Printer Friendly

Better placement of soil moisture point measurements guided by 2D resistivity tomography for improved irrigation scheduling.


The suitability of two-dimensional surface electrical resistivity tomography (ERT) (Dahlin 2001), also referred to as resistivity imaging, is evaluated for mapping the spatial variability of soil moisture conditions throughout the growing season of irrigated cotton. ERT has been applied to a wide variety of near-surface environmental and soil science problems. The fundamentals and the history of the development of ERT are reviewed by Griffiths and Tumbull (1985), Griffiths et al. (1990), Griffiths and Barker (1993), Reynolds (1997), and Dahlin (2001). Time-lapse ERT can monitor the movement of the wetting front (Daily et al. 1992; al Hagrey and Michaelsen 1999; Mirus et al. 2009; Nimmo et al. 2009), an indicator of deep drainage. Deep drainage is defined as the portion of water that migrates beyond the root-zone. Rucker (2009) numerically modelled the sensitivity of various resistivity arrays for detecting the wetting front, and concluded for the Wenner array, which is commonly used in ERT surveys, that the voltage measurement became insensitive after the wetting front progressed to two electrode spacings. Rings and Hauck (2009) numerically simulated the influence of a wetting front on ERT measurements, and showed if the upper-most layer is conductive that it is difficult to retrieve accurate resistivity values beneath the conductive layer. This is commonly the situation when Vertosols overly sands and gravels. Both Rucker (2009) and Rings and Hauck (2009) indicated the need for independent measurements of soil and water properties for quantitative hydrophysical interpretations of resistivity soundings and ERT measurements.

There is the desire for improved water use efficiency in the irrigation sector in Australia, as well as a need to quantify the contribution of irrigation drainage for regional water balance studies. There is also a worldwide need for better water monitoring in irrigation farming. ERT surveys, not currently used by the irrigation sector in Australia or internationally, have the potential to contribute to on-farm water use efficiency.

To examine the suitability of ERT for mapping the wetting front and moisture storage throughout the profile operating under normal growing and water conditions, two trials were undertaken in the cotton-growing region of the lower Namoi catchment (Fig. 1). Field trial 1 examines changes in the soil profile resistivity due to one watering using traditional flood irrigation in furrows (Fig. 2). This typically delivers 1 ML/ha per watering. The second trial, which is also watered using flood irrigation, examines the changes in resistivity at the beginning and end of the cotton-watering season, just before the last scheduled watering for the growing season.

Many farmers in the Murray-Darling Basin (MDB), Australia, have recently had cutbacks in water allocations due to changes in water management and drought conditions throughout the region. The MDB is the major agricultural region of Australia and is where most of Australia's irrigated cotton is grown. The need for improved water use efficiency can be understood by looking at the decline of cotton production in Australia during the recent drought. The first commercial crops of the modern cotton era were planted in the lower Namoi catchment, within the MDB, in 1961 (Kahl 2006), and the industry reached peak production in 1998-99 when ~535 400 ha of cotton was planted, producing 3 221 340 bales (Cotton Yearbook 2008). Due to the extent of the drought throughout the MDB in the 2007-08 growing season, only 68 500 ha of cotton was planted, producing 607 500 bales.


Deep drainage is one portion of the water balance that is being examined closely as part of improving water use efficiency in irrigation agriculture (Radford et al. 2009; Silburn et al. 2011). Estimates of water loss due to deep drainage in Australian cotton-growing districts are as high as 3 ML/ha.year (Silburn and Montgomery 2004). Smith et al. (2005) examined 79 flood irrigation in-furrow events on cracking clay soils in cotton-growing areas of southern Queensland. Losses to deep drainage averaged 0.425 ML/ha per irrigation, representing an annual loss of up to 2.5 ML/ha throughout the growing season. Thus, there is a strong incentive to map and quantify deep drainage.


To minimise deep drainage, irrigation must be timed so that the amount of water applied does not exceed the field capacity of the soil, while maintaining sufficient soil moisture so that the plants are not water-stressed (Smith et al. 2005). Frequency domain capacitance systems such as C-probe are increasingly being used for water scheduling of irrigated cotton (Sloane 2004). However, due to cost limitations, only one to three probes are placed in a field. Typically, one probe is placed near the head of the field, one near the middle of the field, and one at the tail end of the field. Given that furrow lengths can range from 200 to 1200m, point measurements provide insufficient information on the spatial moisture content throughout the field. These methods include: porous media systems (e.g. tensiometers and granular matrix sensors), time domain reflectometry (TDR), wetting front systems (fullstop or lysimeters), and neutron logging (Charlesworth 2005).

The applicability of ERT for monitoring soil moisture changes associated with crop water use has been verified by collective studies on the soil water content beneath irrigated corn grown in loamy clay Calcisol in Beauce, France (Michot et al. 2001, 2003; Panissod et al. 2001). Observed responses in the dipole--dipole ERT images (0.2-m electrode spacing) were compared with resistivity point measurements, TDR, and thermal probe measurements recorded in a test pit 1.4 m from the surface ERT line. These studies established that the resistivity of each block in the inverted resistivity model increases linearly with decreasing water content. They also suggest that ERT may allow verification of sprinkling irrigation efficiency.

Acworth et al. (2005) used a Wenner ERT array (2.5-m electrode spacing) to monitor the soil moisture beneath sorghum, lucerne, and fallow fields in the Vertosol soils of the Liverpool Plains, NSW, Australia. This site is 150km south of the lower Namoi catchment (the case study site) and has soils with parent material similar to the soils of the case study site, which is downstream from this location. They demonstrated that electrical resistivity decreased with increasing neutron soil moisture meter counts, with a significant degree of correlation (an exponential equation fit with an R-squared value of 0.77). Significant seasonal variations and crop-induced variability in the ERT images were also noted.

At a site in Avignon, France, where the soils varied from silty clay (0-0.6m) to silty clay loam (0.6-1.0 m) and clay loam (1.0-1.4 m), Srayeddin and Doussan (2009) demonstrated that ERT can map high spatial variability of water uptake under sprinkler-irrigated maize and sorghum. A Wenner ERT array was used, with 0.3 m between electrodes. They found distinct differences in the heterogeneity of water uptake observable in the ERT images relating to the different root depths of the maize and sorghum throughout the growing season, and different watering levels (fully, moderately, and poorly watered). In the upper 0.6m of the soil profile, there was a linear relationship between root water uptake and the electrical resistivity variation over the 5 days between two irrigations. Srayeddin and Doussan (2009) concluded that ERT images must be calibrated against independent measurements of water content.

Jayawickreme et al. (2008, 2010) used ERT to examine the difference in soil moisture content beneath adjacent forest and grasslands. They demonstrated that ERT could measure seasonal variations in soil moisture content. Of relevance to this research, they showed that rainfall infiltration under grasslands drained to greater depths because the soils beneath the grasslands were closer to field capacity.

These examples of ERT for mapping moisture changes in soils beneath crops indicate that ERT has the potential to map the wetting associated with deep drainage beneath irrigated cotton grown on a Vertosol. Prior to the trials, it was not known how sensitive the ERT technique would be for delineating the moisture changes under a single watering under commercial growing conditions. Due to the high cation exchange capacity of the soils in the cotton-growing districts, these soils are already electrically conductive with recorded resistivity values of 1-20 ohm-m (Acworth and Jankowski 1997); in addition, the soil structure is extremely dynamic as it goes through wetting and drying cycles (McGarry and Chan 1984; Timms et al. 2001), resulting in the formation of soil cracks which greatly increase the soil's electrical resistivity (Greve et al. 2010).

At the time of the surveys, commercial cotton crops were growing, so it was not possible to heavily instrument the survey lines to record the actual deep drainage. An indication of the likely changes in the measured soil resistivity can be obtained from models of the electrical properties of sediments (Bussian 1983; Kelly 1994; Revil et al. 1998; Kelly et al. 2009). Resistivity ([rho]) is the reciprocal of conductivity ([sigma]). The ERT field measurements and inversions are usually presented as resistivity sections, while the electrical properties of clay surface electrical conduction and water are often discussed using the units of conductivity. The discussion below links these independent measurements.

Soil resistivity is influenced by the porosity of the sediment, the degree of saturation, salinity of the pore water, clay content, and clay type (via surface conduction on clay particles) (Bussian 1983; Revil et al. 1998). Under irrigation, the dominant variable that will cause any change in the measured resistivity is the degree of saturation. As the pore water content increases after irrigation, the resistivity will decrease, and as the cotton plants use the water in the root-zone, the resistivity will increase. The expected change in the measured bulk resistivity ([[rho].sub.b]), also called the whole sample resistivity, is predicted using the electrical model of Revil et al. (1998). Extending the effective medium theory of Bussian (1983), Revil et al. (1998) defined the bulk electrical conductivity (oh) as a nonlinear function of pore-water conductivity ([[sigma].sub.w]), grain surface conductivity ([[sigma].sub.s]), porosity ([phi]), and the cementation exponent (m), which depends on the aspect ratio of the grains and ranges from 1.5 for rounded quartz sand to 4 for clays (Kelly 1994). Revil et al. (1998) extended the effective medium theory to incorporate different charge carriers (anions and cations) and derived the following approximation:


where [t.sup.f.sub.(+)] is the proportion of electrical current transferred in the flee electrolyte by the cations:

[xi] = [[sigma].sub.s] / [[sigma].sub.w] (2)

and F is the formation factor defined by Archie (1942) as:

1 / F = [[phi].sup.m]. (3)

Assuming all other major variables are constant during wetting and drying phases, then as the degree of saturation, [S.sub.w], decreases, the resistivity of the sample increases by:

[[rho]] = 1/[[sigma]] = 1/[[sigma].sub.b][S.sup.n.sub.w] (4)

where [[rho]] is the resistivity of the partially saturated sediment and n is the saturation exponent, which typically has a value of 2. In the field, the saturated Vertosol had a bulk resistivity of 1 ohm-m. Under partial saturation conditions, and assuming n = 2 (Revil et al. 1998), the bulk resistivity is expected to decrease as shown in Fig. 3. Regions with sandier soils would be expected to have higher values for [[rho].sub.b] and [[rho]], but the sandy sediments known to exist at depths greater than 5 m (Martin 1981; Ward 1999; Kelly et al. 2009) were not sampled.


The field trial sites

Soils at both case study sites are very similar grey clays (Vertosol), and are typical of the irrigation districts in the lower Namoi catchment. The clays extend to ~6m and overlie sands and gravels of the weathered and transported sedimentary Early Cretaceous to Permian formations and reworked volcanic Tertiary alkali basalts and trachytes (Martin 1981; Ward 1999). Within the gravel units, there are also clay-rich layers and lenses. Both sedimentary and volcanic rocks form the Nandewar Range to the east, the Liverpool Range in the south, and the Pilliga plateau to the west.


Soils in the trial area have been studied extensively (McGarry et al. 1989; Ward 1999). Over 200 soil cores have been collected in the field trial region and are described by McGarry et al. (1989). The nearest soil core to the field trial sites 1 and 2 is ed125. For case study 3, the nearest soil core description is ed055.

Ward (1999) classifies ed055 and ed125 as grey clay soils which are slow-draining, self-mulching, and formed from water-sorted aeolian clays. In places, brown clays occur and these tend to be more freely draining. Sand has been observed to fill some of the cracks. The soils have a high cation exchange capacity (CEC). At site ed055, in the upper 2.6 m of the profile the CEC varied from 0.420 to 0.473 [mol.sub.c]/kg, with a clay content range of 64-67%. At site ed125, in the upper 1.3 m of the profile the CEC range was 0.403-0.483 [mol.sub.c]/kg and clay content 61-72%, and in the interval 2.5-2.6m the CEC was 0.283 [mol.sub.c]/kg and clay content 42%. This high CEC means that the clay fraction contribution to the bulk electrical conductivity of the sediments is significant (Bussian 1983; Revil et al. 1998).

Defining the hydraulic properties of the clays has proven difficult for two reasons: the upper portions of the soils have been reworked extensively over the past 50 years; and the soils are highly dynamic, cracking when dry and swelling when wet (McGarry and Chan 1984; Timms et al. 2001). The clays have a high water storage capacity and, after being watered by traditional flood irrigation, require 3 days to drain to -20 kPa soil water potential (Rengasamy and Olsson 1993). For similar Vertosol soils in Queensland, Foley et al. (2006) reported near-saturated permeability on the Black Vertosols of 0.3-2 mm/h (7-48mm/day) for shallow plough pans (annual cropping and lucerne leys, controlled traffic, minimum tillage), and 8-25 mm/h (192-600 mm/day) under grass pasture.

Data acquisition and processing

All resistivity measurements were recorded using an ABEM SAS 4000 Terrameter system connected to a multi-electrode resistivity cable. Selection of the four electrodes used in the measurement of a point within the ERT image was automatically controlled by the meter. The current ([C.sub.1] and [C.sub.2]) and potential ([P.sub.1] and [P.sub.2]) electrodes were arranged in the Wenner configuration ([C.sub.1], [P.sub.1], [P.sub.2], [C.sub.2]), and the minimum spacing (a) between the electrodes was 1 m at sites 1 and 3, and 1.25 m at site 2. A 32-electrode cable was used for measurements at sites 1 and 3, allowing for a maximum spacing of 10a. At site 2, two parallel, but offset, 2.5-m cables were used, giving an effective electrode spacing of 1.25 m, with a maximum spacing of 21a.

All resistivity tomography sections collected during the course of this research were recorded with the electrodes aligned parallel to the furrows. The electrodes were placed midway between the crest and the trough of the furrows. The cotton plants are grown along the crest, and the water migrates along the trough (Fig. 2).

At both sites, only every second furrow was watered per irrigation, swapping for the next irrigation (alternate irrigation). However, in the field it was observed under flood irrigation in furrows that the water would break through into the adjacent furrow within 10 min, and that all furrows eventually had water flowing. The water used for irrigating the crop throughout the season at sites 1 and 2 had an electrical conductivity of ~0.500 mS/cm, at 25[degrees]C. At both sites, the target delivery rate was 100 mm/ha per irrigation run (1 ML/ha). In all ERT images, the water is applied starting on the left and flows to the right.

At site 1, the first electrode of the ERT line was placed 90 m into the field from the head ditch end (the end of the field from which the water is delivered via siphons). This was done to centre the ERT on a previously installed capacitance probe (C-probe) used for scheduling the crop irrigation. The total length of the field is 280m. Thirty-two electrodes were spaced 1 m apart, which gives an inverted model depth of ~5 m. The ERT data were recorded on 11 and 12 February 2004, during the sixth watering of the cotton. C-probe measurements were also collected at site 1. The installation of C-probes, their measurements, and web delivery of results were all provided by a commercial service (Agrilink Pty Ltd). The system consists of an array of capacitance sensors installed at 0.2, 0.3, 0.4, 0.6, and 0.8m below the ground surface. In the cracking clays, the C-probes are difficult to calibrate and a universal calibration was used. The difficulty in calibrating the system is due to the high shrinking/swelling properties of the clay. Thus, the degree of saturation reported in the graph is only relative, within the bounds indicated within the graph.

Site 2 is on the same property as Wee Waa site 1, but on a different field. Flood irrigation in furrows was also used to irrigate the cotton on this field. The field has a total length of 670 m. Two offset cables were used to give a longer survey line and a greater depth of investigation. The spacing between electrodes on each cable was 2.5 m, and the offset spacing between electrodes was 1.25 m. This gives an inverted model depth in the order of 10 m. Electrode 1 of cable 1 was positioned at the base of the head ditch, where the siphon water enters the furrow, and the cotton plants started at electrode 6 on cable 1. Between the two sets of measurements, the electrodes were completely removed from the ground. The only reference point was a survey location marker at the head of the furrow. The resistivity images were recorded at this location on 26 October 2004, one month after planting, and on 7 February 2005, when the cotton plants were 0.5-1 m tall. The plants had been irrigated five times.

Resistivity inversions (the numerical procedure for converting apparent resistivity data, where each data point is a volume-averaged measurement, to a discrete-element resistivity model of the earth) were calculated using RES2DINV (available from The inversion algorithm is described by Loke and Barker (1996) and Loke (2010). At field trial site 1, the resistivity images were automatically inverted using the default inversion settings. This simulates the way that a resistivity tomography system could be used as part of water scheduling for irrigated crops. At site 2, an extended model was used for the inversion because it improved the match between the two sets of independent inversions. The Gauss-Newton method was selected for the recalculation of the Jacobian matrix of partial derivatives for solving the least-squares equation when inverting the apparent resistivity to estimate a 'true' resistivity model. In areas with large resistivity and vertical structure contrast, the Gauss Newton method is significantly more accurate than the faster quasi-Newton method also available in RES2DINV (Dahlin and Loke 1998). Each resistivity section was inverted independently. This was considered adequate because of the stable readings and large contrasts observed. Difference inversion, where all time lapse ERT images are modelled with reference to the first set of measurements, may be used to resolve finer features in ERT images (LaBrecque and Yang 2001; Loke 2010). In all of the ERT images presented below, the root mean square error of the inversion was <1%, indicating that there was little noise in the resistivity data and smooth transitioning between adjacent readings.

Results and discussion

Field trial 1: Time-lapse monitoring of flood irrigation in furrows

Cotton-plant water stress is indicated by the plateau (Sloane 2004) in the rate of decline in the relative saturation curves for C-probe depths 0.2-0.6m (Fig. 4). Based on the C-probe observation in Fig. 4, the field was irrigated. The ERT image at site 1 just before the start of irrigating is shown at the top of Fig. 5. Dry soil zones, indicated by the orange/red semi-circles at shallow depths, highlight the positions of some of the cotton plants. Similar semi-circles around the roots of the plants are observed in the ERT images presented by Michot et al. (2003). Thus, the semi-circles of high resistivity show the extent of the zone where the cotton plants have extracted the water from the soil pores. The consistent development of the semi-circles around the root-zones between irrigations indicates that the ERT images, once calibrated, could be used to monitor water uptake by the plants.

From the ERT image it can be observed that the C-probe was positioned in a portion of the field that was relatively dry in the upper 0.6 m and moist at the base of the C-probe. Had the C-probe been placed at 13.5 m (wetter in the shallower profile and drier at the base) or 12 m (wetter throughout the profile), the irrigation scheduling would have been very different. This highlights the need for an electromagnetic or resistivity survey before the placement of the C-probe, in order to determine a representative location for the moisture measurements that are used to schedule irrigation. In the upper 1.5 m of the section, resistivity values vary from 4 to 8 ohm-m. Assuming uniform soil properties along the section, and referring to the expected relationship between the degree of saturation and the measured bulk resistivity for the Vertosols shown in Fig. 3, the degree of saturation varies from ~36% to 50%. In Fig. 4 for probe depths 20-60cm, at 50% saturation the cotton plants are not water-stressed. For over half of the crop, the next watering could have been delayed by 4-5 days. Over a growing season, extending the period between watering by this interval would potentially reduce the need for one irrigation run, saving ~100 mm of water per hectare.


Irrigation started at 08 : 20 and ran for 6 h. ERT results 5 h after the start of irrigation are shown in the middle section presented in Fig. 5. The wetting of the upper soil is indicated by the diminishing size of the semi-circles around the roots of the plants (more of the shallow portion of the image is green). At a depth of 1-2 m, the zone with a resistivity of <4.5 ohm-m (blue), which is more saturated, now extends across most of the image. The wetting front is indicated by the downward extension of the zone where the soil resistivity is <6 ohm-m (green). The wetting front, which indicates deep drainage, has reached a depth of ~4 m after 5 h and extends beyond the bottom of the ERT image after 24 h (Fig. 5), possibly indicating the movement of water beyond 5 m. To quantify the resistivity images, and to reconcile these findings with the permeability values reported by Foley et al. (2006), ERT images of irrigation infiltration need to be compared with chloride recharge estimates similar to those done by Radford et al. (2009) and Scanlon et al. (2007). In the top 0.5 m of the image, there are still impressions of the cotton-plant root-zone, although the resistivity has decreased significantly since being irrigated. In the C-probe graph (Fig. 4), probes from depths 0.2-0.6 m all show a quick response to the irrigation water wetting the profile. Water movement to beyond 0.8 m is observable in the C-probe graph indicated by the small rise in the relative moisture content for the probe at 0.8 m.

This sequence of ERT images demonstrates that this technique can monitor the movement of water in a Vertosol after irrigation. The zones of water uptake by the cotton plants can also be monitored and the potential exists to calibrate the images for monitoring plant water stress.

Field trial 2: Seasonal variability

Seasonal variation of the water stored in the soil profile and the repeatability of electrical resistivity tomography were examined by comparing measurements taken on 26 October 2004 and 7 February 2005 (Fig. 6). Comparing the October and February measurements, it can be seen that the overall pattern is very similar, consisting of a highly electrically conductive layer (blue/ green) overlying a slightly more resistive layer (red/purple). These two layers correspond well with the nearby excavations, which show that the upper 5 m of the soil profile consists of grey cracking clay (Vertosol). Underneath the Vertosol layer is clayey-silty sand.

The upper 0.5 m of the February ERT image shows numerous dry zones around the roots of the established cotton plants that have dried out the upper soil profile. However, there is very little change in the first 6m in the upper 0.5m between the two sections (left of section). This is the interval with no plants between the head channel and the established crop (Fig. 2). These images clearly show that electrical tomography can be used for monitoring moisture content around the roots of cotton plants. In the upper 1.5m of the February inverse model resistivity section, there is considerable variability in the measured resistivity ranging from ~4 to 15 ohm-m, indicating saturation ranging from ~26% to 50%. This variability has a significant implication for irrigation scheduling depending on the placement of a soil moisture point-measurement device.



In the depth interval 1-5 m, another major difference between the October and February sections can be observed. In this interval there are large zones of decreased resistivity (mid-blue, 4-5 ohm-m, on a background of light blue, 5-6 ohm-m). This indicates a decrease in resistivity by as much as 33%. This would be consistent with increased stored water in the soil profile. This interpretation is supported by the ERT observations of Jayawickreme et al. (2008) and the soil core work by Silburn et al. (2011). Both observed increased stored water beneath grasslands/crops in the depth interval 1-5 m compared with locations with trees and shrubs. There is no indication of changes in resistivity below 6 m, which would suggest that water which reaches the sandy sediments either drains below the measured interval, leaving no discernable changes in moisture content, or is insufficient to change the measured resistivity.

A 4-m interval below the root-zones with increased water content represents a large volume of water and has significant water management implications that require further investigation. The ERT sections would need to be calibrated with core moisture measurements to be able to quantify the change in soil moisture content between October and February.


It has been demonstrated that ERT can be used to monitor water movement through Vertosol soil profiles after cotton plants have been watered by flood irrigation. ERT shows that the wetting front migrates beyond the root-zone of the cotton plants within a few hours of irrigation and that, under flood irrigation in furrows, water continues to move down several metres over the 24-h interval of monitoring. ERT images provide information on the uniformity of the soil/sediment at depth along the furrow, detect major zones of infiltration associated with cracking clays or palaeochannels, and differentiate between clayey and sandy soils in areas of uniform pore water salinity.

It was observed that ERT detects seasonal changes in the water content stored within the Vertosols. The water content of the soils below the root-zone increases throughout the season. At case study site 2, this represented a 4-m-thick interval that had electrical resistivity values decrease by as much as 33%. This is a substantial volume of water that is not accessible to the cotton plants. The management of this water needs further investigation.

To accurately quantify the volume of water seen in the ERT images, the electrical properties of the soils need to be calibrated in the laboratory. Alternately, soil water can be quantified by correlating the inverted ERT images with several direct soil moisture measurements, e.g. TDR or neutron probes. Where there is reasonable uniformity in the soil type at the field scale, the ERT images could be converted to moisture sections.

Initial estimates indicate that the soil moisture can vary by ~25% along the field. This indicates that relying on one to three soil-moisture probes (common in the irrigation sector) to schedule watering could result in inefficient irrigation scheduling. This 25% difference in soil moisture means that watering of the plants in the wetter soil could be delayed for up to another 5 days before the plants start to become water-stressed. Over a growing season, extending the time between irrigation runs by this period would save one watering, equal to 100 mm/ha. For plants in the drier soil, the delay in watering may start to stress the plants. The plant water stress window is still being investigated (Sloane 2004). Further research is required on the statistical distribution of the degree of saturation over a field to maximise yield while minimising water used per hectare.

The ERT images recorded in this study highlight the need for electromagnetic and ERT surveys before selecting capacitance or neutron probe measurement points in order to place the probes in regions that are representative of the majority of the field. The current practice of placing one probe near the head, middle, and tail of the field in just one furrow could lead to poor water scheduling. ERT has a role as part of the geophysical family of moisture measurements in filling the gap between the point measurement methods and the broad-scale electromagnetic surveys.


The authors would like to thank the following people and organisations for helping us to obtain access to fields under irrigation: Mike Logan, David Meppem, and Ben Steven. We would also like to acknowledge the financial assistance of Cotton Research Development Corporation and the Cotton Catchment Communities CRC. We thank the reviewers for improving the focus of this manuscript.


Acworth RI, Jankowski J (1997) The relationship between bulk electrical conductivity and dryland salinity in the Narrabri formation at Breeza, Liverpool Plains, New South Wales, Australia. Hydrogeology Journal 5, 109-123. doi:10.1007/s100400050259

Acworth RI, Young RR, Bernadi AL (2005) Monitoring soil moisture status in a black Vertosol on the Liverpool Plains, NSW, using a combination of neutron scattering and electrical image methods. Australian Journal of Soil Research 43, 105-117. doi:10.1071/SR04064

al Hagrey SA, Michaelsen J (1999) Resistivity and percolation study of preferential flow in vadose zone at Bokhorst, Germany. Geophysics 64, 746-753. doi:10.1190/1.1444584

Archie GE (1942) The electrical resistivity log as an aid in determining some reservoir characteristics. Transactions of the American Institute of Mining, Metallurgical and Petroleum Engineers 146, 54-62.

Bussian AE (1983) Electrical conductance in a porous medium. Geophysics 48, 1258-1268. doi:10.1190/1.1441549

Charlesworth P (2005) 'Soil water monitoring.' Irrigation Insights Number 1. 2nd edn (Land & Water Australia: Canberra) Available at: http://lwa.

Cotton Yearbook (2008) 'The Australian cottongrower. Cotton yearbook 2008.' (Greenmount Press: Toowoomba, Qld)

Dahlin T (2001) The development of DC resistivity imaging techniques. Computers & Geosciences 27, 1019-1029. doi:10.1016/S0098-3004 (00)00160-6

Dahlin T, Loke MH (1998) Resolution of 2D Wenner resistivity imaging as assessed by numerical modelling. Journal of Applied Geophysics 38, 237-249. doi: 10.1016/S0926-9851(97)00030-X

Daily W, Ramirez A, LaBrecque D, Nitao J (1992) Electrical resistivity tomography of vadose water movement. Water Resources Research 28, 1429-1442. doi:10.1029/91WR03087

Foley JL, Tolmie PE, Silburn DM (2006) Improved measurement of conductivity on swelling clay soils using a modified disc permeameter method. Australian Journal of Soil Research 44, 701-710. doi: 10.1071/SR05195

Greve AK, Acworth RI, Kelly BFJ (2010) Detection of subsurface soil cracks by vertical anisotropy profiles of apparent electrical resistivity. Geophysics 75, WA85-WA93. doi:10.1190/1.3474590

Griffiths DH, Barker RD (1993) Two-dimensional resistivity imaging and modelling in areas of complex geology. Journal of Applied Geophysics 29, 211-226. doi:10.1016/0926-9851(93)90005-J

Griffiths DH, Turnbull J (1985) A multi-electrode array for resistivity surveying. First Break 3, 16-20.

Griffiths DH, Turnbull J, Olayinka AI (1990) Two-dimensional resistivity mapping with a computer-controlled array. First Break 8, 121-129.

Jayawickreme DH, van Dam RL, Hyndam DW (2008) Subsurface imaging of vegetation, climate, and root-zone moisture interactions. Geophysical Research Letters 35, L18404. doi: 10.1029/2008GL034690

Jayawickreme DH, van Dam RL, Hyndam DW (2010) Hydrological consequences of land-cover change: quantifying the influence of plants on soil moisture with time-lapse electrical resistivity. Geophysics 75, WA43-WA50. doi: 10.1190/1.3464760

Kahl P (2006) 'Cotton pickin' pioneer.' (P. Kahl: Wee Waa, NSW)

Kelly BFJ (1994) Electrical properties of sediments and the geophysical detection of groundwater contamination. PhD Thesis, The University of New South Wales, Sydney. Australia.

Kelly BFJ, Allen D, Ye K, Dahlin T (2009) Continuous electrical imaging for mapping aquifer recharge along reaches of the Namoi river in Australia. Near Surface Geophysics 7, 259-270.

LaBrecque DJ, Yang X (2001) Difference inversion of ERT data: A fast inversion method for 3-D in situ monitoring. Journal of Environmental & Engineering Geophysics 6, 83-90. doi: 10.4133/JEEG6.2.83

Loke MH (2010) 'RES2DINV vet. 3.59 for Windows XP/Vista/7. Rapid 2-D Resistivity & IP inversion using the least-squares method.' Geoelectrical Imaging 2D & 3D Manual. (Geotomo Software: Penang) Available at: (verified June 2011).

Loke MH, Barker RD (1996) Rapid least-squares inversion of apparent resistivity pseudosections using a quasi-Newton method. Geophysical Prospecting 44, 131-152. doi: 10.1111/j.1365-2478.1996.tb00142.x

Martin HE (1981) An early Cretaceous age for subsurface Pilliga sandstone in the Spring Ridge district, Mooki Valley. Journal and Proceedings, Royal Society of New South Wales 114, 29-31.

McGarry D, Chan KY (1984) Preliminary investigation of clay soils' behaviour under furrow irrigated cotton. Australian Journal of Soil Research 22, 90-108. doi:10.1071/SR9840099

McGarry D, Ward WT, McBratney AB (1989) 'Soil studies in the lower Namoi Valley: Methods and data 1: The Edgeroi data set.' Vols 1 and 2. (CSIRO Division of Soils)

Michot D, Benderitter Y, Dorigny A, Nicoullaud B, King D, Tabbagh A (2003) Spatial and temporal monitoring of soil water content with an irrigated corn crop cover using surface electrical resistivity tomography. Water Resources Research 39, 1138. doi: 10.1029/2002WR001581

Michot D, Dorigny A, Benderitter Y (2001) Mise en evidence par resistivite electrique des flux hydrauliques et de l'assechement par le mais d'un CALCISOL de Beauce irrigue. Comptes Rendus de l'Academie des Sciences Paris A332, 29-36.

Mirus BB, Perkins KS, Nimmo JR, Singha K (2009) Hydrologic characterization on of desert soils with varying degrees of Pedogenesis: 2. Inverse modeling for effective properties. Vadose Zone Journal 8, 496-509. doi: 10.2136/vzj2008.0051

Nimmo JR, Perkins KS, Schmidt KM, Miller DM, Stock JD, Singha K (2009) Hydrologic characterization of desert soils with varying degrees of pedogenesis: 1. Field experiments evaluating plant-relevant soil water behavior. Vadose Zone Journal 8, 480-495. doi: 10.2136/vzj2008.0052

Panissod C, Michot D, Benderitter Y, Tabbagh A (2001) On the effectiveness of 2D electrical inversion results: an agricultural case study. Geophysical Prospecting 49, 570-576. doi: 10.1046/j.1365-2478.2001.00277.x

Radford BJ, Silburn DM, Forster BA (2009) Soil chloride and deep drainage responses to land clearing for cropping at seven sites in central Queensland, northern Australia. Journal of Hydrology 379, 20-29. doi: 10.1016/j.jhydrol.2009.09.040

Rengasamy P, Olsson KA (1993) Irrigation and sodicity. Australian Journal of Soil Research 31, 821-837. doi: 10.1071/SR9930821

Revil A, Cathles LM, Losh S, Nunn JA (1998) Electrical conductivity in shaly sands with geophysical applications. Journal of Geophysical Research 103, 23925-23936. doi: 10.1029/98JB02125

Reynolds JM (1997) 'An introduction to applied and environmental geophysics.' (John Wiley & Sons Ltd: Chichester, UK)

Rings J, Hauck C (2009) Reliability of resisitivity quantification for shallow subsurface water processes. Journal of Applied Geophysics 68, 404-416. doi:10.1016/j.jappgeo.2009.03.008

Rucker D (2009) A coupled electrical resistivity infiltration model for wetting front evaluation. Vadose Zone Journa1 8, 383-388. doi:10.2136/vzj2008.0080

Scanlon BR, Reedy RC, Tachovsky JA (2007) Semiarid unsaturated zone chloride profiles: Archives of past land use change impacts on water resources in the southern High Plains, United States. Water Resources Research 43, W06423. doi: 10.1029/2006WR005769

Silburn DM, Montgomery J (2004) 'Deep drainage under irrigated cotton in Australia A review.' WATERpak--a guide for irrigation management in cotton. Section 2.4, pp. 29-40. (Cotton Research and Development Corporation/Australian Cotton Cooperative Research Centre: Narrabri, NSW)

Silburn DM, Tolmie PE, Biggs AJW, Whish JPM, French V (2011) Deep drainage rates of Grey Vertosols depend on land use in semi-arid subtropical regions of Queensland, Australia. Soil Research 49, 424-438.

Sloane D (2004) Using C-probes: Irrigation decisions from the plants' perspective. The Australian Cottongrower 25, 52-55.

Smith RJ, Raine SR, Minkevich J (2005) Irrigation application efficiency and deep drainage potential under surface irrigated cotton. Agricultural Water Management 71, 117-130. doi:10.1016/j.agwat.2004.07.008

Srayeddin I, Doussan C (2009) Estimation of the spatial variability of root water uptake of maize and sorghum at the field scale by electrical resistivity tomography. Plant and Soil 319, 185-207. doi:10.1007/ s11104-008-9860-5

Timms W, Acworth RI, Berhane D (2001) Shallow groundwater dynamics in smectite dominated clay on the Liverpool Plains of New South Wales. Australian Journal of Soil Research 39, 203-218. doi: 10.1071/SR00002

Ward WT (1999) Soils and landscapes near Narrabri and Edgeroi, NSW, with data analysis using fuzzy k-means. CSIRO Land and Water Technical Report 22/99. Available at: technica199/tr22-99.pdf (accessed 11 March 2011).

Manuscript received 16 March 2011, accepted 23 June 2011

B. F. J. Kelly (A),(B), R. I. Acworth (A), and A. K. Greve (A)

(A) Connected Waters Initiative, The University of New South Wales, affiliated with the National Centre for Groundwater Research and Training, The University of New South Wales, Sydney, NSW 2052, Australia.

(B) Corresponding author. Email:

10.1071/SR11145 1838-675X/11/060504
COPYRIGHT 2011 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kelly, B.F.J.; Acworth, R.I.; Greve, A.K.
Publication:Soil Research
Article Type:Report
Geographic Code:8AUST
Date:Sep 1, 2011
Previous Article:Long-term effects of afforestation with Pinus radiata on soil carbon, nitrogen, and ph: a case study.
Next Article:Effect of pasture buffer length and pasture type on runoff water quality following prescribed burning in the Wivenhoe catchment.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters