My Account: Log In | Join | Renew
Table of Contents
Select All Chapters
Application of the technique known as computer assisted tomography (CAT) to x- and gamma-ray attenuation measurements has provided an exciting new method for nondestructive imaging within a solid matrix with considerable potential for studying soil behaviour in space and time. The information provided, however, is currently limited by the capabilities of the instrumentation available. Commercially available medical CAT scanners have proven useful for visual studies of soil structure, the advancement and stability of wetting fronts and the structural changes following wetting and drying. The usefulness of these systems and of single source gamma CAT scanning systems in studying soil-plant-water systems, however, is invariably restricted by their inability to distinguish between changes in water content and bulk density in swelling and shrinking soils, and by the associated physical relocation of soil elements that can occur. Thus their quantitative applications have been limited to the measurement of water drawdowns in proximity to plant roots in nonswelling soils or statistical assessments of macroporosity distributions before and after complete wetting and drying cycles. Though fast in operation, the quantitative usefulness of x-ray scanners is limited by the polychromatic nature of the beam and the process known as beam hardening. Furthermore the proprietary nature of these commercial systems also makes software modification or extensions impossible. Simultaneous measurement of the spatial distributions of water content and bulk density in soils that exhibit swelling and dispersion, has been shown to be feasible using CAT applied to dual source (Cs-137 and Yb-169) gamma-ray attenuation. The relatively low photon emission from gamma sources and the propagation of statistical errors, however, necessitates large counting times to provide acceptable accuracy and restricts the use of present gamma systems to the study of steady state or only slowly changing systems. Realization of the full potential of this technique will require substantial improvements in scanning geometry and counting electronics to improve the speed and precision of measurements. With further development, however, dual source gamma-ray systems should ultimately prove most effective for quantitative soil studies. It is now more than a decade that CAT (Hounsfield, 1972) (also termed computed tomography [CT]) in various forms has been applied to a number of different energy beams including: electrons, protons, positrons, alpha particles, lasers, radar, ultra-sound, and nuclear magnetic resonance (NMR), to provide three dimensional imaging of the internal structure of solid objects. Detailed reviews of various aspects of CAT scanning have been presented by Newton and Potts (1981) and by Kak and Slaney (1988) and of its use in studying water movement around plant roots by Aylmore (1993). From the point of view of soil and plant scientists interested in understanding the processes of soil structure development, water movement in soils and its availability for plant growth, application of CAT to measurements of the attenuation of energy beams undoubtedly provides one of the most exciting new techniques ever developed. Its promise of the ability to see inside soil columns and monitor the processes occurring in a continuous, nondestructive manner, clearly has the potential to resolve the major controversies in soil physics and soil-plant-water relations. This special publication provides the opportunity to review and evaluate the progress that has and has not been made in the application of CAT to soil-plant-water studies and to define the steps necessary to obtain the maximum benefit from this technique.
X-ray computed tomography presents an exciting alternative for soil and water measurements. Not only is the measurement nonstructured and noninvasive, it also provides soil data with a spatial resolution of 1 mm or less. In this regard, x-ray tomographic measurements can further help the soil physicist to understand flow and transport processes. Our objective is to present the capabilities and limitations of current medical x-ray tomography scanners, by giving two examples. The first example summarizes the results of an earlier reported study, and demonstrates the spatial and temporal variability of soil moisture in draining soil cores. In the second example, we applied x-ray tomography to characterize soil porosity variability. Both experiments show that, although soil and water properties can be determined with a spatial resolution of 1 mm or smaller, further development of this technique is needed to beneficially use x-ray tomography in flow and transport studies.
Nondestructive measurements of soil properties using computed tomography (CT) have been extensively utilized over the past decade. The properties measured have included: bulk density, porosity, water content, solute concentration, macropore size, and fracture width. Several porous media flow processes have also been monitored and quantified with CT measurements. The majority of the published results have been generated with medical CT scanners utilizing low energy, low resolution x-ray sources. In some cases, researchers have constructed custom CT scanners with gamma-ray sources. We present CT images obtained with industrial CT scanners equipped with high energy and high resolution (microfocus) x-ray sources. To demonstrate high energy CT, a large diameter (572 mm) sand sample was imaged before and after wetting. Image subtraction was used to isolate the x-ray attenuation due to the presence of water. The high resolution capability was demonstrated by generating 500 images of a small diameter (5 mm) sandstone sample. These images have spatial resolution on the order of 0.010 mm. These images clearly show the sizes and shapes of the grains and pores. The representative CT images presented demonstrate that CT measurements can be generated on scales spanning several orders of magnitude.
Synchrotron x-ray computed microtomography (CMT) can be used to make nondestructive tomographic sections with spatial resolutions of a few micrometers. This resolution presents possibilities for study of soil-fluid interactions on a spatial scale hitherto unreachable. Details of a CMT apparatus in operation at the Brookhaven National Synchrotron Light Source X26 beam line are presented and prospects for future development at new synchroton facilities are considered. Several investigations of test systems have been made and results for wet and dry samples of glass beads and sand samples are given to show the power of the system.
Solute travel time in a soil core measured from the usual breakthrough experiment is based on flux-averaged outflow concentrations. For soil cores containing preferential flow paths, the computed travel time and dispersivity are not representative of the majority of the core cross section, nor is this travel time representative of the velocities in the preferential flow paths. X-ray computed tomography was used to measure volume averaged relative concentrations inside the soil core during a breakthrough experiment. These results show velocities 50% less than and dispersivities three to four times as large as those based on flux-averaging for undisturbed silt loam forest cores. Smaller differences resulted for undisturbed loamy sand cores from a cultivated field. Negligible differences resulted for uniformly-packed glass bead cores.
The spatial continuity of soil macroporosity in serial sections obtained by computed tomography (CT)-scanning of soil cores was analyzed by geostatistical analysis. The soil samples involved compacted soils and subsoiled soils. A total of 50 1-mm scans (serial sections) were taken from each soil core by a CT-scanner. The x-rays of the serial sections were analysed by an image analyzer. Conventional analysis of soil structure suggested greater soil macroporosity and greater continuity of soil macroporosity in the subsoiled soils. Image analysis of serial sections from CT-scans taken from the soil cores showed that the soil macroporosity in the subsoiled soils was considerably larger, but less variable than the macroporosity in the compacted soils. The soil macroporosity in the compacted soils was greatest just above the compacted soil layer. The geostatistical analysis of soil macroporosity indicated a greater degree of spatial continuity in the subsoiled soils, especially in the largest size class of macropores. Geostatistical analysis was a useful method for indicating spatial variability and for quantifying the apparent continuity of soil macroporosity.
Small scale density variations in soil cores were measured and quantified using a gamma ray tomography scanner. The dry bulk density distribution was measured on samples of Norge silty loam (fine-silty, mixed, thermic Udic Paleustoll). The samples differed in size and sampling method. With these measurements the natural density variation within the samples and the effects of sample collection were defined on the scale of 5 mm3. Driven, 50.8-mm cores were adequate to describe the density variations of the soil matrix. None of the samples were considered adequate to define the soil macropores. The mean density distribution was found to vary over small vertical distances. All samples had density distributions with negative skew and kurtosis less than three. The standard deviation and kurtosis of the density distribution was found to increase with increasing numbers of macropores.
The nondestructive study of plants and soils in situ has historically been difficult due to the inherently opaque nature of the system. Repeated study of root growth and physiological processes in bulk soil has been impossible without techniques for remote sensing, such as x-ray, computed tomography, rhizotrons, or magnetic resonance imaging (MRI). Magnetic resonance imaging techniques are now providing tools for the direct, nondestructive study of plants and soils. New smaller bore, high field MRI scanners can provide resolutions (10–100 μm) far superior to resolutions provided by clinical scanners (>1 mm), and can truly be considered magnetic resonance microscopy (MRM). Images can be repeatedly acquired over time, providing both spatial and temporal information. The technique has provided information on water extraction by roots, root growth and function, flow of water through porous media, water distribution, and binding patterns in sand-sandstone and in planta, and plant physiological processes. The chemical environment can also be probed by examinations of relaxation times and diffusion coefficients measured by MRI. As the technique is further developed, many new and unique applications will be found.
Quantification of root activity in terms of root growth and indirectly through water uptake is necessary for understanding plant growth dynamics. A greenhouse study of soil columns (Lakeland sand, thermic, coated Typic Quart zipsamment, bulk density 1.4 Mg m−3) planted with soybean [Glycine max (L.) Merr.], bahiagrass (Paspalum notatum Fluegge), and a control (not planted) was conducted in 1989. A Treflan based chemical barrier was placed in half of the soil columns. Columns were watered every other day except when being subjected to water stress tests at selected times. Columns were scanned using x-ray computer tomography (CT) once weekly and daily during the stress evaluation periods. X-ray CT enabled qualitative (images) as well as quantitative outputs (statistical moments derived from pixel absorption histograms). The mean x-ray absorption correlated with water content. Results suggested that root presence can also be indirectly inferred based on water content drawn down during planned stress events. Images and statistical moments (e.g., mean, standard deviation, or skewness) from the pixel absorption histogram yielded information related to plant rooting activity especially when roots are large enough to be clearly visible. Quantitative information was extracted from the images themselves using edge detection and binarization image analysis techniques. Image analysis procedures suggested a continuous development of porosity in upper layers that would be expected with root development. X-ray CT may be a valuable tool in plant root studies particularly when one can couple water extraction observations with statistical moment and image analysis results.
Two nuclear magnetic resonance (NMR) imaging techniques, spin echo and gradient recalled echo (GRE) with radio frequency (RF) spoiling, were used to characterize the spatial distribution of bush bean (Phaseolus vulgaris L.) and sunflower (Helianthus annuus L.) roots in plaster of paris columns, 73 mm in diam. and 70 mm in height. The degree of root decay as well as water movement in root channels were observed by sequential NMR imaging studies during 6 months. The average infiltration rate of porous matrix, and the volumetric discharge rate of permeating root channels were also measured during the decay period. High quality NMR images in live roots in situ were obtained with a three-dimensional resolution of 0.4 by 0.4 by 0.6 mm. Root decay reduced the signal to noise ratio of NMR because of dessication and tissue disintegration. Root channels became preferential flow paths after a 4-wk decay period. Dry dead roots were hydrophilic, and could be rehydrated within 5 to 10 min after wetting. Spatial characteristics of roots, such as size, permeation, tortuosity, and orientation of elongation, did not affect the occurrence of preferential flow. The preferential flow velocity approximated from the measured volumetric discharge and the estimated total cross sectional area of permeating root channels was 10 to 50 mm s−1, about 104 to 105 times higher than the average infiltration rate of the porous matrix.