Page Banner

United States Department of Agriculture

Agricultural Research Service

Assessment of Salinity and Irrigation/Drainage Practices
headline bar
J. D. Rhoades, S. M. Lesch, R. D. LeMert, and W. J. Alves
This "poster paper" introduces the reader to a state-of-the-art, practical technology to measure soil salinity and its spatial relations in detail in irrigated fields developed by J. D. Rhoades and colleagues at the U. S. Salinity Laboratory using
  • Mobilized, EM and four-electrode sensors
  • GPS, geostatistics
  • Efficient ground-truthing techniques
This technology was developed for
  • Inventorying (mapping) and monitoring soil salinity
  • Evaluating the adequacy and appropriateness of irrigation and drainage systems and practices
  • Locating the areal sources of salt-loading within diffuse irrigated landscapes
  • Discussions of the theory and software developed for ground-truthing are presented
  • Descriptions of the equipment are given with illustrations
  • Examples are presented demonstrating the method's utility
  • References are provided to publications where the technology is presented in more detail

The achievement of efficient irrigation and effective salinity control requires periodic information of soil salinity levels and distributions within the crop rootzones and fields of irrigated lands, in order to
  • Inventory conditions of soil salinity
  • Assess the adequacy of leaching and drainage
  • Guide management practices
In addition, practical procedures are needed for delineating the sources of salt-loading and for mapping the distribution and extent of drainage problem areas.
The achievement of such an assessment technology begins with a practical methodology for measuring soil salinity in the field, which is complicated by its spatially variable and dynamic nature caused by the effects and interactions of
  • Varying edaphic factors
    • Soil permeability
    • Water table depth
    • Salinity of perched groundwater
    • Topography
    • Soil parent material
    • Geohyrology
  • Management induced processes
    • Irrigation
    • Drainage
    • Tillage
    • Cropping practices
  • Climate-related factors
    • Rainfall, amount and distribution
    • Temperature
    • Relative humidity
    • Wind
When the need for repeated measurements and extensive sampling requirements are met, the expenditure of time and effort to characterize and map a project's salinity condition with conventional soil sampling and laboratory-analysis procedures becomes prohibitive.
A more rapid, field-measurement technology is needed that:
  • Accounts for the spatial location of the measurement sites involved with the required large intensive and extensive data sets
  • Provides a systematic methodology for evaluating management effects
  • Is capable of proving changes or differences in an area's salinity condition over time
Such a technology has been recently developed. It is an integrated system comprised of:
  • Rapid, mobile instrumental techniques for measuring bulk soil electrical conductivity (ECa) in the field as a function of spatial position on the landscape
  • Procedures and software for inferring salinity from ECa
  • Computer-assisted mapping techniques capable of associating and analyzing spatial databases
  • Appropriate spatial statistics to infer salinity distributions in root zones and changes over space and time
Two kinds of mobilized, instrumental systems have been developed for purposes of field salinity measurement:
  • One based solely on the use of four-electrode units to measure ECa
  • The other using an electromagnetic induction sensor, either alone or together with four electrodes
The remainder of this document briefly describes these two systems and gives some examples of their utility.

Mobile Four-Electrode Sensing System
Figure 1: mobile Four-Electrode Sensing System
Figure 1. Mobile "fixed-array" four-electrode system
with GPS antenna mounted on top of the mast
  • The electrodes are combined into the "heels" of tillage shanks and mounted on a hydraulically controlled tool-bar attached to a tractor via a conventional three-point hitch
  • The electrodes run at a depth of about 10 cm in the soil as the tractor moves across the field
  • A Global Positioning System (GPS) antenna is positioned above the tractor cab and used to determine the spatial position of each sensor reading
  • The ECa and the GPS signals are sensed at adjustable frequencies (as often as every second) and logged into memory for later analysis of salinity condition and spatial relations
  • The four-electrode conductivity meter and the GPS receiver, their respective power supplies and their data loggers are contained in the water-tight, stainless steel box mounted behind the tool-bar shown in Figure 1.
  • The tractor operator is provided with a remote monitor displaying time, ECa reading and logging status
  • The analysis of the spatial data is carried out at the side of the field in a mobile office equipped with a computer work station and soil-salinity testing facilities

Examples of output data obtained with the mobile four-electrode sensing system

Figure 2: output of the four-electrode sensing system
Figure 2. Relation between bulk soil electrical conductivity and distance along a transect
across a furrow-irrigated, tile-drained alfalfa field (Imperial clay soil)
located in the Imperial Valley of California
  • Figure 2 shows ECa readings collected every 1 m apart as the tractor moved across a furrow irrigated, tile-drained alfalfa field in the Imperial Valley of California
  • The "minimum" in the ECa readings occurring at about 380 meters from the irrigation-intake end of the field corresponds to the position of a suite of subsurface drains
  • Otherwise, the ECa values increased toward the "tail end" of the field, presumably due to reduced application and infiltration of irrigation water with distance "down" the furrows
Figure 3: average rootzone soil salinities
Figure 3. Average rootzone soil salinities on a laboratory, soil sample-extraction basis (ECe), as predicted from the measured ECa data along the transect
  • Figure 3 show correspondence between soil salinity predictions (ECe basis) based on soil electrical conductivity measurements obtained by mobile, electromagnetic induction (EMh) and four-electrode systcms along a transect across a furrow-irrigated, tile drained alfalfa field (Imperial clay soil) located in the Imperial Valley of California
  • Also shown are salinities predicted from the EM-sensor system discussed later; the accuracy of these predictions is generally excellent
  • These data suggest that much of the variability in average rootzone salinity across the field is caused by the interactive, effects of the drainage and irrigation systems

Example of the marked effect that a subsurface drainage system can have on average rootzone salinity

Figure 4: relation between bulk soil electrical conductivity and distance
Figure 4. Relation between bulk soil electrical conductivity (ECa) and distance along a transect crossing two sets of tile-drains in a field (silty loam soil) located in the Coachella Valley of California
  • Figure 4 shows an example of a field of silty loam soil in the Coachella Valley which has two sets of buried "tile-lines"; one set being about 2.7 m deep and spaced about 90 m apart and another set being about 1.7 m deep and located at one-third and two-third distances between the deeper lines
  • Soil salinity levels "mimicked" the drainage system, with high values of ECa measured in the soil located between tile-spacings and low values in the soil overlying them
  • Concurrently, salinity tended to increase in the direction of irrigation (to the left in the figure), although the trend is "tempered" somewhat by the effect of the drainage system

Mobile Electromagnetic Sensing System
Figure 5: mobile electromagnetic sensing system
Figure 5. mobile salinity assessment vehicle with combined electromagnetic induction and four-electrode soil conductivity sensing systems
This system involves a Geonics EM-38 instrument mounted in front of the transport vehicle within a vinyl ester pipe, as well as two sets of four-electrode arrays mounted underneath the vehicle.


The "EM-pipe" can be rotated, to enable the EM-38 readings to be made in both horizontal (EMh) or vertical (EMv) configurations. The tube and "rotator" are mounted on a hydraulic apparatus which elevates the EM-38 sensor sequentially to various heights above ground and translates it sequentially in the horizontal direction, so as to allow both EMh and EMv measurements to be made sequentially at various heights above both the furrow and seedbed. These changes in the height and orientation of the EM sensor are undertaken in order to alter the depth and distribution of the EM signal in the soil and, thus, to permit the determination of the salinity-distribution in the rootzone in two dimensions. The four-electrode arrays are mounted on a hydraulically operated scissor-action mechanism which includes a sensor and control mechanism to insert the probes into the soil and to measure ECa at both 1-m and 2-m array spacings in both the furrow and seed bed. In the picture the EM-sensor and four-electrode arrays are in the "up", or "travel", position.
An automated control system was developed to carry out the sequence of 52 operations involved in the full range of possible sequential "EM-38 and four-electrode" measurements. The engineering design of this system is described elsewhere (Carter, et. al., 1993). The control system is operated via an interface control panel with enable-buttons for activating EM and four-electrode sensor measurements and a 6-position selection switch for positioning the sensors over (and at various heights above for the EM sensor) the furrow and seed-bed. When the EM button is enabled, the EM sensor is rotated to the vertical (EMv) configuration and the carriage moves both the EM and four-electrode sensors to the selected position. The EM "start" button then initiates the following automated sequence:
  1. the EMv reading is made and logged
  2. the EM-38 sensor is rotated to the horizontal position
  3. the EMh reading is made and logged
  4. the EM-38 sensor is rotated back to the vertical position
This sequence is repeated for each Y-Z position selected. Depressing the four-electrode "start" button initiates the following automated sequence:
  1. the scissors apparatus inserts the electrodes into the soil
  2. ECa is measured at the 1-m array spacing
  3. a delay is provided for data logging at the 1-m spacing
  4. the meter/logger is switched to the 2-m array
  5. the ECa is read after a delay at the 2-m array spacing
After completion of the last logging, the scissors apparatus lifts the electrodes from the soil and stores them in the travel position. A small printed circuit board provides the necessary time delay functions. The mobile unit then moves to the next measurement site. All measurements at each site can be made in about 30-45 seconds. A Cooperative Research and Development Act contract has been developed by USDA/ARS with AG Industrial Manufacturing Inc. of Lodi, California to commercialize this system.

Example 1

With the EM equipment, salinity distributions within the rootzone can be inferred.
Figure 6: two-dimensional pattern of salinity
Figure 6. The average two-dimensional pattern of salinity in the soil profiles of the transect of figure 3
Figure 6 shows the average two-dimensional distribution of salinity in the soil (Imperial clay) profiles along a transect across a furrow-irrigated, tile-drained alfalfa field located in the Imperial Valley of California.
Salinity in the center of the seed-bed of the fine-textured soil is not as high as might be expected. A likely reason for this is the presence of an extensive network of cracks within the bed which allowed water movement through it, especially in the later stages of the irrigation season. This "inter-flow" likely leached out salts which otherwise would have accumulated by capillarity and upward flow in the bed, if it was completely isolated from the furrows. The patterns of salinity within the soil profiles were very similar at various points along the transect; however, in relation to the average profile shape, salinity increased in the upper part of the profile and decreased in the lower part of the profile with distance towards the down gradient end of the furrow-irrigated field, as shown in the next figure.

Example 2

Figure 7: changes in salinity distribution down a furrow
Figure 7. Changes (percentage basis) in salinity distribution, with reference to the mean profile, within soil profiles along a transect across a furrow-irrigated, tile-drained alfalfa field (Imperial clay soil) located in the Imperial Valley of California.
  • The data in figures 6 and 7 show that the pattern of salinity within the bed and throughout the soil profile varied systematically in response to the imposed irrigation system.
  • Salinity distribution in the rootzone can also be affected by the drainage system; this was observed in the Coachella Valley field previously discussed. Lower salinities occurred in this field in the soil overlying the tile-lines and higher salinities occurred in the soil located in between the tile lines.

Example 3

Additionally in this field, as shown below, the distribution of salinity in the soil profile varied with the mean level of salinity.
Figure 8: relation between salinity distribution and mean level of salinity
Figure 8. Relation between salinity distribution and mean level of salinity in a tile drained field (silty loam soil) located in the Coachella Valley of California
These distributions imply that salinity is high in areas where the net flux of water has been upward in the region of the field located in between the drain lines and is low in the areas where the flux has been downward, that is where leaching has occurred in the soil overlying the tile lines.

Example 4

The salinity distribution in the upper part of the rootzone (0-0.5 m) of the Coachella Valley field is shown below:
Figure 9: two-dimensional distributions of salinity
Figure 9. Two-dimensional distributions of salinity in the upper half-meter of the soil profiles of a field located in the Coachella Valley of California, as influenced by mean (0-0.5 m) salinity level
  • These data indicate that the salinity levels and patterns within the seed bed of this field are also related to the mean profile salinity levels, which in turn are related to the drainage pattern.
  • As shown in Figure 9, the salinity distributions in this silty-loam soil are clearly two-dimensional in contrast to the one-dimensional profiles observed for the clay textured Imperial Valley soil (Figure 6).
  • This difference in salinity distribution is thought to be due to differences in the cracking properties of the two soils.
  • Taken together, all these data Figures 4,8 & 9 indicate that the drainage system in this field is inadequate given the manner of irrigation, or geohydrologic situation, or both, existing there.

Mapping Theory and Software
Several of the examples given to show the utility of the assessment equipment involved results converted to soil salinity units, as determined using soil samples and conventional laboratory-extraction procedures (ECe). In fact, most effective use of the mobile sensor-systems described above requires a rapid, accurate method for converting ECa measurements to ECe values.
We previously showed (Rhoades et al., 1989b, 1990) that ECe can be determined from ECa with sufficient accuracy for practical assessment using knowledge, or reasonably accurate estimates, of the clay and water contents in the soil profile at each ECa measurement site. While this method is suitable when ECa measurements are made by hand, it is impractical for processing the large amounts of data generated with the mobile measurement systems.
For this reason we developed a practical methodology based on multiple linear regression (MLR) to estimate soil salinity from
  • Extensive ECa survey data
  • Limited ground-truth data of ECe
  • Trend surface parameters
With the assessment system described herein, a series of easily obtained EM, or four-electrode, or both, instrument readings are acquired across a field using a relatively dense, systematic survey scheme. A limited number of soil samples are then acquired from a specially selected, subset of measurement-sites (as explained below) and measured for salinity (the Rapid Field Method of Rhoades et al, 1989a is most practical for this purpose).
An MLR equation is subsequently established with the co-located data (Lesch et al,1992).
An important requisite of the MLR approach is that the locations of the soil salinity calibration sites must be spatially representative of the entire survey area. This requisite was satisfied by implementing a newly developed spatial sampling algorithm.
Theory and tests of appropriateness of both the MLR approach and the calibration sampling/siting algorithm are described in detail elsewhere (Lesch et al, 1993a,b). Software for this approach is available from the U. S. Salinity Laboratory (Lesch, et. al., 1995). Additionally, other software has been developed to process the mobile, four-electrode transect data for the purposes of plotting transect "profiles" and producing salinity maps.

An example soil salinity map produced using the above procedures and software

Figure 10: predicted soil salinity survey map
Figure 10. Map of average rootzone (0-1.2 m) soil salinity (ECe basis) in a tile-drained field (silty loam soil) located in the Coachella Valley of California. which shows the spatial pattern (average rootzone basis) of the Coachella Valley field previously discussed
  • The median ECe value of 10-12 dS/m measured within the 0-1 m depth in the field shown in figure 10, is excessive for crop production.
  • Additionally, the type of salinity distribution found within the profile, as previously discussed, implies that the direction of net water within the rootzone varies across the field depending upon the invoked patterns of irrigation and drainage.
  • The net-flow direction is upward over much of the field implying inadequate irrigation and drainage systems and related management.
  • These results are described in more detail in two papers: submitted to the Journal of Irrigation and Drainage, ASCE entitled:
"Improving Irrigation/Drainage/Salinity Management using Spatially Referenced Salinity Measurements" by J D Rhoades, S M Lesch, R D Le Mert, and W J Alves.

"Salt Transport in Cracking Soils: Salt Distributions in Soil Profiles and Fields, and Salt Pickup by Runoff Waters" by J D Rhoades, S M Lesch, S L Burch, J Letey, R D Le Mert, P J Shouse, J D Oster, and T O'Halloran.

This poster describes an integrated package of instrumental systems for:
  • Intensively determining the spatial patterns of soil salinity within irrigated fields
  • Assessing the adequacy of irrigation and drainage systems
  • Establishing the area locations of excessive deep percolation and prime sources of salt-loading to the ground water
The technology package described is unique and represents a major breakthrough in our ability to rapidly and accurately assess soil salinity in irrigated lands.
Results presented here indicate that much of the apparent chaos in the spatial pattern of soil salinity found in irrigated fields is man-induced and can be explained in terms of deterministic processes caused by such management practices as:
  • Irrigation
  • Drainage
  • Cultivation
  • Tillage
The edaphic factors and management practices causing the salinity patterns can often be determined using the integrated salinity assessment approach and procedures described here. The system offers a unique technology for accurately and rapidly mapping and monitoring salinity-distributions in the field as well as for determining:
  • Causes of salinity
  • Diffuse sources of salt-loading from irrigated lands
  • Evaluating the effectiveness (possibly efficiency) of irrigation/drainage management practices
Since salinity is a tracer of water flow, the instrumental systems and associated data analysis may have a much broader application than just salinity assessment. For example, the methodology could potentially be used to:
  • Explain or define the underlying processes affecting the transport of nitrate or some pesticides in irrigated fields
  • Assess irrigation uniformity and degree of leaching

  1. Carter, Lyle M., J. D. Rhoades and J. H. Chesson. 1993. Mechanization of soil salinity assessment for mapping. Paper presented at 1993 Water Meeting of American Society of Agricultural Engineering, Chicago, Illinois, Dec. 12-17, 1993.
  2. Kaddah, M. T. And J. D. Rhoades. 1976. Salt and water balance in Imperial Valley,California. Soil Sci. Soc. Am. J. 40:93-100.
  3. Lesch, Scott M., J. D. Rhoades, L. J. Lund, and D. L. Corwin. 1992. Mapping soil salinty using calibrated electromagnetic measurements. Soil Sci. Soc. Am. J. 56(2):540-548
  4. Lesch, Scott M., D. J. Strauss and J. D. Rhoades. 1995. Spatial prediction of soil salinity using electromagnetic induction techniques. I. Statistical prediction models: A comparison of multiple linear regression and cokridging. Water Resour. Res. 31:373-386.
  5. Lesch, Scott M., D. J. Strauss and J. D. Rhoades. 1993. Spatial prediction of soil salinity using electromagnetic induction techniques. II. An efficient spatial sampling algorithm suitable for MLR model identification and estimation.
    Water Resour. Res. 31:387-398.
  6. Lesch, Scott M., J. D. Rhoades, D. J. Strauss, K. Lin and M. A. A. Co. 1995. The ESAP user manual and tutorial guide. Version 1.0, U. S. Salinity Laboratory Research Report #138, 108 pp.
  7. Rhoades, J. D. 1992. Recent advances in the methodology for measuring and mapping soil salinity. Proc. Int'l Symp. on Strategies for Utilizing Salt Affected Lands, Bangkok, Thailand, Feb. 17-25, 1992. pp. 39-58.
  8. Rhoades, J. D. 1994. Soil salinity assessment: Recent advances and findings.
    Proc. ISSS Sub-Commission Salt-Affected Soils Conference, Acapulco, Mexico, July 10-126, 1994. (In press).
  9. Rhoades, J. D., N. A. Manteghi, P. J. Shouse, and W. J. Alves. 1989. Estimating soil salinity from saturated soil-paste electrical conductivity. Soil Sci. Soc. Am. J. 53(2):428-433. Soil electrical conductivity and soil salinity: New formulations and calibrations.
    Soil Sci. Soc. Am. J. 53 (2):433-439.
  10. Rhoades, J. D., P. J. Shouse, W. J. Alves, N. A. Manteghi and S. M. Lesch. 1990. Determining soil salinity from soil electrical conductivity using different models and estimates. Soil Sci. Soc. Am. J. 54(1):46-54.

Last Modified: 4/7/2006