Report No: 744/1/01

August 2001



An understanding of hydrological processes is essential for assessing water resources as well as the changes to the resources caused by changes in the land use or climate. Moreover, hydrological simulation models which represent hydrological processes can only be used to predict the consequences of land use and climate change successfully, if they are built on a sound understanding of the processes. The understanding and definition of these hydrological processes, in turn, can only be accomplished by appropriate and careful observation and experimentation of key components of the hydrological cycle, To this end, a long-term project has been initiated to observe and measure specific hydrological processes in order to develop and refine appropriate models for:

As part of this focus on hydrological process research and in order to support other hydrological research, a laboratory facility has been initiated so that the measurement and monitoring of key variables can be accomplished. This report describes the instrumentation that has been developed, presents a comprehensive analysis of selected data and offers an evaluation of the various instruments and techniques used in this first phase of hydrological processes research.


The range of measuring and monitoring instrumentation and techniques for defining hydrological process variables is immense. Hence, this study was not intended to be a comprehensive overview of such instrumentation and techniques, but to present the specific instruments and techniques developed during the course of this establishment phase of the hydrological processes laboratory in the School of Bioresources Engineering and Environmental Hydrology, University of Natal. Similarly the analyses of the soil hydraulic data in this study are not meant to result in a representative set of characteristics for South African soils, but rather, in a selected set of detailed characteristics which can be used for assisting with model parameter estimation and for comparative studies.

Indeed, since many databases of soil hydraulic characteristics are either determined using disturbed samples or are biased towards those materials for which hydraulic properties are more conveniently determined (Leij et al., 1996; Hutson, 1983), this present study is valuable for the assessment of techniques and soil characteristics developed in-situ as well as in the laboratory. In addition, the materials measured are not confined to agricultural soils, but include natural grassland profiles, forested soils, engineered layers as well as laboratory packed samples at a range of bulk densities. The study comprises four parts.


In the first part of this report instrumentation and methodologies are described and evaluated. This project contributed towards the establishment of a laboratory facility and the stocking of selected instrumentation to initiate hydrological processes research. A summary of the instrumentation developed for measuring porous media liquid retention, hydraulic conductivity and solute transport characteristics as well as for monitoring soil water status, is presented in Table ES- 1. The techniques, comprising both laboratory and in-situ, methods, are described briefly where the techniques may differ from conventional methods. After each description an example set of data is presented and this is followed by an evaluation of both the instrumentation and the method, including a list of advantages and shortcomings.

The descriptions and evaluations of the instrumentation offer researchers, consultants and water resources planners an insight into the types of measurement available, the value and range of typical results and the complexity of the various tests. This will assist in determining strategies for field and laboratory measurements commensurate with the problem being addressed. In addition, the database of physical and hydraulic properties will serve as a first port-of-call for professionals requiring initial estimates of the behaviour of various soils and porous media. Table ES- 1. Summary of Laboratory and Field Equipment.
Item Description Qty Purpose
L1 Controlled Outflow Cells 4 Accurate water retention and hydraulic conductivity characteristics
L2 Saturated Hydraulic Conductivity Permeameters 3 Saturated hydraulic conductivity, Ks, by constant head method
L3 Short Column Permeameters 3 Unsaturated hydraulic conductivity characteristic, K(h) using a unit gradient
L4 Bruce-Klute Diffusivity Cells 3 Diffusivity vs water content relationship D(θ). (Also with L1, K(h)).
L5 Diffusion half-cells 3 Solute diffusion vs water content characteristic, Do(θ)
L6 Equlibrated Soluble Mass Leach Columns 3 Measurementof solute behaviour in displacement leach test.
F1 Automatic-feed Double Ring Infiltrometers 4 In-situ saturated hydraulic conductivity, Ks, by ponded infiltration
F2 Tension Infiltrometers 6 In-situ unsaturated hydraulic conductivity, K(h) to h=20cm by direct measurement.
F3 Automatic Tensiometers 300 In-situ soil matric potential using differential pressure transducers.
F4 Automatic Groundwater Level Recorders 25 In-situ groundwater elevation monitoring.
F5 Weir Water Level Recorders 10 Ntshongweni swale market gardening water balance.
F6 Drill-bit Corers for Tensiometers 3 Inserting automatic tensiometers
F7 Field Loggers (4 single ended sensors/logger) 100 Monitoring of automatic tensiometers, groundwater and weir level sensors.
Sundry Equipment (1996/97Acquisition)
S1 Laboratory Scale (6 kg at 1/100 gm) 1 Soil water content, dry mass, bulk and specific density measurements.
S3 MCS Nylon Soil Moisture Sensors (SMS) 4 Monitoring in-situ soil matric potential.
S4 Campbell Scientific Heat Dissipation SMS 10 Monitoring in-situ soil matric potential.
S5 Chloride Ion Probe 1 Measuring diffusion characteristic concentrations.


The results of some 850 measurements of various porous media characteristics are summarised in Chapter 5, while Appendix A and B are dedicated to listing the results of key parameters measured. The porous media measured in this study exhibit a range of textures and bulk densities, a range not normally encountered in databases of soil properties from agricultural profiles (Leij et al., 1996). The texture of 207 of the porous media analysed in this study are compared to those in an international database, UNSODA, in Table ES-2. This present study does not share the bias towards measurements on the more conveniently managed materials that is evident in LTNSODA and other similar studies. In the UNSODA study some 55% of the materials fall in the sandy range, while only 39% of the materials tested in this study classify as sand or sandy materials, albeit the quantity of samples tested in this study is smaller. Nevertheless, a large percentage of the measurements performed in this study were on undisturbed samples or in-situ profiles, unlike many other studies where laboratory packed samples are tested An additional 92 soils, for which the textural class was not categorised, were also included in the measurement and evaluation of hydraulic characteristics.

Table ES-2. Summary of the textural classes of the porous media tested.
Texture Class UNSODA This Study
Quantity % Quantity %
Sand 184 23.6 8 3.9
Loamy sand 64 8.2 3 1.4
Sandy loam 133 17.1 46 22.2
Sandy clay loam 52 6.6 24 11.6
Silt 3 0.4 0 0.0
Silt loam 142 18.2 2 1.0
Clay loam 36 4.6 43 20.8
Loam 70 9.0 52 25.1
Silty clay loam 33 4.2 3 1.4
Sandy clay 3 .4 4 1.9
Silty clay 21 2.7 1 .5
Clay 39 5.0 21 10.1
TOTAL 780 100.0 207 100.0

An assessment of various pedotransfer functions, used to estimate soil hydraulic characteristics from the soil texture and bulk density, was performed. This assessment revealed that the best agreement between the predicted parameters and measured data occur for disturbed samples packed in a laboratory as summarised in Table ES-3. This is a significant result and emphasises the many cautions, tabled in the literature, warning against the misuse of these functions. Nevertheless, the data set presented in the text and appendices can still be used for valuable direct estimations of parameters used in deterministic hydrological models.

Table ES-3. Summary of correlation between predicted water contents and those estimated from measurements with the variance (R) and standard error of predicted water content (s).
Sample Type No. of samples DUL WP
R2 s R2 s
Grassland 31 0.226 0.062 0.221 0.055
Forested 7 0.450 0.089 0.500 0.071
Disturbed 16 0.897 0.036 0.809 0.033


Parameter sets of specific mathematic functions are derived by best fit analysis against the measured hydraulic characteristic data. The mathematical functions chosen for the curve fitting exercise included the well known van Genuchten, Brooks-Corey and Campbell functions. These comprise the typical functions required in many models currently used in hydrological, soil-plant-atmosphere and solute transport simulation. (ACRU, SWAP, HYDRUS2-D, SWAT, SWB, LEACHM, HSPF and others). The parameters derived for these functions are listed in appendices C to E for convenient reference in assessing model input and evaluated against other surveys as demonstrated in Figure ES-1.

Again, selected parameters, describing hydraulic characteristics that are relatively easy to measure (e.g. water retention characteristic), are used to estimate those parameters that require sophisticated or time consuming test techniques (e.g. the hydraulic conductivity characteristic). These estimations are not accurate and emphasise the difference between in-situ or undisturbed measurements, which may contain significant macro-porosity, and those derived from disturbed, laboratory packed samples, which exhibit little macro-porosity.

A typical range of saturated hydraulic conductivities, derived from retention characteristic parameters, is shown in Figure ES-2. The predicted values often vary more than an order of magnitude from the measured data. This reflects the difficulties in predicting the hydraulic conductivities of in-situ soils, where macro-pore structure may dominate the saturated hydraulic conductivity.


The description and evaluation of methods and instrumentation for measuring hydraulic characteristics of porous media and for monitoring the status of soil water will assist researchers, consultants and water resources practitioners in determining strategies for field and laboratory measurements commensurate with the problem being addressed. In addition, the database of physical and hydraulic properties will also serve as a first port-of-call for professionals requiring initial estimates of the behaviour of various soils and porous media.

The database of soil physical and hydraulic characteristics comprise a valuable set of parameters for evaluating the behaviour of subsurface hydrological processes. Moreover, the hydraulic characteristic function parameters, derived for a wide range of porous media, provide not only valuable estimates for input into numerous simulation models, but they also provide insight into characterising materials which exhibit significant macro-pore structure.

Continued instrumentation development and measurement strategies aimed at evaluating the various techniques are recommended. In addition, further population of the soil hydraulic characteristics database with measurements from a wide range of porous media is recommended so that robust methods of estimating these characteristics from easily measured properties can be developed. It is particularly important to develop measurement and predictive techniques for in-situ hydraulic characteristics in the many instances where the dual porosity structural features of the porous medium dominate the subsurface hydrology.

The continued development of techniques for evaluating hydrological processes will allow for the typical accuracies required in the prediction and planning of a scarce water resource which is often subject to multiple demands.