, 1911. https://
doi.org/10.3390/w15101911 Keywords: apparent electrical conductivity; salinity; soil moisture; em38-mk; inverse distance weighting;
spatial distribution
Academic Editor: Yanshan Cui

Received: 12 April 2023

Revised: 7 May 2023
Accepted: 16 May 2023 1. Introduction
Published: 18 May 2023
Soil moisture is a widely perceived component of the water balance, crucial for crop
development and nutrient transport from the soil to crops [1,2]. Apart from its role in agri-
culture, soil moisture plays an important role in the Earth’s energy balance by controlling
Copyright: © 2023 by the authors.
the changes of surface water and energy fluxes to balance atmospheric water demands
Licensee MDPI, Basel, Switzerland. through evaporation [3]. Generally, soil moisture in arid environments is a limited compo-
This article is an open access article nent of the hydrological cycle due to limited and variable rainfall and high temperatures,
distributed under the terms and along with long spells of droughts and lack of water sources [4]. However, most arid
conditions of the Creative Commons regions contain arable land, and usually the supplemental irrigation that they rely on
Attribution (CC BY) license (https:// consists of water transfers from areas with ample availability [5].
creativecommons.org/licenses/by/ Given the challenges faced by arid areas, and with the scarcity of water, the need
4.0/). to monitor, manage, and control water application rates through irrigation scheduling

Water 2023, 15, 1911. https://doi.org/10.3390/w15101911 https://www.mdpi.com/journal/water

Water 2023, 15, 1911 2 of 23

remains a key mitigation approach [6]. That being stressed, various factors come into
play when irrigation events are scheduled; one of the most important components of
accurately determining prior irrigation scheduling is soil moisture. Accurate measurement
and estimations of soil moisture allow farmers to fully understand the water dynamics
in their cropped fields, both spatially and with depth [7]. The spatial distribution of
soil moisture in water-scarce South Africa is a major challenge, particularly within the
agricultural sector where moisture information is critical for irrigation scheduling. Most
measurements are based on point data, which lack the spatial representativeness of the
area. With improvements in smart agriculture, point information does not provide much
guidance on spot treatment of soil or on scheduling irrigation only in needy areas.
Previous studies have demonstrated various methods of monitoring soil moisture from
plant level, field scale, and global scales to assist farmers in proper irrigation scheduling [8,9].
Several strategies to measure and monitor soil moisture have been implemented, including
the application of gravimetric methods where soil is sampled, weighed, dried, and weighed
again to calculate the water loss. Apart from the above methods, others such as nuclear-
based soil moisture assessment, tensiometric approaches, and hygrometric approaches
have been widely explored [10–12]. However, all soil moisture measuring and estimation
methods define moisture from a fixed point perspective without taking into account the
spatial variability of soils within the field [13]. Furthermore, methods such as gravimetric
approaches come with some limitations, such as induced errors on measuring scales, human
errors when recording data, and high cost [14,15].
The use of probes, such as the capacitance probes, measuring the dielectric current
to estimate moisture at a point level also have challenges such as inducing errors, partic-
ularly in clay soils where mud cracks are more common during moisture deficit periods.
Such cracks create an air space, removing the sensor contact to the soil, which results in
null readings occurring. This is not only the case with capacitance probes installed in
fields; it applies to any contact sensors that may be used to obtain instant readings [16,17].
Although such approaches provide more detailed information about soil water content
depths and salinity status, they also have spatial limitations [18]. Despite the fact that point
measurements lack spatial representation of an area, they are still crucial for calibration
and evaluation of spatial estimation approaches [19].
With increasing interest in precision farming and smart agriculture, soil properties
such as soil electrical conductivity (EC), texture, permeability, and soil moisture are some
of the most crucial, which farmers and agricultural advisors need to understand prior to
any land use [20,21]. Spatial data on soil salinity and moisture availability carry crucial
information, which aids in farm irrigation management and soil treatment due to the
heterogeneity of soil properties in space and time [22]. The transformation from traditional
point soil moisture and EC measurement, through destructive soil sampling, to geospatial
approaches has gained more attention in attempts to solve water scarcity problems while
increasing water use efficiency and resource sustainability [23,24].
The development of satellite-based remote-sensing soil moisture products has added
value in understanding soil water dynamics spatially. Global soil moisture products have
been developed from satellite remote sensing and are widely used in large-scale investi-
gations. These products include the Soil Moisture Active Passive (SMAP) at 36 km with
a daily temporal resolution retrieval; Soil Moisture and Ocean Salinity (SMOS) at 25 km
and daily retrieval; and Scanning Multichannel Microwave Radiometer (SMMR) at 150 km
on daily basis, to list just a few [25,26]. However, due to their coarse resolutions, their
applicability at farm scales is challenging. With such challenges, various researchers have
integrated different satellite data, such as the Moderate Resolution Imaging Spectrora-
diometer (MODIS) with 250 m resolution [27,28], Landsat 8 OLI at 30 m resolution [29],
and Sentinel-1 at 10–20 m resolution, to aid in understanding moisture dynamics at field
scales [30]. Despite these advances, soil moisture spatial data obtained from remotely
sensed observations only estimate soil moisture within the top 0–5 cm; this already is a
limitation when assessing moisture dynamics for deep-rooted crops [31].
Water 2023, 15, 1911 3 of 23

The limited coverage of soil depth and low spatial resolution of these datasets do
not adequately address many issues at national scale and, more importantly, at local field
scales. The wide range of image coverage and ease of access enables their applications from
regional to global scales, however, mixed pixels are common in remote sensing data, making
monitoring soil moisture variability in heterogeneous terrains more difficult [32]. With
such challenges in the past decade, advances and innovations in science and technology
have led scientists to develop sensors that can estimate soil moisture indirectly, with
reasonable acquisition depths such that the root zone of various crops can be reached
and their information can be integrated in geospatial tools to develop spatial digital soil
moisture maps [33–35]. Some of the most promising sensors include the non-destructive
electromagnetic induction sensors and ground penetration radar geophysical tools used in
groundwater exploration [36,37].
However, such sensors also do not directly measure volumetric water content but
rather measure a different parameter, which can be related to soil water content as a
surrogate [38,39]. Unfortunately, apart from the use of hyperspectral sensors, until now
there have been no satellite sensors providing information on soil parameters such as EC,
soil types, and soil fertility, which are important properties of the soil that play a crucial role
in water absorption by plants [40]. Consequently, spatial information of soil EC remains
the biggest data gap challenge. In areas with high salinity problems, EC plays a critical
role in delineating zones for salinity treatment. The lack of spatial data on soil EC requires
ground measurement of soil EC or laboratory analysis of field sampled soil, which is a
tedious exercise in larger areas. High salinity zones are more prone to water logging and
poor crop performance.
Devices such as electromagnetic induction sensors (EM38) have indirectly come a long
way in the history of soil moisture estimation by measuring apparent electrical conductivity
(ECa) and relating it to soil water content through calibration methods [41–48]. The EM38
is a non-destructive soil sensor, which retrieves soil electrical conductivity as a function
of soil moisture availability to move the charges available and conduct electrical current
flow [47]. The sensor takes measurements to a depth of 1 m. The EM38 provides high
temporal resolution data, at as high as 1 s intervals, however, EM38 sensors are expensive.
The use of EM38 sensors for salinity assessment used to be a time-consuming activity prior
to the introduction of smart sensors and precision technology. Currently, EM38 devices can
be towed in farm machineries and protected with non-metallic covers to quickly survey an
entire field in a short space of time. Its ability to estimate soil properties at four soil depths
makes it attractive and necessary for understanding soil moisture dynamics at different
root zones, especially in irrigated fields where scheduling irrigation is a prerequisite.
Salinity and soil moisture content availability are the two major concerns in irrigated
farming regarding soil quality and water quantity. However, the availability of spatial soil
water content information in the South African farming environment is very limited. As
such, point information dominates most areas. This information provides crucial guidance
in agricultural planning and farm management in order to solve water and treatment
loses. EC can be used as a surrogate for soil water content estimation in situations where
moisture sensors are not available to provide the spatial representativeness of an area.
This study aims at mapping the spatial (Hz) and vertical (Vz) distribution of soil electrical
conductivity and soil moisture at different soil depths by integrating electromagnetic data
with geospatial techniques under different wetness conditions post irrigation.

2. Materials and Methods

2.1. Study Area Description
The study was carried out on an 18 ha experimental farm owned by the South
African Barley Breeding Institute (Figure 1a). The area is located 12 km from Hartswa-
ter town and is part of the Phokwane Local Municipality in the Frances Baard Dis-
trict Municipality (Figure 1c) of the Northern Cape Province (Figure 1b). Geographi-
cally, the 18 ha experimental farm is located between the coordinates 27◦ 430 14.4100 S and
 2. Materials and Methods
2.1. Study Area Description
The study was carried out on an 18 ha experimental farm owned by the South African
Water 2023, 15, 1911 Barley Breeding Institute (Figure 1a). The area is located 12 km from Hartswater town and4 of 23
is part of the Phokwane Local Municipality in the Frances Baard District Municipality
(Figure 1c) of the Northern Cape Province (Figure 1b). Geographically, the 18 ha experi-
24mental farm
◦ 440 34.18 is located between the coordinates 27°43′14.41″ S and 24°44′34.18″ E. The ex-
00 E. The experimental farm is located within the Vaalharts irrigation and cov-
perimental farm is located within the Vaalharts irrigation and covers an area of over 36,950
ers an area of over 36,950 hectares, being the largest irrigation scheme in South Africa.
hectares, being the largest irrigation scheme in South Africa. Ojo et al. [49] reported that
Ojo et al. [49] reported that the Vaalharts Irrigation Scheme was planned in 1933 to address
the Vaalharts Irrigation Scheme was planned in 1933 to address unemployment, hunger,
unemployment, hunger, and poverty. The proposal to develop this scheme was approved
and poverty. The proposal to develop this scheme was approved in 1934, which resulted
in in
1934, which resulted
the construction of the in Vaal
the construction of the
dam. Water was thenVaal dam.from
diverted Water thewas
Vaalthen
Riverdiverted
weir on from
thetheVaal River weir on the eastern side of Warrenton [50]. As the water table
eastern side of Warrenton [50]. As the water table rose, the scheme’s salinity problems rose, the
scheme’s
initiatedsalinity problems
and worsened initiated and
waterlogging, worsened
which waterlogging,
to date prevents the usewhich to date prevents
of groundwater in
thethe
use of due
area groundwater in the area[50].
to salt accumulations dueThe
to salt
farmaccumulations [50]. 1088
is situated between The farm is situated
and 1090 m
between
above sea1088 and 1090 m above sea level.
level.

Figure 1. Locality map of the study area showing the experimental farm and the area of interest
Figure 1. Locality map of the study area showing the experimental farm and the area of interest
used for the EM38 survey where (a) is the 18 ha experimental farm, the red boundary is the EM38
used for thearea,
surveyed EM38 (b) survey
shows the where (a) is the
provincial 18 of
layout haSouth
experimental farm, the Northern
Africa, highlighting red boundary is the EM38
Cape (NC)
surveyed
where thearea, (b)area
study shows the provincial
is located layout
in green, while (c)of South
shows theAfrica,
locationhighlighting
of the surveyNorthern
area in theCape
dis- (NC)
trict municipality.
where the study area is located in green, while (c) shows the location of the survey area in the
district municipality.
According to Vermeulen and van Niekerk [51], sandy soils and inadequate natural
drainage in Vaalharts
According are well known
to Vermeulen and van forNiekerk
causing waterlogging.
[51], sandy soilsThe months of December
and inadequate natural
and January typically have the highest monthly average temperature,
drainage in Vaalharts are well known for causing waterlogging. The months of December of approximately
and32 January
°C, whiletypically
July has the havelowest
the monthly
highest average
monthlytemperature of 0.5 °C. Theof
average temperature, summer sea-
approximately
son,
◦ which runs from October to April, has an average precipitation
32 C, while July has the lowest monthly average temperature of 0.5 C. The summer of
◦ 450 mm [52]. Theseason,
area receives summer rainfall with hot days, and has a cold winter
which runs from October to April, has an average precipitation of 450 mm [52]. The area season [53]. The
Vaalharts irrigation scheme is in operation for one of the major irrigated pecan producers,
receives summer rainfall with hot days, and has a cold winter season [53]. The Vaalharts
and it uses pivot irrigation systems [54]. The fact that the area is the biggest pivot-irriga-
irrigation scheme is in operation for one of the major irrigated pecan producers, and it uses
tion-dominant scheme in South Africa and located in a dry area made it attractive for the
pivot irrigation systems [54]. The fact that the area is the biggest pivot-irrigation-dominant
undertaking of this study for crop water use assessment, especially with the fact that ag-
scheme in South
riculture dominates Africa
theand located
use of in a dry
freshwater area
in the made[54].
country it attractive for the undertaking of
this study for crop water use assessment, especially with the fact that agriculture dominates
the use of freshwater in the country [54].

2.2. Methods and Approaches

2.2.1. Soil Electrical Conductivity Surveying
Four field survey campaigns were conducted within a selected area, located in an exper-
imental barley crop field, using an electromagnetic induction EM38-mk device (Figure 2).
The device is equipped with a signal transmitting coil at one end and a receiver coil on the
other end, spaced 1 m apart.
 2.2. Methods and Approaches
2.2.1. Soil Electrical Conductivity Surveying
Four field survey campaigns were conducted within a selected area, located in an
experimental barley crop field, using an electromagnetic induction EM38-mk device (Fig-
Water 2023, 15, 1911 5 of 23 coil
ure 2). The device is equipped with a signal transmitting coil at one end and a receiver
on the other end, spaced 1 m apart.

Figure 2. An
Figure electromagnetic
2. An electromagneticinduction
induction instrument (EM38MK)
instrument (EM38MK) andand its its components,
components, wherewhere
(a) is (a)
the is the
EM38EM38device, (b)(b)
device, is is
the
thedevice
deviceonon vertical modeduring
vertical mode duringsurvey,
survey,
(c) (c) shows
shows the the setting
setting controls,
controls, and and
(d) is
(d)the Bluetooth
is the Bluetooth GPSGPSlinked
linkedto
to the devicefor
the device forsample
sample location.
location.

TheThe device
device wasused
was usedtotosurvey
survey the
the same
samearea
areatwice
twice onon
each
eachfield campaign
field campaign to allow
to allow
for two survey orientations wherein one survey was done with the instrument
for two survey orientations wherein one survey was done with the instrument on vertical on vertical
orientation, and another on horizontal orientation. The soil signal penetration depth when
orientation, and another on horizontal orientation. The soil signal penetration depth when
the device is on vertical mode (Vz) is 0–50 cm and 0–100 cm, and when the device is on
the its
device is on orientation
horizontal vertical mode (Hz)(Vz) is 0–50 cm and
the measurement 0–100
depth cm,cm
is 0–25 and when
and 0–75the
cm.device
The same is on its
horizontal
height of 10 cm above the soil was maintained using belt straps, reducing the seesawsame
orientation (Hz) the measurement depth is 0–25 cm and 0–75 cm. The
height ofon
effect 10 the
cm instrument.
above the soil wasthe
Before maintained using
start of each belt the
survey, straps, reducing
battery thethe
status of seesaw
deviceeffect
on was
the instrument.
checked. TheBefore the start
EM38 device wasofcalibrated
each survey,
at 1.5 the battery
m height status
above of the
ground device
each time was
before the survey was done, according to the Geonics ® instructions published in Canada.
checked. The EM38 device was calibrated at 1.5 m height above ground each time before
theThe surveyed
survey was area
done,was cleared ofto
according allthe
metallic objects
Geonics to avoid the introduction
® instructions published inofCanada.
artificial The
magnetic
surveyed areanoise,
waswhich
clearedincluded the removal
of all metallic of metal
objects belts,the
to avoid watches, cell phones,
introduction and anymag-
of artificial
electronic devices that could affect the measurements. The survey lines followed the raw
netic noise, which included the removal of metal belts, watches, cell phones, and any elec-
paths available between the barley crops, which changed with different field campaigns as
tronic
the devices
crop grew that could
and canopyaffect thebecame
cover measurements.
dense. AllThe survey
stations lines followed
surveyed the using
were located raw paths
available
a GPS connected to the computer, and the computer was connected to the instrument foras the
between the barley crops, which changed with different field campaigns
crop grew
data and during
logging canopy thecover
surveybecame
(Figuredense.
2). All stations surveyed were located using a
GPS connected to the computer, and the computer was connected to the instrument for
2.2.2.
data Calibration
logging duringofthe
thesurvey
Electromagnetic Induction Instrument (EM38)
(Figure 2).
To calibrate the electromagnetic induction instrument (EM38), apparent soil electri-
2.2.2. Calibration ofwas
cal conductivity themeasured in the horizontal
Electromagnetic Induction mode (0.25–0.75(EM38)
Instrument m) and vertical mode
(0.50–1 m) within the EM38 survey area of interest. Following the EM38, soil samples were
To calibrate
collected the
at four electromagnetic
different sites withininduction
the EM38instrument
survey area (EM38), apparent
of interest soil electrical
to determine soil
conductivity
properties. was measured
At each point, a in theauger
hand horizontal
was usedmode (0.25–0.75
to take m)0–1
cores from andmvertical modeof(0.50–
at an interval
1 m) within
0.25 thedepth
m with EM38 survey
added areacores
to four of interest. Following
being taken from eachthepoint
EM38, soil samples
(Figure were col-
3a). For every
lected
coreat fourphysical
taken, different sites within
properties the in
including EM38 survey
situ soil waterarea of interest
content, electricaltoconductivity,
determine soil
and temperature
properties. At eachwere measured
point, a hand before
augersample
was usedpackaging.
to take cores from 0–1 m at an interval
of 0.25 m with depth added to four cores being taken from each point (Figure 3a). For
every core taken, physical properties including in situ soil water content, electrical con-
ductivity, and temperature were measured before sample packaging.
Water 2023,
Water15, x FOR
2023, PEER REVIEW
15, 1911 6 of 25
6 of 23

Figure
Figure 3. Soil
3. Soil samplingatatdifferent
sampling different depths
depths for
forEM38
EM38calibration.
calibration.(a)(a)
Different layers
Different of soil
layers oftaken at
soil taken at
each calibration point, (b) surveyors taking core samples at different depths equivalent to the
each calibration point, (b) surveyors taking core samples at different depths equivalent to the EM38 EM38
measurement
measurement depth.
depth.
All core soil samples were stored in zip lock bags labelled with GPS coordinates, depth,
Allsample
and core soil samples
number; were
zip lock stored
bags were in ziptolock
used bags
prevent labelled
moisture with
loss fromGPS coordinates,
the samples
depth, and sample number; zip lock bags were used to prevent moisture
(Figure 3b). The samples were taken to the laboratory for physiochemical analysis. Similar loss from the
samples (Figure 3b).
soil properties, The samples
including were
soil water takenand
content to the laboratory
electrical for physiochemical
conductivity analysis.
saturated at 1:2.5,
Similar soil properties,
were analyzed including
at the laboratory. soil water
Additional soilcontent andthat
properties electrical
includedconductivity saturated
the SAR properties
(Mg, Ca,
at 1:2.5, were Na, and K), cations,
analyzed and the soilAdditional
at the laboratory. pH were also analyzed
soil as other
properties that soil attributes
included the SAR
influencing the ECa response. Linear regression analyses were applied for
properties (Mg, Ca, Na, and K), cations, and the soil pH were also analyzed as other soil all depths and
sample points
attributes to establish
influencing a relationship
the ECa response.between temperatureanalyses
Linear regression ECa-EM38,
correctedwere applied ECafor all
from EC-meter, ECe, soil water content, and other soil properties, which best describes
depths and sample points to establish a relationship between temperature corrected ECa-
their correlations.
EM38, ECa from EC-meter, ECe, soil water content, and other soil properties, which best
describes their Distribution
2.2.3. Spatial correlations.of Soil Electrical Conductivity
EM38 readings were used to estimate the spatial distribution patterns of soil electrical
2.2.3. Spatial Distribution
conductivity of Soil
during various Electrical
irrigation Conductivity
events. The survey’s EM38 measurements were
interpolated in ArcGIS
EM38 readings were10.6used Esri© softwarethe
to estimate using an inverse
spatial distance
distribution weighting
patterns interpo-
of soil electrical
conductivity during various irrigation events. The survey’s EM38 measurements The
lation to map the spatial distribution of electrical conductivity within the barley. were in-
inverse distance weighting interpolation approach was selected due to its ability to inter-
terpolated in ArcGIS 10.6 Esri© software using an inverse distance weighting interpola-
polate unknown areas without overestimating or underestimating electrical conductivity
tionmeasurement
to map the spatial
readingsdistribution
[55]. The mapsof electrical
depictingconductivity
the apparentwithin theconductivity
electrical barley. The inverse
of
distance
the soilweighting interpolation
were produced with a pixel approach
size of 35was
cm selected
× 40 cm. due to four
For all its ability to interpolate
field campaigns,
unknown areas without
spatial distribution maps overestimating
were generatedor forunderestimating
four depths of farm electrical conductivity
soil at the selected areameas-
urement readings
of interest, [55].
with the The
first maps
layer depictingthe
representing theroot
apparent
zone ofelectrical
the barleyconductivity of the soil
crop. All apparent
electrical conductivity readings from the EM38 were inspected for out
were produced with a pixel size of 35 cm × 40 cm. For all four field campaigns, spatialof range, including
negative and
distribution zero,were
maps readings. Electrical
generated for conductivity
four depthsreadings
of farmfrom each
soil at thesurvey station
selected areawere
of inter-
corrected for temperature effects using [56] Equation (1) given as:
est, with the first layer representing the root zone of the barley crop. All apparent electrical
conductivity readings from the EM38 were inspected for 26.815
out
T of range, including negative
ECa( T25) = ECa[0.4479 + 1.3801e ] (1)
and zero, readings. Electrical conductivity readings from each survey station were cor-
rected forECa
where temperature effectsconductivity
is the apparent using [56] Equation
measured(1)
at given
a pointas:and T is the mean daily
temperature with depth 0–1 m obtained from Dirk Friedhelm Mercker (DFM) probes
located in the field. 𝐸𝐶𝑎 𝑇25 = 𝐸𝐶𝑎 0.4479 1.3801𝑒 . (1)

2.2.4.𝐸𝐶𝑎
where is the apparent
Assessment conductivity measured at a point and T is the mean daily
of Soil Salinity
temperature with depth 0–1 m
Soil salinity within the studyobtained from
area was Dirk based
assessed Friedhelm
on theMercker (DFM) probes
salinity, alkalinity, and lo-
cated in theoffield.
sodicity soils in South Africa listed in Table 1 [57].

2.2.4. Assessment of Soil Salinity

Soil salinity within the study area was assessed based on the salinity, alkalinity, and
sodicity of soils in South Africa listed in Table 1 [57].
Water 2023, 15, 1911 7 of 23

Table 1. Salinity, alkalinity, and sodicity of South African soils.

Soil Class EC Concentration (mS/m)

Non-Saline <200
Slightly Saline 200−400
Moderately Saline 800−1600
Strongly Saline >1600
Saline-Sodic (Non-Alkaline) >400 pH < 8.5
Saline-Sodic (Alkaline) >400 with pH > 8.5
Sodic <400 with pH > 8.5

2.2.5. Conversion from ECa to SWC

From the interpolated electrical conductivity spatial distribution layers, soil water
content was estimated by establishing a calibration relationship between in situ soil water
content and soil electrical conductivity. A power regression model was developed from
EM38 field campaign data to estimate the soil water content percentage in four layers of
each campaign. In the developed calibration equation, electrical conductivity was set as
the independent variable, while soil water content was selected as a dependent variable.
The relationship between soil water content (θ) and soil electrical conductivity (ECa) was
expressed as:
θ = ECa B (2)
where θ represents the soil water content, and ECa is the apparent electrical conductivity;
to determine θ, Equation (2) was applied to the ECa raster files in the raster calculator in
the ArcMap environment to produce θ maps.

2.2.6. Data Processing and Evaluation Approach

The interpolation approach was used to process and develop the spatial distribu-
tion maps of soil electrical conductivity, while the statistical approach for evaluating the
relationships between the apparent electrical conductivity (ECa), soil water content (θ),
and other soil attributes were used. The relationships were established using regression
analysis. A correlation coefficient (r) with a coefficient of determination (R2 ) values was
established in every relationship between soil electrical conductivity and soil attributes,
as the primary measure or model of quality of fit. Furthermore, correlation analysis using
Pearson’s correlation (p-value) was conducted to understand the relationship between
soil water content (θ) and the soil electrical conductivity, in order to assess whether the
relationship had a statistically significant difference.

3. Results
3.1. Assessment of Soil Electrical Conductivity to Determine Soil Water Content
3.1.1. Spatial Distribution of Soil Electrical Conductivity
The mapping of apparent soil electrical conductivity was conducted at a selected area
at a farm scale using an electromagnetic induction instrument (EM38-MK), integrated data
with an geospatial interpolation inverse distance weighting method, and ground-based
sampling approaches with a focus on using electrical conductivity in determining soil water
content at different depths. The spatial variability maps of soil electrical conductivity within
the study area are shown in Figures 4–7. Variability in the soil’s electrical conductivity
within the selected area varied both spatially and with depth. To minimize sampling of soil
and measuring of soil water content manually at each depth during each campaignfrom
the 22 September 2020 EM38 field campaign, to the last EM38 field campaign on 14 October
2020 (Figures 4–7)—soil water content measurements were only conducted on the first top
layer (0–25 cm). These measurements were used to quantify the variability of the spatial
soil water content on other layers by calibration from the first day of the EM38 campaigns.
 of the EM38 campaigns. Higher values were obtained from 0.75 m to 1 m depths, while
the upper layers between 0 and 0.50 m had low values. The low values obtained could
possibly be related to the root zone water abstraction by the crops and possible rapid soil
Water 2023, 15, 1911 water evaporation as a result of loose soils due to tillage on the near surface layers, char- 8 of 23
acterized as sandy-loam as described in the study area description. The bottom layers with
high conductivity can be related to high water content, which slowly responds to the near
surface activities, including root zone water extraction by roots and top surface atmos-
Higher values were obtained from 0.75 m to 1 m depths, while the upper layers between 0
pheric demand. In August, the EM38 ECa response values ranged between 1.1 mS/m and
and 0.50 m had low values. The low values obtained could possibly be related to the root
11.85 mS/m, with an average soil water content of 6.63%. On 22 September 2020, the ECa
zone water abstraction by the crops and possible rapid soil water evaporation as a result of
values ranged between 5.1 mS/m and 19 mS/m at a 16.26% soil water content average. The
loose soilsfrom
results due 23to tillage
Septemberon the near
2020 surface
show layers,
that the characterized
ECa readings rangedas sandy-loam
from 7.7 mS/m as described
to 22.4
in the study area description. The bottom layers with high conductivity
mS/m at an average soil water content of 18.77%, while predictions of the last campaign can be related to
high water content, which slowly responds to the near surface activities,
on 14 October 2020 showed ECa ranging from 1.1 mS/m to 18.6 mS/m, with 5.33% average including root
zonesoilwater
waterextraction
content. by roots and top surface atmospheric demand. In August, the EM38
ECa response
The soilvalues ranged
electrical between
conductivity 1.1 in
shown mS/m
Figureand 11.85shows
4 clearly mS/m, with an pattern
a variability average soil
of ECa
water at four
content ofdifferent
6.63%. On layers
22 between
September0 m 2020,
and 1 the
m depths. It is clear
ECa values that the
ranged ECa at each
between 5.1 mS/m
andlayer differs at
19 mS/m from that of soil
a 16.26% otherwater
layers; this is due
content to theThe
average. fact results
that ECafrom
and 23
anySeptember
other soil 2020
showproperty
that thevary
ECaspatially no matter
readings rangedhow small
from 7.7the area is.
mS/m toIt22.4
is also
mS/m evident that
at an the range
average of water
soil
ECa during this survey period was between 1.1 and 11.85 mS/m,
content of 18.77%, while predictions of the last campaign on 14 October 2020 showed which can be linked to ECa
soil wetness or dryness. Low ECa in this context can be a sign of low soil water content, as
ranging from 1.1 mS/m to 18.6 mS/m, with 5.33% average soil water content.
described in the works of Bai et al. [57] and Turkeltaub et al. [58].

Figure
Figure 4. 4. Spatialvariability
Spatial variability of
ofsoil
soilelectrical conductivity
electrical for four
conductivity for depths ((A–D)((A–D)
four depths represent top to top to
represent
bottom layers on 11 August 2020).
bottom layers on 11 August 2020).
The soil electrical conductivity shown in Figure 4 clearly shows a variability pattern of
ECa at four different layers between 0 m and 1 m depths. It is clear that the ECa at each
layer differs from that of other layers; this is due to the fact that ECa and any other soil
property vary spatially no matter how small the area is. It is also evident that the range of
ECa during this survey period was between 1.1 and 11.85 mS/m, which can be linked to
soil wetness or dryness. Low ECa in this context can be a sign of low soil water content, as
described in the works of Bai et al. [57] and Turkeltaub et al. [58].
 Figure 5 indicates the spatial distribution of soil ECa for four depths in the surveyed
area. It is clear that higher ECa values are more pronounced on the western part of the
map, cutting out almost half of the surveyed area, while the eastern part has slightly low
values, which are higher compared to the distribution indicated in Figure 4. This is an
indication of high soil water content allowing more conductivity in soils. The ECa in this
Water 2023, 15, 1911 9 of 23
case ranges between 5 and 19.4 mS/m. On the map Figure 4A the mean SWC = 6.33%, as
compared to the Figure 5 where the mean SWC = 16.26%. It is evident that this has in-
creased, while the ECa values also increased with maximum ECa = 13.75 mS/m.

Figure 5. Spatial
5. Spatial variabilityof
variability of soil
soil electrical
electricalconductivity for four
conductivity fordepths
four ((A–D)
depthsrepresent
((A–D) top to
represent top to
Water 2023, 15, x FOR PEER Figure
REVIEW 10 of 25
bottom layers on 22 September 2020).
bottom layers on 22 September 2020).
Figure 6 depicts an increase in soil ECa across all depths of layers. The ECa ranges
between 7.7 and 20.3 mS/m. A higher ECa is evident in the northern and southern parts of
the surveyed area, while it appears to be lower in the central parts of the 0–25 m upper
layer and 0.75 m, while in 0.50 and 1 m ECa appears to be higher in the western part of the
surveyed area. The higher ECa might be attributed to high soil water content allowing
high conductivity in soils. This is due to the fact that the mean θ% continues to increase
when compared with Figure 5, where mean θ is 16.26%, and Figure 6, where it is 18.77%,
while the ECa values on the maps also increase. The relationship where ECa increases with
SWC has been narrated by Bai et al. [57].

FigureFigure 6. Spatial
6. Spatial variabilityof
variability of soil
soil electrical
electricalconductivity for four
conductivity fordepths
four ((A–D)
depthsrepresent
((A–D)top to
represent top to
bottom layers on 23 September 2020).
bottom layers on 23 September 2020).
Figure 7 shows a slight decline in soil ECa, which might be attributed to a decrease
in soil water content, limiting soils to conducting more current as opposed to wet soils. It
is evident across all layers that soil ECa varies with space, with the lower layer containing
higher ECa values compared to the rest of the layers.
Water 2023, 15, 1911 10 of 23
Water 2023, 15, x FOR PEER REVIEW 11 of 25

Figure7.7.Spatial
Figure Spatial variability
variability of
of soil
soil electrical
electricalconductivity
conductivityforfor
four depths
four ((A–D)
depths represent
((A–D) top top
represent to to
bottom layers on 14 October 2020).
bottom layers on 14 October 2020).

3.1.2. Relationship
Figure between
5 indicates In Situ distribution
the spatial Soil ElectricalofConductivity
soil ECa for and
fourIn Situ Soil
depths Water
in the surveyed
Content
area. It is clear that higher ECa values are more pronounced on the western part of the map,
cuttingPrior
out to the conversion
almost half of theof the soil’sarea,
surveyed electrical
whileconductivity
the eastern part into has
soil slightly
water content, it
low values,
was important
which are higher tocompared
clearly understand the relationship
to the distribution between
indicated the measured
in Figure 4. This soil
is anelectrical
indication
ofconductivity and soil
high soil water water
content content from
allowing morethe data obtained
conductivity during
in soils. theECa
The firstin
day calibration
this case ranges
sampling
between with19.4
5 and depth. The On
mS/m. relationships between
the map Figure 4Athethein situ SWC
mean soil electrical
= 6.33%,conductivity,
as compared to
taken
the using
Figure EC meter,
5 where and soil
the mean SWCwater contentItmeasured
= 16.26%. is evidentusing amplitude
that this domain while
has increased, reflec-the
tometry soil moisture (ADR), probe at different
ECa values also increased with maximum ECa = 13.75 mS/m. soil depths and different soil sampling
points (Figure
Figure 8a–d). an
6 depicts Theincrease
results show
in soilthat
ECasoil electrical
across conductivity
all depths is generally
of layers. The ECainflu-
ranges
between 7.7 and 20.3 mS/m. A higher ECa is evident in the northern and corresponds
enced by soil water content availability. For this reason, low conductivity southern parts
ofwith low soil water
the surveyed area,content,
while itwhile the to
appears opposite
be loweris the casecentral
in the with high values
parts of theof0–25
soil m
water
upper
content and electrical conductivity. This is evident in Figure 8a–c while
layer and 0.75 m, while in 0.50 and 1 m ECa appears to be higher in the western part of the Figure 8c shows
very lowarea.
surveyed correlation, whichECa
The higher might have
might bebeen attributed
attributed to instrument
to high soil watermeasurement
content allowingerrors.
high
conductivity in soils. This is due to the fact that the mean θ% continues to increase when
compared with Figure 5, where mean θ is 16.26%, and Figure 6, where it is 18.77%, while
the ECa values on the maps also increase. The relationship where ECa increases with SWC
has been narrated by Bai et al. [57].
Figure 7 shows a slight decline in soil ECa, which might be attributed to a decrease in
soil water content, limiting soils to conducting more current as opposed to wet soils. It is
evident across all layers that soil ECa varies with space, with the lower layer containing
higher ECa values compared to the rest of the layers.
 Water 2023, 15, 1911 11 of 23

3.1.2. Relationship between In Situ Soil Electrical Conductivity and In Situ Soil Water Content
Prior to the conversion of the soil’s electrical conductivity into soil water content, it
was important to clearly understand the relationship between the measured soil electrical
conductivity and soil water content from the data obtained during the first day calibration
sampling with depth. The relationships between the in situ soil electrical conductivity, taken
using EC meter, and soil water content measured using amplitude domain reflectometry
soil moisture (ADR), probe at different soil depths and different soil sampling points
(Figure 8a–d). The results show that soil electrical conductivity is generally influenced
by soil water content availability. For this reason, low conductivity corresponds with low
Water 2023, 15, x FOR PEER REVIEW soil water content, while the opposite is the case with high values of soil water content 12 of 25
and electrical conductivity. This is evident in Figure 8a–c while Figure 8c shows very low
correlation, which might have been attributed to instrument measurement errors.

Figure
Figure 8. Relationshipbetween
8. Relationship between inin situ
situ soil
soilelectrical
electricalconductivity andand
conductivity soilsoil
water content
water at varying
content at varying
soilsoil depths
depths at at sampledlocations,
sampled locations, (a)
(a) represent
representthetherelationship of ECa
relationship andand
of ECa Soil Soil
water content
water at 0 toat 0 to
content
0.250.25 m depth,
m depth, (b)(b) 0.25toto0.50
0.25 0.50m
m depth,
depth, (c)(c)0.50
0.50toto0.75
0.75mmdepth
depthand (d) (d)
and denotes a relationship
denotes at 0.75at 0.75
a relationship
to 1 m depth.
to 1 m depth.
3.1.3. Soil Electrical Conductivity Conversion to Soil Water Content
3.1.3. Soil Electrical Conductivity Conversion to Soil Water Content
Once the relationship between soil water content and electrical conductivity was
Once the relationship
understood, the findings between soil water
were combined intocontent and electrical
one general model. Soil conductivity
water contentwas un-
derstood, the findingsmaps
spatial distribution werewere
combined intoby
generated one generalthe
applying model. Soil water
developed modelcontent
from thespatial
relationship of soil water content with the soil electrical conductivity to spatial
distribution maps were generated by applying the developed model from the relationship distribution
maps
of soil (Figure
water 9). with the soil electrical conductivity to spatial distribution maps (Figure
content
9).
 to 1 m depth.

3.1.3. Soil Electrical Conductivity Conversion to Soil Water Content

Once the relationship between soil water content and electrical conductivity was un-
derstood, the findings were combined into one general model. Soil water content spatial
Water 2023, 15, 1911 distribution maps were generated by applying the developed model from the relationship 12 of 23
of soil water content with the soil electrical conductivity to spatial distribution maps (Figure
9).

Figure
Figure 9. Model
9. Model used used to convert
to convert ECawater
ECa to soil to soil water
content (θ).content (θ).

3.2. The Spatial Distribution of Soil Water Content

The predicted spatial variability maps of soil water content determined based on the
ECa-SWC relationship are given in Figures 10–13, and soil water content is given as a
percentage of wetness in the soil. Soil water content was found to be higher in the bottom
two layers, located at 0.75 m and 1 m depths. This is due to a high ECa shown on the spatial
distribution maps of ECa for different survey days. On 8 August 2020, the predicted soil
water content ranged between 4.15% in the top layer and increased up to 27.82% in the
bottom layer. The amount of soil water content was lower in the first layer and increased
with
Water 2023, 15, x FOR PEER REVIEW
depth (Figures 10–13). The soil water content predicted on 22 September 2020 ranged
14 of 25
between 13.85 and 39.85 percent, and soil water content increased with an increase in depth.

Figure
Figure 10. Predicted
10. Predicted θ at
θ at fourdepths
four depths ((A–D)
((A–D)represents 0.250.25
represents m tom
1mtoon
1m8 August 2020). 2020).
on 8 August

Figure 11 shows soil water content spatial distribution maps over four depth intervals,
related to the survey done on 22 September 2020. Figure 12A shows the spatial distribu-
tion of soil water content between 0 and 0.25 m depth, with high moisture being observed
on the western top and bottom of the map. The rest of the layers also indicate a shift of
high moisture towards the left side of the surveyed area. The bottom layer contains high
Water 2023, 15, 1911 13 of 23
Water 2023, 15, x FOR PEER REVIEW 15 of 25

Water 2023, 15, x FOR PEER REVIEW 16 of 25

Figure
Figure 11.11. Predictedθθat
Predicted at four
four depths
depths((A–D) represents
((A–D) 0.250.25
represents m tom1m
toon
1m22 on
September 2020). 2020).
22 September

Figure 12 portrays the spatial distribution of soil water content across various soil
depths up to 1 m. Low soil water content is evident on the right hand sections of the sur-
veyed area on the top two layers (0–0.25 m and 0.50 m), which can be attributed to the
active response of the layers relating to soil water evaporation on the top layer and crop
water use occurring on the two layers containing the crop roots. The lower layer contains
high soil water content ranging between 16% and 27%. This is lower when compared to
the same layer on other survey dates, which can be attributed to time gap between the
survey date and the last day of irrigation in the section.

Figure
Figure 12. 12. Predicted
Predicted θ θatatfour
fourdepths
depths ((A–D)
((A–D)represents 0.250.25
represents m tom1 to
m on
1m23on
September 2020). 2020).
23 September

Figure 13 shows the spatial distribution maps of soil water content under different
soil depths. It is evident that soil water content on the top layer contains the lowest soil
water content range, while the bottom layer contains the highest soil water content. The
variability in soil water content can be attributed to the exposure and shielding of soil
surface by crop canopy. The upper layer at 0.25 m depth is closer to the surface where the
crop density increases as the leaf area and biomass increases, casting protection on the soil
from fast evaporation of SWC, where water is only lost through transpiration [59]. In soils
exposed to the atmosphere, water evaporates quickly due to the atmospheric demands,
Water 2023, 15, 1911 14 of 23
Water 2023, 15, x FOR PEER REVIEW 17 of 25

Figure
Figure 13.13. Predictedθθatatfour
Predicted four depths
depths ((A–D)
((A–D)represents
represents0.25 m to
0.25 m 1tom1on
m14
onOctober 2020).2020).
14 October

3.2.1.
OnRelationship between
the following day, the Soil Water Content
23 September Measured
2020, the and Predicted
soil water content Electrical
predicted ranged
Conductivity
between 5.26 and 39.85 percent, which increased from around the upper soil layer with
depth, Theup tovalues extracted
the bottom from On
layer. the the
samelast geographic
campaign, location of measurement
conducted as the soil
on 14 October 2020, the
water content and the predicted electrical conductivity were
predicted soil water content ranged between 4.34 and 38.58 percent, indicating used to determine the corre-a rise in
soillation
waterbetween
content thepercentage
two variables, andusing linear relationships
increasing with depth. (Figure
An14a–d). During
interesting August
observation is
2020, the relationship between ECa and SWC was negative, with a coefficient of determi-
that the bottom layer contains higher soil water content than the top 0.25 m layer, which
nation R2 of 0.85 and a correlation coefficient of −0.92 (Table 2). On 22 September 2020, the
can be related to evaporation processes and crop water use on the top layer, which is the
relationship between ECa and SWC was negative, with an R2 value of 0.95 and a correla-
root zone.
tion coefficient of −0.98. On 23 September 2020 (Table 2), the relationship between ECa and
SWC Figure
was 10 depicts
positive, soilan
with water content
R2 value variability
of 0.71 across different
and a correlation coefficient soil
ofdepths, undertaken
0.84 (Table 2).
during
Altdorff the et
8 August 2020 survey.
al. [41] claim Soil water
that a number contentthat
of variables during
varythis
from survey
regionranged between
to region, in- 4%
andcluding
28%. The top layer atregulate
soil treatments, 0.25 m the
depth showed between
relationship very lowECa soiland
water
SWC. content across
The silica ferti-almost
thegation,
entirewhich
survey wasarea, which
applied at acan
depthbe ofattributed
3 mm on to 23 soil drying2020,
September up as mayit moves away from
be the reason
thefor the shift date.
irrigation from aAn negative association
interesting betweenisECa
observation and
that theSWC to a positive
bottom relationship.
layer contains higher soil
Thiscontent
water may have affected
than soil conductivity
the bottom layer, whichand moisture
can be regulation.
related to The soil’s electrical
evaporation con- and
processes
ductivity increases as a result of silica absorbing
crop water use on the top layer, which is the root zone. cations from the soil medium, particu-
larly Ca, mg,
Figure and K. soil
11 shows These cations
water function
content as carriers
spatial of electrical
distribution mapscurrent in thedepth
over four soil [60].
intervals,
Silica was applied to increase the strength of the barley by thickening the cell walls, with
related to the survey done on 22 September 2020. Figure 12A shows the spatial distribution
the aim of preventing lodging [61]. On 14 October 2020, the relationship between ECa and
of soil water content between 0 and 0.25 m depth, with high moisture being observed
SWC was positive, with an R2 value of 0.75 and a correlation coefficient of 0.87 (Table 2).
on For
theall
western top and bottom of the map. The rest of the layers also indicate a shift of
the survey dates, the relationship between ECa and SWC portrayed a statistically
high moisture towards the left side
significant difference p ≤ 0.05 (Table 2). ofTherefore,
the surveyed area. show
the results The bottom
that high layer contains high
soil electrical
moisture zones compared to the rest of
conductivity was a function of soil water content. the layers, with soil water content ranging between
23% and 39.8%.
Figure 12 portrays the spatial distribution of soil water content across various soil
depths up to 1 m. Low soil water content is evident on the right hand sections of the
surveyed area on the top two layers (0–0.25 m and 0.50 m), which can be attributed to the
active response of the layers relating to soil water evaporation on the top layer and crop
water use occurring on the two layers containing the crop roots. The lower layer contains
Water 2023, 15, 1911 15 of 23

high soil water content ranging between 16% and 27%. This is lower when compared to the
same layer on other survey dates, which can be attributed to time gap between the survey
date and the last day of irrigation in the section.
Figure 13 shows the spatial distribution maps of soil water content under different
soil depths. It is evident that soil water content on the top layer contains the lowest soil
water content range, while the bottom layer contains the highest soil water content. The
variability in soil water content can be attributed to the exposure and shielding of soil
surface by crop canopy. The upper layer at 0.25 m depth is closer to the surface where the
crop density increases as the leaf area and biomass increases, casting protection on the soil
from fast evaporation of SWC, where water is only lost through transpiration [59]. In soils
exposed to the atmosphere, water evaporates quickly due to the atmospheric demands,
while soils shielded by the crop canopy retain moisture for longer periods [59]. The top
layers of the surveyed depths contain lower soil water compared to the bottom two layers
because they are the root zone levels where crops uptake their water.

3.2.1. Relationship between the Soil Water Content Measured and Predicted Electrical Conductivity
The values extracted from the same geographic location of measurement as the soil
water content and the predicted electrical conductivity were used to determine the cor-
relation between the two variables, using linear relationships (Figure 14a–d). During
August 2020, the relationship between ECa and SWC was negative, with a coefficient of
determination R2 of 0.85 and a correlation coefficient of −0.92 (Table 2). On 22 September
2020, the relationship between ECa and SWC was negative, with an R2 value of 0.95 and a
correlation coefficient of −0.98. On 23 September 2020 (Table 2), the relationship between
ECa and SWC was positive, with an R2 value of 0.71 and a correlation coefficient of 0.84
(Table 2). Altdorff et al. [41] claim that a number of variables that vary from region to
region, including soil treatments, regulate the relationship between ECa and SWC. The
silica fertigation, which was applied at a depth of 3 mm on 23 September 2020, may be
the reason for the shift from a negative association between ECa and SWC to a positive
relationship. This may have affected soil conductivity and moisture regulation. The soil’s
electrical conductivity increases as a result of silica absorbing cations from the soil medium,
particularly Ca, mg, and K. These cations function as carriers of electrical current in the
soil [60]. Silica was applied to increase the strength of the barley by thickening the cell walls,
with the aim of preventing lodging [61]. On 14 October 2020, the relationship between
ECa and SWC was positive, with an R2 value of 0.75 and a correlation coefficient of 0.87
(Table 2). For all the survey dates, the relationship between ECa and SWC portrayed a
statistically significant difference p ≤ 0.05 (Table 2). Therefore, the results show that high
soil electrical conductivity was a function of soil water content.
Table 2. Statistical summary of the relationship between the measured soil water content and
estimated electrical conductivity.

Day of the Year Model r p-Value

11 August 2020 θ = −3.93 ECa + 12.54 −0.92 0.04
22 September 2020 θ = −3.10 ECa + 39.83 −0.98 0.02
23 September 2020 θ = 1.09 ECa − 10.91 0.84 0.05
14 October 2020 θ = 0.67 ECa + 3.99 0.87 0.04
 Water15,
Water 2023, 2023, 15, x FOR PEER REVIEW
1911 18 of 2516 of 23

Figure 14.14.The
Figure Therelationship betweenpredicted
relationship between predicted
soilsoil electrical
electrical conductivity
conductivity and measured
and measured soil watersoil wa-
ter content,
content,where
where (a–d)
(a–d) represents
represents (11 August
(11 August 2020, 2020, 22 September
22 September 2020, 232020, 23 September
September 2020, and2020,
14 and
October 2020)
14 October 2020)dates
datesofof
thethe
survey.
survey.

3.2.2. The2. Relationship

Table between
Statistical summary Soil
of the Electricalbetween
relationship Conductivity and Soil
the measured soil Physiochemical
water content andProperties
esti-
mated electrical conductivity.
A correlation matrix was developed to evaluate the influence of soil properties on
Day of the Year electrical conductivity (ECa)
Modeland (ECe) readings (Table r 3). The tablep-Value
shows a very low cor-
11 August 2020 relation between θECe,
= −3.93 ECa + 12.54(mg), and calcium
magnesium −0.92(Ca), with higher0.04positive correlation
22 September 2020between ECe and θ sodium
= −3.10 ECa + 39.83
(Na), and a moderate −0.98
correlation between0.02 ECe and potassium
23 September 2020(K). Soil pH and θECe showed
= 1.09 a relatively low correlation.
ECa − 10.91 0.84 The correlation
0.05 matrix below
14 October 2020 depicts how physiochemical
θ = 0.67 ECaproperties
+ 3.99 0.87to the ECa measured
of soil relate 0.04 by the EM38 and
an ECe paste of sampled soil with the purpose of obtaining properties, which have higher
3.2.2. The
influence in Relationship between Soil Electrical Conductivity and Soil Physiochemical
soil ECa variability.
Properties
Table 3.A correlationmatrix
Correlation matrixofwas
soil developed to evaluate
physiochemical the and
properties influence of soilcations
extractable properties
with on
electrical
electrical conductivity
conductivity ECa and ECe.(ECa) and (ECe) readings (Table 3). The table shows a very low
correlation between ECe, magnesium (mg), and calcium (Ca), with higher positive corre-
ECe lation betweenKECe and sodium
Na Mg (Na), and Ca a moderate correlation
PH between
SWCECe and po- ECa
tassium (K). Soil pH and ECe showed a relatively low correlation. The correlation matrix
ECe 1.00
below depicts how physiochemical properties of soil relate to the ECa measured by the
Na 0.88 1.00
EM38 and an ECe paste of sampled soil with the purpose of obtaining properties, which
K 0.49 have higher influence
0.77 1.00 in soil ECa variability.

Mg 0.49 0.52 0.43 1.00

Ca 0.05 0.41 0.75 0.03 1.00
PH 0.09 −0.13 −0.31 −0.01 −0.28 1.00
SWC −0.15 −0.02 0.03 −0.16 0.18 0.51 1.00
Eca 0.54 0.05 0.07 −0.56 0.37 −0.08 −0.13 1.00
 SWC −0.15 −0.02 0.03 −0.16 0.18 0.51 1.00
Eca 0.54 0.05 0.07 −0.56 0.37 −0.08 −0.13 1.00

3.3. Irrigation Events

Figure 15 shows soil water content throughout the barley cropping season, obtained
Water 2023, 15, 1911 17 of 23
from a DFM soil water content probe. Green arrows indicate the gain in SWC%, which
indicates irrigation events. Conversely, a decline in gradient between irrigation events de-
picts soil water content drawn down through crop water use, indicated by the red line
3.3. Irrigation
relapse. SoilEvents
water content in the figure below was measured between 0–0.5 m depth,
Figure
which was15 theshows
activesoil water
layer andcontent throughouttothe
highly responsive barley cropping
irrigation and cropseason, obtained
water use.
from aTheDFM soil water
quantity content
of water probe.
given to theGreen arrowseach
crop before indicate
EM38the gain in
survey SWC%,
is given which4,
in Table
indicates
including irrigation
the date events.
and timeConversely, a decline
of each survey in gradient
campaign. On mostbetween irrigation
irrigation dates,events
10 mm
depicts
drops were given to crops while a 3 mm amount was given during fertigation. Higherline
soil water content drawn down through crop water use, indicated by the red ECa
relapse.
and SWC Soilwere
water content during
obtained in the figure below
survey dayswas measured
closer between 0–0.5
to the irrigation m while
events, depth,low
which
ECa
was
andthe
SWCactive layer
values andobtained
were highly responsive to irrigation
on days further and crop
away from water
the last use.
days of irrigation.

Figure 15.Soil
Figure15. Soilwater
watercontent
contentdepicting
depictingirrigation
irrigationevents.
events.

The
Table 4. quantity
Irrigationof water
and EM38given to the
survey crop
dates before
with time,each EM38the
including survey is given
amount in Table
of water 4,
irrigated
including
throughoutthe
thedate and time of each survey campaign. On most irrigation dates, 10 mm
season.
drops were given to crops while a 3 mm amount was given during fertigation. Higher ECa
Irrigation Date and Time
and SWC were obtainedSurveyduring
Date and Timedays closer to
survey Water Irrigated events, while
the irrigation Condition
low ECa
5 August 2020/14:00and SWC values were 11 August 2020/15:00
obtained on days further away from 10 mm Dry
the last days of irrigation.
21 September 2020/08:00 22 September 2020/15:00 10 mm Wet
Table 4. Irrigation23and
23 September 2020/12:00 EM38 survey
September dates with time, including
2020/16:00 3 mmthe amount Wetof water irrigated
(Fertigation)
throughout the season.
12 October 2020/21:00 14 October 2020/16:00 10 mm Wet-Dry

Irrigation Date and Time Survey Date and Time Water Irrigated Condition
3.4. Irrigation Water and Soil Quality
5 August 2020/14:00 11 August 2020/15:00 10 mm Dry
Table 5 shows the water quality analysis for the irrigation water used in the farm from
21 September 2020/08:00 the22canal.
September 2020/15:00
The EC 10 mmcalcium, and potassium were
value, magnesium, sodium, Wet analyzed. The
23 September 2020/12:00 sodium absorption
23 September ratio (SAR) value obtained
2020/16:00 3 mm from the irrigation water
Wet is also reported
(Fertigation)
12 October 2020/21:00 14 October 2020/16:00 10 mm Wet-Dry

3.4. Irrigation Water and Soil Quality

Table 5 shows the water quality analysis for the irrigation water used in the farm from
the canal. The EC value, magnesium, sodium, calcium, and potassium were analyzed. The
sodium absorption ratio (SAR) value obtained from the irrigation water is also reported
here. The purpose of this analysis was to compare the EC values between irrigation water
and soil ECa. From the soil analysis, the SAR average value of soils in the study area was
found to be 0.34 mileq/L.

Table 5. Irrigation water quality analysis result.

Measured
SAR
Parameter EC mS/m Ca mg/L Mg mg/L Na mg/L K mg/L
mileq/L
Unit
Value 66.6 44 22 51 10.91 1.108
Water 2023, 15, 1911 18 of 23

4. Discussion
This study focused on assessing the applicability of a non-destructive soil sensing
electromagnetic induction device (EM38-MK) in determining the soil’s apparent electrical
conductivity, with the aim of using it as a proxy to determine soil water content variability
post irrigation events, and salinity state of the soil in the Vaalharts irrigation scheme at farm
level. To achieve spatial distribution maps of the soil’s electrical conductivity at different
depths, the geospatial inverse distance weighting interpolation method was used. The
model developed based on the relationship between soil water content and soil moisture
was used to convert soil electrical conductivity layers into soil water content.
Agricultural soils have different structures, nutritional composition, water holding
capacities, and strengths. Electrical conductivity laboratory and field experiments were
carried out to investigate their relationship with other soil properties known to have
influence on them. Linear regression analysis was conducted between the measured
and predicted values of both soil water content and the soil’s electrical conductivity to
evaluate the capabilities of the EM38 in determining the conductivity of the area and
soil water content based on statistical measures. The main influencing factor of high soil
ECa in this study was SWC%. This is supported by the laboratory tests relating ECa and
physiochemical properties, which showed fewer correlations, whereas the correlation of
ECa and SWC was high, with R2 values ranging between 0.71 and 0.95 during different
wetness conditions, provided in Figure 8. Similar findings have been reported in some
studies [62,63]. These findings clearly demonstrate that as the number of days increase,
starting from the days of irrigation, SWC decreases, while ECa also has the same pattern.
Regression based analysis between the apparent electrical conductivity and soil water
content in this study shows higher correlations with an R2 = 0.87. This illustrates the
ability of the EM38-based ECa measurements in estimating the soil water content of a given
area. Wong and Asseng [63] reported similar correlations with an R2 of 0.78 between ECa
and available water for plants, while Reedy and Scanlon [62] determined the volumetric
water content with a linear regression based model obtaining an R2 of 0.80. Both are not
far from what is demonstrated in the findings of this study. These findings are further
supported by the work of Sherlock and McDonnell [64] who, upon using the EM38 and
EM31 in water table determination, found that simple linear regression explained more
than 80% of variations. High correlation coefficients (r) were obtained between measured
soil water content and estimated ECa (−0.92, −0.98, 0.84, and 0.87) making it evident that
the relationship between ECa and soil water content can be used as a surrogate in order to
estimate one of the two soil variables. High ECa values can be a sign of salinity according
to Bai et al. [57], however, based on our findings this was not the case. Furthermore, the
relationship between ECa and the physiochemical properties were low, leaving SWC as the
main driver of ECa.
The findings of this study demonstrate that the soil water content is the main driver
that allows the flow of electrical current in the soil medium, making it easier for the
electromagnetic induction device to determine the soil’s apparent electrical conductivity.
Such findings are also reported by others [36,43,65]. The linear relationship between soil
water content and the soil’s electrical conductivity made it possible to easily convert ECa
(mS/m) into SWC%, influenced largely by irrigation in the area. As the days got further
away from the irrigation days, the decline in soil electrical conductivity was evident.
This was based on the fact that soil water content declines as the crop uses water and
evaporation takes place in the absence of precipitation, which makes it easier for farmers to
determine and track their soil water depletion. This provides an accurate basis for irrigation
scheduling and quantifying crop water use.
Misra and Padhi [66], when using the EM38 for soil water distributions, found that
the best time to use the EM38 for soil water content mapping was when the ECa was highly
influenced by soil water content. In this study, we found that with the sandy-loam soils in
the arid environment where the study was done, ECa can only determine soil water content
under the equality: ECa > 0 < 113 mS/m. For that reason, ECa cannot determine SWC if
Water 2023, 15, 1911 19 of 23

it is less than or equal to zero, or if it exceeds 113 mS/m as per the limits of the model
developed in this study.
On assessing the relationship between ECa and other soil parameters, the findings
demonstrated that magnesium (mg) correlated slightly above average with ECa, while
other physiochemical properties showed very low correlation, which provides information
that ECa in our study area is not directly influenced by other physiochemical properties
of the soil apart from soil water content. These observations provide evidence that the
EM38-based ECa measurement can accurately determine SWC in the area. Kurt Heil and
Schmidh [67], on assessing the influence of soil properties on ECa, also found similar results
on the basis that the relationship between ECa and soil properties was very weak in their
surveys, making soil water content the main driver of ECa.
The SAR average value of soils in the study area was found to be 0.34 mileq/L, which
is low sodium when SAR ranges between 0–10. As such, it indicates a low risk of irrigation
water being used as described by Burger and Celkova [68]. Results from irrigation water
quality analysis showed a higher value in SAR with 1.11 mileq/L, which is higher than
the value obtained in the soil, however, although the SAR value in water is higher than
the one in soil, it still falls within the safe irrigation water boundary according to Burger
and Celkova [68]. The EC obtained in water is also higher than the one obtained from soil,
which is still within the safe saline levels according to another study [56].
Furthermore, on analyzing the spatial distribution maps of ECa, we compared the
observed, predicted, and local salinity scale values of South Africa found in one study [56]
to assess if there was any evidence of salinity in the area as reported in its history. We found
that soils within the study area had no sign of salinity, with ECa values < 100 mSm−1 . The
findings of this study relate to the findings of other studies [43,65], which also used the
EM38-mk to estimate soil moisture based on ECa measurements, and found similar trends
in terms of correlations and behavior of ECa under given SWC conditions.
The use of a non-destructive electromagnetic induction device has the potential to
reduce the time and costs of installing point sensors in the field or avoid the necessity of
making actual measurements using portable sensors, particularly in solving soil water re-
lated problems in agriculture, hydrogeological environments, and the natural environment
as a system [43]. The capability of the EM38 to map soil parameters with depth that covers
most root zones of common crops that are grown around the world offers the opportunity to
solve the spatial limitations of point fixed data [43]. In this study, the findings demonstrate
that SWC in the top layer is always lower when compared to the bottom layers, which can
be related to the fact that the top 0–0.25 m soil layer is occupied by roots, although some
crops can go deeper with growth stages.
The spatial variability of soil electrical conductivity and soil water content was success-
fully determined, and the generated maps come with a number of benefits. These findings
can be used to assist with scheduling irrigation for irrigation technicians, who would be
able to understand the soil water content variability beyond the naked eye on the surface
and eliminate the issue of providing information using spatial disjoined figures. The pro-
duced maps in precision agriculture, where spot soil treatment is done using RTK-linked
auto-machineries, could be the used as base maps to treat only areas which are necessary,
saving the costs of soil treatment resources and labor costs.
Although the EM38 is capable of mapping soil ECa and SWC, there are limitations
of such devices; in this study the existence of the metallic pivot irrigation tower, fencing,
metallic storage facilities, and metal pegs in the field restricted the area of survey. The
EM38 devices can only be used for limited periods unless they are vehicle towed. Each
soil type requires local calibrations when the EM38 is used as a proxy to determine soil
water content because the EM38 does not directly determine soil water content. The soil’s
electrical conductivity is the main component that is measured, and can be influenced by
different physiochemical properties [69]. It was found that the model can predict soil water
content based on the soil’s apparent soil electrical conductivity if ECa > 0 < 113 mS/m for
our study area.
Water 2023, 15, 1911 20 of 23

The model developed in this study was based on a small section of the farm (100 m
by 70 m) to avoid the effects of metallic objects during the electromagnetic survey. For
that reason, not many samples were taken from the selected area. The number of samples,
however, does not have an impact on the performance of our model in retrieving SWC from
ECa; a similar case has been reported in one study [70], although they used linear models.
Future studies are recommended to incorporate more samples and cover an extensive area
without signal obstacles to assess the model performance. With the benefits of localized
scales, this study recommends the use of EM38 surveys in conjunction with spatial covering
sensors that are capable of estimating soil water content, including the use of UAV-based
sensors, which are capable of providing high resolution images and information on soil
water content at shallow depths.

5. Conclusions
Soil water content variability at different depths under different wetness conditions
was successfully estimated using electromagnetic induction as a proxy. The determination
of soil salinity using soil electrical conductivity was also successful, finding that the study
area soils are below the saline levels of ECa. The water analysis results also indicated
EC levels that are below the salinity levels. Based on the integrated analysis of different
soil physiochemical parameters, it was observed that ECa in the study area was more
influenced by soil water content. Soil moisture using direct point specific techniques is
usually reported using spatial disjoined figures, which do not really provide information
that can be used in variable irrigation treatments. The electromagnetic induction instrument
(EM38) is worth using to quantify the spatial variability of soil water content by measuring
the spatial variability of soil electrical conductivity. A significant observation, in the
simulation of soil water content variability based on soil electrical conductivity, was that
soil electrical conductivity response to the instrument occurs when there is enough water
content available in the soil to move the charges around. Although the results proved
that soil electrical conductivity measurement could be used as a surrogate for determining
soil water content using EM38, calibration for every soil type is required for better results.
On the other hand, the instrument is expensive, and the interpretation of data requires
scientific analysts rather than simply people with agricultural knowledge. We recommend
future research to look at large field coverage with more point samples. The approaches
and methods used in this study can be replicated in other areas to attain an in-depth
understanding of the soil water dynamics as well as understanding the salinity status of
the soil.

Author Contributions: Conceptualization, methodology, analysis, and original draft preparation

P.E.R.; writing—review and editing, M.A.M.A.E.; writing—review and editing, E.A.; writing—review
and editing, J.G.C.; writing—review and editing, G.L.; review and editing, funding acquisition,
M.A.M.A.E. and J.G.C.; fieldwork operations and editing, P.E.R., E.B.E. and M.A.M.A.E. All authors
have read and agreed to the published version of the manuscript.
Funding: This research was funded by the Agricultural Research Council-Natural Resources and
Engineering (ARC-NRE) Project number: P07000128, Department of Science and Innovation; Na-
tional Research Foundation and the Water Research Commission of South Africa Project number:
C2022/2023-00978.
Data Availability Statement: Available on request.
Acknowledgments: We would like to thank Fagan Scheepers and the entire South African Barley
Institute (SABBI) for permitting us to use their experimental farm for this study, we acknowledge the
Water Research Commission of South Africa (WRC) for additional support in this work. Finally we
would also like to acknowledge the creditable work done by the reviewers and editors at water and
Thomas Fyfield for helping with in-house review.
Conflicts of Interest: The authors declare no conflict of interest.
Water 2023, 15, 1911 21 of 23

References
1. Rodriguez-Iturbe, I.; Porporato, A.; Laio, F.; Ridolfi, L. Plants in Water-Controlled Ecosystems: Active Role in Hydrologic
Processes and Response to Water Stress. Adv. Water Resour. 2001, 24, 695–705. [CrossRef]
2. Verstraeten, W.; Veroustraete, F.; Feyen, J. Assessment of Evapotranspiration and Soil Moisture Content Across Different Scales of
Observation. Sensors 2008, 8, 70–117. [CrossRef] [PubMed]
3. Wang, L.; D’Odorico, P.; Evans, J.P.; Eldridge, D.J.; McCabe, M.F.; Caylor, K.K.; King, E.G. Dryland Ecohydrology and Climate
Change: Critical Issues and Technical Advances. Hydrol. Earth Syst. Sci. 2012, 16, 2585–2603. [CrossRef]
4. Williams, A.G.; Ternan, J.L.; Fitzjohn, C.; De Alba, S.; Perez-Gonzalez, A. Soil Moisture Variability and Land Use in a Seasonally
Arid Environment. Hydrol. Process. 2003, 17, 225–235. [CrossRef]
5. Hrozencik, R. Trends in U.S. Irrigated Agriculture: Increasing Resilience Under Water Supply Scarcity. SSRN Electron. J. 2021, 229,
1–55. [CrossRef]
6. Taghvaeian, S.; Andales, A.A.; Allen, L.N.; Kisekka, I.; O’Shaughnessy, S.A.; Porter, D.O.; Sui, R.; Irmak, S.; Fulton, A.; Aguilar, J.
Irrigation Scheduling for Agriculture in the United States: The Progress Made and the Path Forward. Trans. ASABE 2020, 63,
1603–1618. [CrossRef]
7. Kumar, H.; Srivastava, P.; Ortiz, B.V.; Morata, G.; Takhellambam, B.S.; Lamba, J.; Bondesan, L. Field-Scale Spatial and Temporal
Soil Water Variability in Irrigated Croplands. Trans. ASABE 2021, 64, 1277–1294. [CrossRef]
8. Foster, T.; Mieno, T.; Brozović, N. Satellite-Based Monitoring of Irrigation Water Use: Assessing Measurement Errors and Their
Implications for Agricultural Water Management Policy. Water Resour. Res. 2020, 56, e2020WR028378. [CrossRef]
9. Seguin, B.; Courault, D.; Guérif, M. Surface Temperature and Evapotranspiration: Application of Local Scale Methods to Regional
Scales Using Satellite Data. Remote Sens. Environ. 1994, 49, 287–295. [CrossRef]
10. Leone, M. (INVITED)Advances in Fiber Optic Sensors for Soil Moisture Monitoring: A Review. Results Opt. 2022, 7, 100213.
[CrossRef]
11. Salouti, M.; Khadivi Derakhshan, F. Biosensors and Nanobiosensors in Environmental Applications. In Biogenic Nano-Particles
and Their Use in Agro-Ecosystems; Ghorbanpour, M., Bhargava, P., Varma, A., Choudhary, D.K., Eds.; Springer: Singapore, 2020;
pp. 515–591. ISBN 9789811529849.
12. Songara, J.C.; Patel, J.N. Calibration and Comparison of Various Sensors for Soil Moisture Measurement. Measurement 2022,
197, 111301. [CrossRef]
13. Ochsner, T.E.; Cosh, M.H.; Cuenca, R.H.; Dorigo, W.A.; Draper, C.S.; Hagimoto, Y.; Kerr, Y.H.; Larson, K.M.; Njoku, E.G.;
Small, E.E.; et al. State of the Art in Large-Scale Soil Moisture Monitoring. Soil Sci. Soc. Am. J. 2013, 1, 1888–1919. [CrossRef]
14. Narayan, U. Retrieval of Soil Moisture from Passive and Active L/S Band Sensor (PALS) Observations during the Soil Moisture
Experiment in 2002 (SMEX02). Remote Sens. Environ. 2004, 92, 483–496. [CrossRef]
15. Rasheed, M.W.; Tang, J.; Sarwar, A.; Shah, S.; Saddique, N.; Khan, M.U.; Imran Khan, M.; Nawaz, S.; Shamshiri, R.R.; Aziz, M.; et al.
Soil Moisture Measuring Techniques and Factors Affecting the Moisture Dynamics: A Comprehensive Review. Sustainability 2022,
14, 11538. [CrossRef]
16. Arregui, L.M.; Quemada, M. Drainage and Nitrate Leaching in a Crop Rotation under Different N-Fertilizer Strategies: Application
of Capacitance Probes. Plant Soil 2006, 288, 57–69. [CrossRef]
17. Starr, J.L.; Paltineanu, I.C. Soil Water Dynamics Using Multisensor Capacitance Probes in Nontraffic Interrows of Corn. Soil Sci.
Soc. Am. J. 1998, 62, 114–122. [CrossRef]
18. Yinglan, Y.; Wang, G.; Hu, P.; Lai, X.; Xue, B.; Fang, Q. Root-Zone Soil Moisture Estimation Based on Remote Sensing Data and
Deep Learning. Environ. Res. 2022, 212, 113278. [CrossRef]
19. Rowlandson, T.L.; Berg, A.A.; Bullock, P.R.; Ojo, E.R.; McNairn, H.; Wiseman, G.; Cosh, M.H. Evaluation of Several Calibration
Procedures for a Portable Soil Moisture Sensor. J. Hydrol. 2013, 498, 335–344. [CrossRef]
20. Okasha, A.M.; Ibrahim, H.G.; Elmetwalli, A.H.; Khedher, K.M.; Yaseen, Z.M.; Elsayed, S. Designing Low-Cost Capacitive-Based
Soil Moisture Sensor and Smart Monitoring Unit Operated by Solar Cells for Greenhouse Irrigation Management. Sensors 2021,
21, 5387. [CrossRef]
21. Zhang, F.; Zhang, L.-W.; Shi, J.-J.; Huang, J.-F. Soil Moisture Monitoring Based on Land Surface Temperature-Vegetation Index
Space Derived from MODIS Data. Pedosphere 2014, 24, 450–460. [CrossRef]
22. Das, M.; Sethi, R.R.; Sahoo, N. Evaluation and Integration of Soil Salinity and Water Data for Improved Land Use of Underpro-
ductive Coastal Area in Orissa. Irrig. Drain. 2010, 59, 621–627. [CrossRef]
23. Svetlitchnyi, A.A.; Plotnitskiy, S.V.; Stepovaya, O.Y. Spatial Distribution of Soil Moisture Content within Catchments and Its
Modelling on the Basis of Topographic Data. J. Hydrol. 2003, 277, 50–60. [CrossRef]
24. Chartzoulakis, K.; Bertaki, M. Sustainable Water Management in Agriculture under Climate Change. Agric. Agric. Sci. Procedia
2015, 4, 88–98. [CrossRef]
25. Entekhabi, D.; Njoku, E.G.; O’Neill, P.E.; Kellogg, K.H.; Crow, W.T.; Edelstein, W.N.; Entin, J.K.; Goodman, S.D.; Jackson, T.J.;
Johnson, J.; et al. The Soil Moisture Active Passive (SMAP) Mission. Proc. IEEE 2010, 98, 704–716. [CrossRef]
26. Liu, Y.; Yang, Y. Advances in the Quality of Global Soil Moisture Products: A Review. Remote Sens. 2022, 14, 3741. [CrossRef]
27. Xu, C.; Qu, J.; Hao, X.; Cosh, M.; Prueger, J.; Zhu, Z.; Gutenberg, L. Downscaling of Surface Soil Moisture Retrieval by Combining
MODIS/Landsat and In Situ Measurements. Remote Sens. 2018, 10, 210. [CrossRef]
Water 2023, 15, 1911 22 of 23

28. Zhang, N.; Wang, M.; Wang, N. Precision Agriculture—A Worldwide Overview. Comput. Electron. Agric. 2002, 36, 113–132.
[CrossRef]
29. Yongnian Zeng, Z.F. Assessment of Soil Moisture Using Landsat ETM+ Temperature/Vegetation Index in Semiarid Environment.
In Proceedings of the IEEE International Geoscience and Remote Sensing Symposium, IGARSS ’04, Anchorage, AK, USA, 20–24 September
2004; IEEE: Anchorage, AK, USA, 2004; Volume 6, pp. 4306–4309.
30. Balenzano, A.; Mattia, F.; Satalino, G.; Lovergine, F.P.; Palmisano, D.; Davidson, M.W.J. Dataset of Sentinel-1 Surface Soil Moisture
Time Series at 1 Km Resolution over Southern Italy. Data Brief 2021, 38, 107345. [CrossRef]
31. Martínez-Fernández, J.; González-Zamora, A.; Sánchez, N.; Gumuzzio, A.; Herrero-Jiménez, C.M. Satellite Soil Moisture for
Agricultural Drought Monitoring: Assessment of the SMOS Derived Soil Water Deficit Index. Remote Sens. Environ. 2016, 177,
277–286. [CrossRef]
32. Devadoss, J.; Falco, N.; Dafflon, B.; Wu, Y.; Franklin, M.; Hermes, A.; Hinckley, E.-L.S.; Wainwright, H. Remote Sensing-Informed
Zonation for Understanding Snow, Plant and Soil Moisture Dynamics within a Mountain Ecosystem. Remote Sens. 2020, 12, 2733.
[CrossRef]
33. Carlson, T.N.; Gillies, R.R.; Schmugge, T.J. An Interpretation of Methodologies for Indirect Measurement of Soil Water Content.
Agric. For. Meteorol. 1995, 77, 191–205. [CrossRef]
34. Geng, Y.-J.; Leng, P.; Li, Z.-L. A Method for Estimating Surface Soil Moisture from Diurnal Land Surface Temperature Observations
over Vegetated Regions: A Preliminary Result over an AmeriFlux Site and the REMEDHUS Network. J. Hydrol. 2023, 617, 129020.
[CrossRef]
35. Schmugge, T.J.; Jackson, T.J.; McKim, H.L. Survey of Methods for Soil Moisture Determination. Water Resour. Res. 1980, 16,
961–979. [CrossRef]
36. Hossain, B. EM38 for Measuring and Mapping Soil Moisture in a Cracking Clay Soil. Ph.D. Thesis, University of New England,
Armidale, UK, 2008. Available online: https://hdl.handle.net/1959.11/2534 (accessed on 15 May 2023).
37. Lu, Y.; Song, W.; Lu, J.; Wang, X.; Tan, Y. An Examination of Soil Moisture Estimation Using Ground Penetrating Radar in Desert
Steppe. Water 2017, 9, 521. [CrossRef]
38. Jackson, T.J., III. Measuring Surface Soil Moisture Using Passive Microwave Remote Sensing. Hydrol. Process. 1993, 7, 139–152.
[CrossRef]
39. Evett, S.R.; Parkin, G.W. Advances in Soil Water Content Sensing: The Continuing Maturation of Technology and Theory. Vadose
Zone J. 2005, 4, 986–991. [CrossRef]
40. Diacono, M.; Rubino, P.; Montemurro, F. Precision Nitrogen Management of Wheat. A Review. Agron. Sustain. Dev. 2013, 33,
219–241. [CrossRef]
41. Altdorff, D.; Galagedara, L.; Nadeem, M.; Cheema, M.; Unc, A. Effect of Agronomic Treatments on the Accuracy of Soil Moisture
Mapping by Electromagnetic Induction. CATENA 2018, 164, 96–106. [CrossRef]
42. Corwin, D.L.; Plant, R.E. Applications of Apparent Soil Electrical Conductivity in Precision Agriculture. Comput. Electron. Agric.
2005, 46, 1–10. [CrossRef]
43. Edeh, J.A. Quantifying Spatio-Temporal Soil Water Content Using Electromagnetic Induction. Ph.D. Thesis, University of the
Free State, Bloemfontein, South Africa, 2017. Available online: http://hdl.handle.net/11660/6471 (accessed on 15 May 2023).
44. Eigenberg, R.A.; Doran, J.W.; Nienaber, J.A.; Ferguson, R.B.; Woodbury, B.L. Electrical Conductivity Monitoring of Soil Condition
and Available N with Animal Manure and a Cover Crop. Agric. Ecosyst. Environ. 2002, 88, 183–193. [CrossRef]
45. Martinez, G.; Huang, J.; Vanderlinden, K.; Giráldez, J.V.; Triantafilis, J. Potential to Predict Depth-Specific Soil-Water Content
beneath an Olive Tree Using Electromagnetic Conductivity Imaging. Soil Use Manag. 2018, 34, 236–248. [CrossRef]
46. Stanley, J.N.; Lamb, D.W.; Falzon, G.; Schneider, D.A. Apparent Electrical Conductivity (ECa) as a Surrogate for Neutron Probe
Counts to Measure Soil Moisture Content in Heavy Clay Soils (Vertosols). Soil Res. 2014, 52, 373. [CrossRef]
47. Sudduth, K.A.; Kitchen, N.R.; Wiebold, W.J.; Batchelor, W.D.; Bollero, G.A.; Bullock, D.G.; Clay, D.E.; Palm, H.L.; Pierce, F.J.;
Schuler, R.T.; et al. Relating Apparent Electrical Conductivity to Soil Properties across the North-Central USA. Comput. Electron.
Agric. 2005, 46, 263–283. [CrossRef]
48. Kitchen, N.R.; Sudduth, K.A.; Drummond, S.T. Soil Electrical Conductivity as a Crop Productivity Measure for Claypan Soils.
J. Prod. Agric. 1999, 12, 607–617. [CrossRef]
49. Ojo, O.I.; Ilunga, F. Geospatial Analysis for Irrigated Land Assessment, Modeling and Mapping. In Multi-Purposeful Application of
Geospatial Data; Rustamov, R.B., Hasanova, S., Zeynalova, M.H., Eds.; InTech: London, UK, 2018; ISBN 978-1-78923-108-3.
50. Verwey, P.M.J. The Influence of the Irrigation on Groundwater at the Vaalharts Irrigation Scheme. Ph.D. Thesis, University of the
Free State, Bloemfontein, South Africa, 2009.
51. Vermeulen, D.; van Niekerk, A. Evaluation of a WorldView-2 Image for Soil Salinity Monitoring in a Moderately Affected
Irrigated Area. J. Appl. Remote Sens. 2016, 10, 026025. [CrossRef]
52. Pretorius, W.M. Vaalharts: Environmental Aspects of Agricultural Land and Water Use Practices. Ph.D. Thesis, North-West
University, Potchefstroom, South Africa, 2018.
53. Mpandeli, S.; Naidoo, D.; Mabhaudhi, T.; Nhemachena, C.; Nhamo, L.; Liphadzi, S.; Hlahla, S.; Modi, A.T. Climate Change
Adaptation through the Water-Energy-Food Nexus in Southern Africa. Int. J. Environ. Res. Public Health 2018, 15, 2306. [CrossRef]
54. Razas, M.A.; Hassan, A.; Khan, M.U.; Emach, M.Z.; Saki, S.A. A Critical Comparison of Interpolation Techniques for Digital
Terrain Modelling in Mining. J. South. Afr. Inst. Min. Metall. 2023, 123, 53–62. [CrossRef]
Water 2023, 15, 1911 23 of 23

55. Tromp-van Meerveld, H.J.; McDonnell, J.J. Assessment of Multi-Frequency Electromagnetic Induction for Determining Soil
Moisture Patterns at the Hillslope Scale. J. Hydrol. 2009, 368, 56–67. [CrossRef]
56. Nell, J.P.; Van Huyssteen, C.W. Prediction of Primary Salinity, Sodicity and Alkalinity in South African Soils. S. Afr. J. Plant Soil
2018, 35, 173–178. [CrossRef]
57. Bai, W.; Kong, L.; Guo, A. Effects of Physical Properties on Electrical Conductivity of Compacted Lateritic Soil. J. Rock Mech.
Geotech. Eng. 2013, 5, 406–411. [CrossRef]
58. Turkeltaub, T.; Wang, J.; Cheng, Q.; Jia, X.; Zhu, Y.; Shao, M.; Binley, A. Soil Moisture and Electrical Conductivity Relationships
under Typical Loess Plateau Land Covers. Vadose Zone J. 2022, 21, e20174. [CrossRef]
59. Wallace, K.J.; Laughlin, D.C.; Clarkson, B.D.; Schipper, L.A. Forest Canopy Restoration Has Indirect Effects on Litter Decomposi-
tion and No Effect on Denitrification. Ecosphere 2018, 9, e02534. [CrossRef]
60. Darwesh, N.; Allam, M.; Meng, Q.; Helfdhallah, A.A.; Ramzy, S.M.N.; El Kharrim, K.; AAl Maliki, A.; Belghyti, D. Using Piper
Trilinear Diagrams and Principal Component Analysis to Determine Variation in Hydrochemical Faces and Understand the
Evolution of Groundwater in Sidi Slimane Region, Morocco. Egypt. J. Aquat. Biol. Fish. 2019, 23, 17–30. [CrossRef]
61. Guerriero, G.; Hausman, J.-F.; Legay, S. Silicon and the Plant Extracellular Matrix. Front. Plant Sci. 2016, 7, 463. [CrossRef]
62. Reedy, R.C.; Scanlon, B.R. Soil Water Content Monitoring Using Electromagnetic Induction. J. Geotech. Geoenviron. Eng. 2003, 129,
1028–1039. [CrossRef]
63. Wong, M.T.F.; Asseng, S. Determining the Causes of Spatial and Temporal Variability of Wheat Yields at Sub-Field Scale Using a
New Method of Upscaling a Crop Model. Plant Soil 2006, 283, 203–215. [CrossRef]
64. Sherlock, M.D.; McDonnell, J.J. A New Tool for Hillslope Hydrologists: Spatially Distributed Groundwater Level and Soilwater
Content Measured Using Electromagnetic Induction. Hydrol. Process. 2003, 17, 1965–1977. [CrossRef]
65. Zeyliger, A.; Chinilin, A.; Ermolaeva, O. Spatial Interpolation of Gravimetric Soil Moisture Using EM38-Mk Induction and
Ensemble Machine Learning (Case Study from Dry Steppe Zone in Volgograd Region). Sensors 2022, 22, 6153. [CrossRef]
66. Misra, R.K.; Padhi, J. Assessing Field-Scale Soil Water Distribution with Electromagnetic Induction Method. J. Hydrol. 2014, 516,
200–209. [CrossRef]
67. Heil, K.; Schmidhalter, U. The Application of EM38: Determination of Soil Parameters, Selection of Soil Sampling Points and Use
in Agriculture and Archaeology. Sensors 2017, 17, 2540. [CrossRef]
68. Burger, F.; Celkova, A. Salinity and Sodicity Hazard in Water Flow Processes in the Soil. Plant Soil Environ. 2003, 49, 314–320.
[CrossRef]
69. Triantafilis, J.; Laslett, G.M.; McBratney, A.B. Calibrating an Electromagnetic Induction Instrument to Measure Salinity in Soil
under Irrigated Cotton. Soil Sci. Soc. Am. J. 2000, 64, 1009–1017. [CrossRef]
70. Foley, J.; Boulton, R. Supporting the Uptake and Application of EMI Technologies on Cotton Farms. Presented during the
Measuring Soil Water Using em38 Technology Workshop. 20–21 July 2015. Available online: https://blackearth.com.au (accessed
on 15 May 2023).

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.