Soil is a thin layer of natural materials on the earths surface containing both organic and inorganic substances and is capable of supporting plant life. This thin layer of soil, may be covered by water or may be exposed to the atmosphere, is a key resource in crop production and crop related decision making. The soil is composed of four main constituents: inorganic material, organic matter, water and air in different proportions. The soils are heterogeneous in nature having different parent materials, nutrients availability, soil physical and chemical properties and many other factors. (Beckett and Webster 1971) reported that soil physical properties can vary significantly within relatively short distances, as short as a few meters. For efficient crop production, accurate characterization of soil is a preliminary step, since both lateral and vertical soil heterogeneities may have an impact on crop growth. Spatial variability in soil properties is due to a complex interaction of biological (e.g. earthworms, pests, microbes), edaphic (e.g. salinity, organic matter, nutrients, texture), anthropogenic (e.g. soil compaction due to farm machinery, leaching efficiency), topographic (e.g. slope, elevation) and climatic (temperature relative humidity, rainfall) factors (Corwin and Lesch 2005a). More generally, the edaphic or soil-related factors can be categorized as soil physical properties (e.g. moisture content, soil texture, soil structure and organic matter or organic carbon content), soil chemical properties (e.g. soil pH, cation exchange capacity and soil salinity), soil mechanical properties (e.g. soil strength or compaction and bulk density) and soil macronutrients (e.g. nitrogen content, phosphorus content, potassium content, calcium content, magnesium content and sulphur content). Soil micronutrients or trace elements such as iron (Fe), copper (Cu), chloride, manganese (Mn), etc. may also contribute to soil spatial variability.
Abbreviations and acronyms: GPS, global positioning system; CEC, cation exchange capacity; EM, electromagnetic; EMI, electromagnetic induction; ER, electrical resistivity; ECa, apparent electrical conductivity; EC, electrical conductivity; ECw, electrical conductivity of soil solution extract; ECe, electrical conductivity of saturation extract; SAR, sodium adsorption ratio; GPR, ground penetrating radar; SAR, surface penetrating radar; RADAR, RAdio Detection And Ranging; etc. etc…. will continue….The importance and effects of soil nutrients and physical, chemical and mechanical soil properties on crops and plants growth and outputs are well known. For instance, soil texture is directly or indirectly critical for recommendations concerning soil cultivation, sowing, base fertilization and first nitrogen application. Numerous soil properties are influenced by texture include: drainage, water holding capacity, aeration, erosion, soil tilth, pH buffering capacity, etc. Soil structure is clumping and binding of sand, silt and clay fractions to form aggregates and arrangement of soil pores between them. Soil structure has a major influence on water and air movement, biological activity, root growth and seeding emergence. Soil water is very essential
for plants because it transports the minerals through the soil to roots, provides a medium for metabolic chemical and biological processes, acts as a solvent for dissolved ions and minerals and keeps the plants and environment at favourable temperature from evapotranspiration. Soil organic matter acts as a reservoir of nutrients which releases nitrogen, sulfur and phosphorus and improves the water holing capacity of soil and that held water is mainly used by plants. It also improves soil structure and helps prevent soil erosion. Soil salinity, a measure of total soluble salts, occurs naturally or due to human activities. Excessive soil salinity disrupts the normal osmotic balance in plant roots and in severe cases the plants become dehydrated even when the soil is wet. Cation exchange capacity (CEC), which is directly linked with colloidal substances like clay content and organic matter, is a measure of soil fertility, nutrients retention capacity and ability to protect groundwater from cation contamination. Soil pH helps uptake of different nutrients e.g. nitrogen, phosphorus and potassium from soil solution by plants. Soil strength affects the ability of roots to penetrate the soil and influences soil tillage and workability of a soil. Soil macronutrients which are required by plants in greater amount are used in building tissues and proteins within plants which make up a bulk of plants biomass. They are commonly associated with chemical reactions, which are essential for plants growth and development. While, plants need micronutrients in very small amount.
Mechanized farming has been in use over the past few decades. The conventional farm machines have an ability to treat large fields as uniform elements even if there exists a large variation in soil/crop within a field and despite all knowing of the farmer. Any area as large as a field containing wide spatial variations in soil types, nutrient availability and other important factors; not taking these variations into account can result in a loss of productivity. On the other hand, treating a large area as a uniform element is essentially wasteful and uses an excess of costly resources in the form of fertilizers, pesticides and herbicides.
To characterize soil variability using conventional methods are too expensive and time consuming to capture soil heterogeneity at the field scale with required spatial resolution. Clearly there is a need for accurate and efficient soil characterization to help manage the variations in soil properties, which is an essential step for application of farm inputs in an efficient manner. Therefore, over the past several years, the attitudes have been changing and a new method of farming called precision agriculture is rapidly emerging.
Precision agricultural is a site-specific management method in soil and crop variations within a field. This is a technological driven system and can give the spatial and temporal information (where, how much and when to apply) about the application of a farm input such as fertilizer, irrigation, pesticide, herbicide, etc. in a field (Corwin and Lesch 2005a). Precision agriculture technologies offer an alternative to complement or even replace the conventional laboratory soil analysis methods. The fundamental components of precision agriculture include high resolution global positioning system (GPS) devices, remote sensing (e.g. aerial photography, satellite and airborne multispectral imagery, microwave and hyperspectral imagery, radiometrics, geophysical sensing, etc.), yield monitors and invasive (e.g. electromagnetic resistivity, dielectric sensors, etc.) and noninvasive (e.g. proximal visible near infrared spectroscopy) field sensors (Plant 2001). These technologies are proving encouraging scope in different future farming applications in the field of agronomy, soil science and engineering. Bullock and Bullock (2000) stated that the efficient and accurate methods of measuring within field variations in soil properties are essential for precision agriculture. However, the inability to obtain soil characteristics rapidly and inexpensively remains one of the biggest limitations of precision agriculture (Adamchuk et al. 2004).
For characterization of soil properties, data from a satellite suffer from inadequate spatial and temporal resolution (McBratney et al. 2003) and crop residue cover and other limitations of remote sensing limit the use of aerial and satellite soil imagery. However, proximal or ground-based (invasive or non-invasive) soil sensors have the ability to rapidly collect inexpensive high-resolution data and even in real-time, by taking measurements as frequently as once every second (Viscarra Rossel and McBratney 1998). Sudduth et al. (2005) described that sensor-based soil analysis provides several advantages over conventional laboratory methods such as lower cost, increased efficiency, more timely results and collection of dense datasets while just traversing a field. In addition, the dense datasets, as compared to conventional sampling methods, increase the overall spatial estimation accuracy even if the accuracy of individual measurements is lower.
(Adamchuk et al. 2004) wrote an article focussing mainly on technology or ground-based soil sensors used for on-the-go soil measurement. It is a comprehensive article which provides information to the reader about various previous attempts that have been made towards real-time soil sensing. However, this article does not provide any information about the basic theory and principles of those sensors which could make the reader understand why the outputs of a specific soil sensor is affected by specific soil factors or characteristics. Therefore, we felt a gap in literature and thought to write a comprehensive article having brief information about basic physical principles of different soil sensors and how various soil properties can be characterized using them.
The purpose of this review article is to review important soil nutrients and physical, chemical and mechanical soil properties that can be characterized by different laboratory-based, ground-based and proximal (invasive, non-invasive, offline and online) soil sensing techniques that have been reported in literature. This review article is furnished with an overview of background information about different sensing concepts, basic principles and theory, various factors affecting the outputs of these sensors and why specific soil properties can be related with a sensor's output. Additional relevant information regarding data handling, cost of a sensor and/or operation, history of invention and usage in precision agricultural applications will also be documented. The final part of this review article is to review the sensor fusion techniques reported in previous literature, to describe the advantages and issues of sensors data fusion and future possibilities and scope of sensor data fusion in engineering, agronomy, soil science and precision agriculture applications.
Ground-based soil sensors
(Adamchuk et al. 2004) categorized different on-the-go soil sensors in six main categories based on their design concepts. They include: 1) electrical and electromagnetic sensors; 2) optical and radiometrics sensors; 3) mechanical sensors; 4) acoustic sensors; 5) pneumatic sensors; and 6) electrochemical soil sensors. They added that the output of majority of the soil sensors is affected by more than one agronomic soil characteristics. In this article, we emphasized on characterizing important soil properties (soil moisture content, texture, organic matter content, soil pH, cation exchange capacity, soil salinity, soil bulk density and soil compaction) and nutrients (N, P, K, Ca, Mg, Fe, S, etc.) using various available laboratory-based, ground-based and close range soil sensors (invasive or non-invasive) reported in literature. Therefore, we categorized different laboratory-based, ground-based or proximal (offline or online) soil sensors in five major categories based on soil properties and physical principles of the sensors. These categories include: 1) soil reflectance sensing; 2) geophysical/geo-electrical sensing; 3) dielectric sensing; 4) soil strength sensing; and 5) soil nutrients sensing. These categories are described below in Table 1. The cited studies reported different level of outputs in laboratory, in-situ in the field and online sensing. The reported methods do not work equally well in all types of soil, instead mixed results were reported that might be based on difference in soil type and parent materials, soil texture and different soil and environmental factors e.g. water content, temperature, humidity, organic matter, topography and soil colour.
1. Geophysical sensing
Geophysical sensing may be defined as the science of measuring physical property changes in subsurface by means of instruments located on or near the ground surface or in the air. Most often these sensing techniques refer to ground-based physical sensing to help elucidate the location and character of buried archaeological and geological features and their mapping. This is a non invasive and non destructive technique. Geophysical sensors have been used since the early 20th century in environmental studies, archaeology and more recently, are being increasingly used to extract information form near-surface and subsurface, also in regolith and soil profile (Farifteh et al. 2006). In regolith and soil pedology these sensors are designed for various applications such as mapping salinity intrusions, terrain conductivity, soil and rock layers, mineral exploration and detection of some general geological features such as determination of fault and fracture zone (Farifteh et al. 2006).
The geophysical techniques are commonly categorized as either active or passive one. An active technique is based on the injection of electromagnetic signal into the ground and measurement of the response at the sensor without disturbing the soil, e.g. ground penetrating radar (GPR) and electromagnetic induction (EMI) sensing. While passive technique rely on the physical soil attributes that always exist and emit their signal and the sensor does not send any signal into the ground, e.g. detection of gamma radiation from soil matrix. There are various soil, water-basin and groundwater sensing geophysical techniques available commercially now a days. (Allen 2004) conducted an international survey, visited several manufacturers and researchers in different countries and reviewed all geophysical techniques that can be applied for groundwater and soil investigation. The most commonly used geophysical techniques in soil science, precision agriculture and engineering are EMI, GPR and radiometrics (gamma ray sensing). Among these techniques, ground-based EMI techniques based on active electromagnetic radiation and geo-electrical techniques based on electrical resistivity (ER) are being widely adopted in the soil science or precision agriculture utilities (Adamchuk et al. 2004).
Table 1. Soil physical, chemical and mechanical properties that can be described by different commonly used in-situ and online soil sensors based on their physical principles and soil properties.
Soil sensors
Soil properties or characteristics
Sensor category based on sensing concept and soil properties
Sensor name
Physical
Chemical
Mechanical
Soil primary macronutrients
Other macro-micro-nutrients
Soil moisture content
Soil texture (sand, silt and clay content)
Organic matter content or total organic carbon
Soil variation or depth/ thickness of top soil or claypan
pH
Cation exchange capacity (CEC)
Salinity or Na content
Bulk density and compaction
Total nitrogen content (N) or nitrates
Phosphorus (P) content or fertility indicator
Potassium (K) content
Fe, Ca, Mg, S, etc.
Reflectance sensors
Ultraviolet
Visible
Near infrared
Mid-infrared
Ultrasonic/X-ray
Geophysical /geo-electrical sensors
Electromagnetic induction sensors
X
X
X
X
Electrical resistivity sensors
X
X
X
X
Gamma sensor or radiometrics
X
X
X
X
Ground penetrating radar
X
X
Dielectric sensors
Time domain reflectometry
Frequency domain reflectometry
Soil strength sensors
Strain gauge type
Penetrometer type
Others
Soil nutrients sensors
Ion-selective electrodes
Ion-selective field-effect transistors
1.1 Electromagnetic induction
1.1.1 Background of electromagnetic induction
Electromagnetic induction (EMI) sensors are based on Faraday's law used in Physics. (de Jong et al. 1979) reported that the use of EMI for mapping sub-surface geology by injecting electrical current in the soil was stared in the beginning of 20th century. However, in agriculture, the EMI techniques were first introduced in the late 1970's with the work conducted by (Corwin and Rhoades 1982; de Jong et al. 1979; Rhoades and Corwin 1981; Williams and Baker 1982) for salinity appraisal. Soil salinity refers to the accumulation of major dissolved inorganic solutes in soil in aqueous phase. These inorganic solutes include soluble and readily dissolvable salts including charged ions (e.g., Na+, K+, Mg+2, Ca+2, Cl−, HCO3−, NO3−, SO4−2 and CO3−2), non-ionic solutes, and ions that combine to form ion pairs (Kitchen et al. 1999). The salinity problem predominantly arises in irrigated lands where evapotranspiration of water leaves a huge concentration of salts in the remaining water.
Now a days, it has become a commonly used practice to characterize several soil properties such as soil salinity, moisture content, bulk density and clay content using EMI sensors. The EMI applications are most suitable in the areas where subsurface properties are reasonably homogeneous and the effect of one soil property dominate over the others. The variation in the dominant soil property can be related with EMI response (Rhoades and Corwin 1990). A number of factors are responsible for EMI signal passage in soil that will be discussed in later sections.
1.1.2 Electromagnetic induction principle and theory consideration
The EMI devices are composed of a transmitter coil and a receiver coil separated by some distance installed on both ends of a nonconductive bar. The principle of EMI devices is described in detail by (McNeill 1980b). According to (McNeill 1980b), the transmitter coil at or above the ground surface is energized with an alternating current, creating a primary, time-varying magnetic field (Hp) in the soil. This magnetic field induces small eddy currents in the soil, while soil matrix produces a week secondary magnetic field (Hs). The magnitude of eddy current loop is directly proportional to the electrical conductivity of soil in its vicinity while the value of secondary magnetic field is proportional to the magnitude of current flowing within the current loop. The receiver coil responds to both the primary and weak secondary magnetic fields. The secondary magnetic field (Hs) is, in general, a complicated function of the inter-coil spacing (s), the operating frequency (f) and ground conductivity (ECa or σa). Operating at low induction numbers (e.g. low salinity areas or N<<1), the ratio between the primary magnetic field, Hp, and secondary magnetic field, Hs, is a linear function of bulk or apparent soil electrical conductivity (ECa). Low induction number is defined by (McNeill 1980b) as the ratio between inter-coil spacing, s, and skin depth, δ, (e.g. s/δ). The skin depth, δ, is defined as the depth at which the primary magnetic field has been attenuated to 1/e or 37 %. As soil conductivity is not homogeneous with depth and laterally, the EMI devices measure electrical conductivity of the total volume of soil contributing to the signal. Therefore, ground conductivity is called apparent or bulk soil electrical conductivity. (McNeill 1980b) defined the ECa as a volume average of a heterogeneous earth letting to act as a homogenous matter. He formulated ECa in the form of following equation.
ECa = ----- (1)
Where, Hs is secondary magnetic field at receiver coil (Hm-1), Hp is primary magnetic field at transmitter coil (Hm-1), ω = 2Ï€f, f = frequency (Hz), µ0 = permeability of free space, and s = inter-coil spacing (m). The word H stands for the magnetic field or magnetic induction in primary and secondary magnetic filed units. The magnitude and phase of secondary magnetic field measured by receiver coil differ from the primary magnetic field due to soil properties, spacing between transmitter and receiver and instrument orientation i.e. horizontal or vertical dipole mode (Hendrickx and Kachanoski 2002). It is also evident from Equation 1 that the exploration depth of EMI signal depends on the separation between transmitter-receiver coils, the orientation of the instrument (Horizontal or vertical dipole) and operating frequency (McNeill 1980b). Increasing the operating frequency will decrease the exploration depth of the measurements.
(Cassel et al. 2009) summarized the EMI principle and theory and different factors influencing the EMI signal and also presented a review of different EMI devices used for salinity appraisal. (Ben-Dor et al. 2009) reviewed different remote sensing to assess soil salinity. The sensitivity or relative response of EMI devices () is a depth-weighted nonlinear function which can be explained in both dipole modes.
----- (2)
----- (3)
Where, is relative response in horizontal dipole mode, is relative response in vertical dipole mode and Z is normalized depth (actual soil depth, z, divided by inter-coil spacing, s). The EMI cumulative response R, which is equal to unity is related to the relative response with the following equation (Equation 4) (Cook and Walker 1992; McNeill 1980b).
----- (4)
For both dipole configurations, the cumulative response functions R(Z) are expressed as following.
----- (5)
----- (6)
Thus ECa measured by the EMI instruments can be computed mathematically as follows (McNeill 1980b).
----- (7)
Graphically, both relative and cumulative responses are shown in Firgure 1 and Figure 2 respectively (McNeill 1980b). In Figure 1, the relative contribution to Hs from a homogeneous conductive layer dZ located at normalized depth Z explains the depth weighted nonlinearity. The figure 1 also reveals that the maximum sensitivity of EMI instrument in horizontal coil configuration is on the surface of soil while the maximum sensitivity of the instrument in vertical dipole mode is at 0.4 m depth (normalized depth). In case of EM38, which is an EMI instrument the inter-coil spacing is exactly 1 m. Therefore, in this case the actual exploration depth and normalized depth are equal. If we assume a depth of exploration of 70 % of the total response then an EM38 instrument operated in horizontal dipole mode can achieve 70 % of total response from 0.75 m depth while in vertical dipole mode this response is achieved from 1.5 m depth Figure 2.
Figure 1: Comparison of relative responses for horizontal and vertical dipole modes. Adopted from (Cassel et al. 2009) and (McNeill 1980b).
Figure 2: Cumulative response versus normalized depth for both vertical and horizontal dipole modes. Adopted from (Cassel et al. 2009) and (McNeill 1980a).
1.1.3 Factors effecting electrical conductivity
Soil ECa as measured by EMI devices is affected by the conductors buried in soil as well as the physical land chemical properties of the soil matrix. The soil conductors other than metallic objects are dissolved electrolytes in the soil water, conductive minerals formed by rocks, clays and clay minerals (McNeill 1980a). In the absence of metal objects the ground conductivity is primarily electrolytic since most soil and rock minerals are poor electrical conductors. The conductivity of all these electrolytes is proportional to the total number of ions in solutions, their charge, and velocity. In addition to electrolytes, several soil physical properties, including porosity (shape, sizes and number of ports and inter pore distances), moisture content filled in macro pores and pore water temperature greatly affect the ground conductivity. For more details, see the article by (McNeill 1980a). An overview of those factors is presented here. According to (McNeill 1980a) most soil and rock minerals are electrical insulators of very high resistivity. However, on rare occasions conductive minerals such as magnetite, specular hematite, carbon, graphite, pyrite and pyrrhotite occur in sufficient quantities in rocks to greatly increase their overall conductivity. The minerals in sand and silt fractions of soil are electrically neutral and are generally excellent insulators. Completely dry clay is also an insulator but introduction of moisture changes the situation radically. A part of silt content that is very fine may also behave like clay content (Lund et al. 1999). Soil electrical conductivity, in general, is electrolytic and takes place through the moisture filled pores and passages which are contained within the insulating matrix. Clay particles are negatively charged particles having large surface area. Positive charged particle (cations) such as Ca+, Mg+, K+, Na+, NH3+ and H+ are attracted by the clay particles. These cations are loosely bound with clay minerals and can exchange with other cations or go into solution when clay is mixed with water. Soil temperature can influence the velocity of dissolved ions and ultimately ECa. Soil organic matter has also large surface area and in the form of humus it may significantly influence the ECa in colloidal form. However, the colloidal characteristics of humus on ECa have not yet been studied extensively.
(Friedman 2005) presented a detailed review of soil properties which affect soil ECa. He grouped the factors affecting soil ECa into three categories, e.g. bulk soil, solid particle, and soil solution. He defined the 'bulk soil' as respective volumetric fractions occupied by the three phases (solid, liquid, and gases) and secondary phases (aggregation), e.g. porosity, water content and soil structure. The 'solid particles' consist on soil and rocks minerals which are time-invariable, e.g. particles shape, size, orientation, particle size distribution, cation exchange capacity (CEC) and soil wettability. The factors in third category 'soil solution' the attributes change quickly with management and environmental conditions, e.g. ionic strength, cation composition (sodium adsorption ratio, SAR=Na+/((Ca2+ + Mg2+)/2)1/2) and temperature. He explained these factors separately in saturated and unsaturated soils with strong theoretical support.
1.1.4 Apparent soil electrical conductivity pathways
The parameters affecting soil ECa are summarized by (Rhoades et al. 1989b) in terms of three pathways of current flow. They named as conducting paths acting in parallel contributing to ECa of soil matrix which are enlisted below:
Liquid phase pathway: dissolved solids contained in the water which is held in large pores;
Solid-liquid phase pathway: this is primarily via alternating layers of soil particles and interstitial soil solution (a solid liquid series coupling element); and
Solid phase pathway: via solid particles that are in direct and continuous contact with one another and exchangeable cations associated with clay minerals.
(Corwin and Lesch 2003) modified the three pathways for ECa measurements from (Rhoades et al. 1989b) as shown in Fig. 3.
Figure 3. Three pathways of ECa. Taken from (Corwin and Lesch 2003).
(Rhoades et al. 1989b) claimed that the contribution of the solid element (via solid particles at number 3 above) is negligible because soil structure does not allow for enough direct particle to particle contact. They proposed a simplified model including two pathways (continues liquid and liquid-solid pathways) named as dual pathway model as described below. They also claimed that the model is a valid one for relating bulk soil electrical conductivity (ECa), soil water content (θw), soil solution electrical conductivity (ECw) and soil salinity (ECe). The original model is shown below.
ECa = ----- (8)
Where, ECws is specific electrical conductivity of the soil-water pathway in series-coupling with each other (dS/m), ECwc is specific electrical conductivity of continuous-liquid pathway (dS/m), ECs is specific electrical conductivity of surface conductance (dS/m), θs is volumetric water content of surface conductance (cm3/cm3), θws is volumetric water content in soil-water pathway (cm3/cm3) and θw is total volumetric water content (cm3/cm3). The model was further evaluated and extended by (Corwin and Lesch 2003). This model is based on the ideas that electrical conductivity of soil can be modeled as a multi-pathway parallel electrical conductance equation. (Corwin and Lesch 2003) showed that this model is applicable for a wide range of typical agricultural situation. They also developed a means of extending equation for extremely dry soil condition by dynamically adjusting the assumed water content function. ECa measurement is influenced by several soil physical and chemical properties e.g. soil salinity, water content, clay content, bulk density and temperature. Electrolytic conductivity increases at a rate of approximately 1.9 % per degree centigrade increase in temperature . Therefore, ECa should corrected at reference temperature of 25oC. To convert the ECa at reference temperature following equation of (U.S. Salinity Laboratory Staff 1954).
EC25 = ft . ECt ----- (9)
Where, ECt is ECa measured at specific temperature and ft is a temperature correction factor which can be determined from different types of equations described in literature. For example, (Sheets and Hendrickx 1995; 2003) described following exponential equation to estimate this factor (ft).
ft = 0.4470 + 1.4034 e(t/26.815) ----- (10)
Where, t is the temperature at which ECa is measured. (Sudduth et al. 2001) discussed various accuracy issues in mobile ECa sensing.
1.1.5 EMI sensors
There are several noninvasive EMI devices used to measure soil ECa in archaeology, geology and agriculture for the last 50 years. These instruments operate at one or different frequencies for variable depth of exploration, therefore, these sensor are also called as frequency domain electromagnetic induction instruments. These instruments ranged from a handheld device to larger vehicle mountable devices and have transmitter-receiver coils spaced from 1 to 4 m. The most commonly used EMI sensors are EM31, EM34, EM39 and EM38 (Geonics Ltd., Mississauga, ON Canada) and newer GEM2 (Geophex, Raleigh, NC). Geonics instruments operate at fixed frequencies while Geophex instrument can operate at various frequencies. However, EM38 is the single most sensor that is being used extensively in agriculture as its depth of exploration corresponds to the root zone of the plants (Corwin and Lesch 2005a). It is a lightweight bar and was initially designed to carry by hand from in different locations for obtaining stationary ECa readings. (Lambot et al. 2009) tabulated different EMI sensors and their manufactures as shown in Table 2.
Table 2. Different EMI sensors and manufacturers list. Taken from (Lambot et al. 2009).
Manufacture
Name
Operating Frequency
(kHz)
Mode of Operation
Conductivity range
(mS/m)
Dimensions
(m)
Tx-Rs Spacing
(m)
Exploration Depth
(m)
Weight
(kg)
Geonics
EM31-MK2
9.8
10, 100, 1000
4.0
3.66
6.0
12.4
EM34-3
0.4
1.6
6.4
H-H, V-V
10, 100, 1000
Receiver console:
0.19x0.135x0.26
Transmitter Console:
1.55x0.08x0.26
Coils: 0.63 dia.
40.0
20.0
10.0
Up to 60.0
20.5
EM38
14.6
H-H, V-V
1000
1.03x0.12x0.036
1.0
H-H:0.75
V-V:1.5
2.5
EM38-MK2
14.6
1000
1.07x0.17x0.08
1 & 0.5
V-V:0.75
V-V:1.5
5.4
EM39
39.2
100, 1000, 10,000
1.63x0.036Ø
0.5
200-500
2.2
Dualem
Duelem-1S
9.0
H-H, V-V
3000
1.41x0.09Ø
V-V:1.0
V-H:1.1
5
Duelem-2
H-H, V-V
2.0
Duelem-4
H-H, V-V
4.0
Duelem-21S
H-H, V-V
1 & 2
Duelem-42S
H-H, V-V
2 & 4
Duelem-421S
H-H, V-V
1, 2 & 4
GSSI
Profiler
1-16; max. 3 freq. at a time
H-H, V-V
1.46x0.24x0.124
1.22
4.5
Geophex
GEM-2
0.3-96
H-H, V-V
1.83x0.125
20-30
2.0
GEM-2A
0.33-96
H-H, V-V
6.43
V-V: 5.1
H-H: 5.92
110
GEM-3
0.3-96
V-V
Ø=0.3, 0.4, 0.64, 0.96
1.1.6 Review of literature
ECa of the soil profile is an indirect indicator of important soil properties and has become one of the most reliable and frequently used measurements to characterize field variability for application to precision agriculture due to its ease of measurement and reliability (Corwin and Lesch 2003; Rhoades et al. 1999a). (Corwin et al. 2003) reported that ECa is a technology that has become an invaluable tool for characterizing soil physical and chemical properties, crop yield and mapping of spatial soil variability and yield patterns in a field. Soil ECa is a product of both static (e.g. clay, bulk density) and dynamic (e.g. moisture content, salinity, mineralogy, temperature, etc.) factors. Mostly one or more than one factors are responsible for the contributing to soil ECa which is different in different fields. Therefore, ECa measurement and interpretation are highly site-specific. In texture driven (static factors) systems the ECa patterns remains consistent over time or temporally stable as the variation in soil ECa is mainly because of dynamic properties e.g. moisture content, salinity, etc. (Corwin and Lesch 2003). Initially soil ECa was measured manually by placing the sensor on the ground. Later, the EMI sensors were also mounted on a mobile system (all terrain vehicle or quad bike) for real-time conductivity sensing with GPS coordinates (Cannon et al. 1994; Kitchen et al. 1996; Sudduth et al. 2003; Sudduth et al. 2005).
As discussed earlier, the ECa was first used for soil salinity appraisals. Later, the utility of ECa was extended to other soil properties e.g. water content, soil texture and other soil parameters. For salinity appraisals many authors focused on calibration of instruments determining the depth-weighted response of various EMI devices by lifting the devices at heights with a series of above ground EMI measurements (Amezketa 2007a; Borchers et al. 1997; Cook and Walker 1992; Corwin and Rhoades 1982; 1984; 1990; Rhoades and Corwin 1981; Rhoades et al. 1989a; Slavich 1990; Wollenhaupt et al. 1986; Zhang and Yin 1993). Above ground ECa was also reported to relate ECa with near-surface soil properties other than salinity in later research (Abdu et al. 2007; Hossain et al. 2008; Jung et al. 2005). Different calibration procedures have been proposed for salinity assessment, including multiple regression coefficients (Rhoades and Corwin 1981), simple depth weighted coefficients (Cameron et al. 1981; Wollenhaupt et al. 1986), established-coefficients (Corwin and Rhoades 1982; 1984), modeled coefficients (Slavich 1990), mathematical coefficients (Cook and Walker 1992) and logistic model with smooth curve fitting (Triantafilis et al. 2000). However, for the soil properties other than salinity, normally regression models are used to relate them with soil ECa (Hedley et al. 2004; Jung et al. 2005; Sudduth et al. 2003; Sudduth et al. 2005).
In the review of EMI sensors, although the EMI responses have been related with various other applications e.g. estimation of soil properties, crop yield, soil drainage classes, herbicide partition coefficients, etc. in the previous literature. However, we restricted our review on the application of ECa to the soil properties which are directly linked with EMI principle and will extend this review to those soil properties which affect crop yield greatly and have indirect relationship with ECa. The literature review can be found in Table 2.
Table 2. Literature review of physical, chemical and mechanical soil properties that have been related with ECa measured by EMI instruments directly or indirectly in previous research.
Soil property
Sensor
Literature review
Directly measured soil properties
Soil salinity (ECe, ECw), salt content or Na+ content
EM31
(Cook and Walker 1992), (Corwin and Lesch 2003), (Kinal et al. 2006), (Cameron et al. 1981)
EM34
(Cook and Walker 1992), (Williams and Baker 1982), (Williams and Fidler 1983), (Williams and Hoey 1987)
EM38
(Abdu et al. 2007), (Amezketa 2007b), (Broadfoot et al. 2002), (Bronson et al. 2005), (Brunner et al. 2007), (Cook and Walker 1992), (Corwin and Lesch 2003), (Corwin and Lesch 2005b), (Corwin and Lesch 2005c), (Corwin and Rhoades 1982), (Corwin and Rhoades 1984), (Hendrickx et al. 1992), (Kachanoski et al. 1988), (Korsaeth 2005), (Lesch et al. 1995a), (Lesch et al. 1995b), (McKenzie et al. 1997), (McKenzie et al. 1989), (Rhoades and Corwin 1981), (Rhoades et al. 1989a), (Rhoades et al. 1989b), (Rhoades et al. 1990), (Sudduth et al. 2003), (Sudduth et al. 2005), (Wollenhaupt et al. 1986), (Triantafilis et al. 2000), (Vaughan et al. 1995), (Yao et al. 2007), (Bennett and George 1995), (Cameron et al. 1981), (Cassel et al. 2009), (Corwin et al. 2006), (Doolittle et al. 2001), (Hendrickx et al. 2002), (Hendrickx and Kachanoski 2002), (Horney et al. 2005), (Kaffka et al. 2005), (Lesch et al. 2005), (McBride et al. 1990), (McNeill 1992), (Rhoades et al. 1999a), (Rhoades et al. 1999b), (Rhoades et al. 1990), (Slavich and Petterson 1990), (Triantafilis et al. 2002), (Amezketa 2007a), (Arriola-Morales et al. 2009), (Doolittle et al. 2001), (Herrero et al. 2003), (McLeod et al. 2010), (Zhang and Yin 1993), (Johnston et al. 1997), (Diaz and Herrero 1992), (Lesch and Corwin 2003; Lesch et al. 1998; Lesch et al. 1992), (Cannon et al. 1994), (Nettleton et al. 1994), (Mankin et al. 1997), (Eigenberg et al. 1998; Eigenberg and Nienaber 1997; 1998),
GEM
(Doolittle et al. 2001), (Huang 2005), (Won et al. 1996)
Water content
EM31
(Corwin and Lesch 2003), (Sheets and Hendrickx 1995), (Mankin et al. 1997)
EM38
(Brevik et al. 2006), (Carroll and Oliver 2005), (Corwin and Lesch 2003), (Corwin and Lesch 2005b), (Delin and Söderström 2003), (Hanson and Kaita 1997), (Hossain et al. 2008), (Huth and Poulton 2007), (Kachanoski et al. 1988), (Kachanoski et al. 1990), (Khakural et al. 1998), (Maier et al. 2006), (Reedy and Scanlon 2003), (Sudduth et al. 2003), (Sudduth et al. 2005), (Vaughan et al. 1995), (Hezarjaribi and Sourell 2007)
Texture (sand, silt and clay) or texture related soil properties e.g. depth to clay pan or sand layers
EM31
(Triantafilis et al. 2001a), (Inman et al. 2001),
EM38
(Bronson et al. 2005), (Carroll and Oliver 2005), (Cockx et al. 2007), (Corwin and Lesch 2005b), (Delin and Söderström 2003), (Domsch and Giebel 2004), (Hedley et al. 2004), (Jung et al. 2006; Jung et al. 2005), (Kachanoski et al. 1988), (Khakural et al. 1998), (Kitchen et al. 1996), (Kitchen et al. 1999), (Korsaeth 2005), (Mertens et al. 2008), (Saey et al. 2009), (Sudduth et al. 2001), (Sudduth et al. 2003), (Sudduth et al. 2005), (Triantafilis et al. 2001b), (Triantafilis et al. 2001a), (Williams and Hoey 1987), (Broge et al. 2004), (Cambouris et al. 2006), (Cockx et al. 2008), (Domsch 2002), (Doolittle et al. 1994), (Heiniger et al. 2003), (Lesch et al. 2005), (McBratney et al. 2005), (Triantafilis and Lesch 2005), (Wong et al. 2008), (Boettinger et al. 1997), (Doolittle et al. 1994),
Compaction (bulk density or compaction)
EM38
(Carroll and Oliver 2005), (Corwin and Lesch 2005b), (Jung et al. 2006; Jung et al. 2005), (Rhoades et al. 1999a),
Indirectly measured soil properties
Organic matter or organic carbon content
EM38
(Carroll and Oliver 2005), (Corwin and Lesch 2005b), (Delin and Söderström 2003), (Jung et al. 2006; Jung et al. 2005), (Korsaeth 2005), (Martinez et al. 2009), (Saey et al. 2009), (Sudduth et al. 2003), (Sudduth et al. 2005), (Broge et al. 2004)
Cation exchange capacity (CEC)
EM31
(Triantafilis et al. 2009)
EM38
(Bronson et al. 2005), (Hedley et al. 2004), (Jung et al. 2006; Jung et al. 2005), (Korsaeth 2005), (Sudduth et al. 2003), (Sudduth et al. 2005), (Triantafilis et al. 2009), (Heiniger et al. 2003), (McBride et al. 1990)
Nitrogen or nitrate (NO3-) content
EM38
(Delin and Söderström 2003), (Jung et al. 2006; Jung et al. 2005), (Korsaeth 2005), (Cambouris et al. 2006), (Kaffka et al. 2005), (Drommerhausen et al. 1995),
Other nutrients e.g. Cl-, Ca2+, Mg2+, pH, P, K, etc.
EM38
(Bronson et al. 2005), (Delin and Söderström 2003), (Hedley et al. 2004), (Korsaeth 2005), (Cambouris et al. 2006), (Heiniger et al. 2003), (McBride et al. 1990)
