Alex B. McBratney

On digital soil mapping

We review various recent approaches to making digital soil maps based on geographic information systems (GIS) data layers, note some commonalities and propose a generic framework for the future. We discuss the various methods that have been, or could be, used for fitting quantitative relationships between soil properties or classes and their ‘environment’. These include generalised linear models, classification and regression trees, neural networks, fuzzy systems and geostatistics. We also review the data layers that have been, or could be, used to describe the ‘environment’. Terrain attributes derived from digital elevation models, and spectral reflectance bands from satellite imagery, have been the most commonly used, but there is a large potential for new data layers. The generic framework, which we call the scorpanSSPFe (soil spatial prediction function with spatially autocorrelated errors) method, is particularly relevant for those places where soil resource information is limited. It is based on the seven predictive scorpan factors, a generalisation of Jenny’s five factors, namely: (1) s: soil, other or previously measured attributes of the soil at a point; (2) c: climate, climatic properties of the environment at a point; (3) o: organisms, including land cover and natural vegetation; (4) r: topography, including terrain attributes and classes; (5) p: parent material, including lithology; (6) a: age, the time factor; (7) n: space, spatial or geographic position. Interactions (*) between these factors are also considered. The scorpan-SSPFe method essentially involves the following steps:

Further results on prediction of soil properties from terrain attributes : heterotopic cokriging and regression-kriging

Several methods involving spatial prediction of soil properties from landform attributes are compared using carefully designed validation procedures. The methods, tested against ordinary kriging and universal kriging of the target variables, include multi-linear regression, isotopic cokriging, heterotopic cokriging and regression-kriging models A, B and C. Prediction performance by ordinary kriging and universal kriging was comparatively poor as the methods do not use covariation of the predictor variable with terrain attributes. Heterotopic cokriging outperformed isotopic cokriging because the former utilised more of the local information from the covariables. The combined regression-kriging methods generally performed well. Both the regression-kriging model C and heterotopic cokriging performed well when soil variables were predicted into a relatively finer gridded digital elevation model (DEM) and when all the local information was utilised. Regression-kriging model C generally performed best and is, perhaps, more flexible than heterotopic cokriging. Potential for further research and developments rests in improving the regression part of model C.

An overview of pedometric techniques for use in soil survey

Quantitative techniques for spatial prediction in soil survey are developing apace. They generally derive from geostatistics and modern statistics. The recent developments in geostatistics are reviewed particularly with respect to non-linear methods and the use of all types of ancillary information. Additionally analysis based on non-stationarity of a variable and the use of ancillary information are demonstrated as encompassing modern regression techniques, including generalised linear models (GLM), generalised additive models (GAM), classification and regression trees (RT) and neural networks (NN). Three resolutions of interest are discussed. Case studies are used to illustrate different pedometric techniques, and a variety of ancillary data. The case studies focus on predicting different soil properties and classifying soil in an area into soil classes defined a priori. Different techniques produced different error of interpolation. Hybrid methods such as CLORPT with geostatistics offer powerful spatial prediction methods, especially up to the catchment and regional extent. It is shown that the use of each pedometric technique depends on the purpose of the survey and the accuracy required of the final product.

Digital Soil Map of the World

Increased demand and advanced techniques could lead to more refined mapping and management of soils. Soils are increasingly recognized as major contributors to ecosystem services such as food production and climate regulation (1, 2), and demand for up-to-date and relevant soil information is soaring. But communicating such information among diverse audiences remains challenging because of inconsistent use of technical jargon, and outdated, imprecise methods. Also, spatial resolutions of soil maps for most parts of the world are too low to help with practical land management. While other earth sciences (e.g., climatology, geology) have become more quantitative and have taken advantage of the digital revolution, conventional soil mapping delineates space mostly according to qualitative criteria and renders maps using a series of polygons, which limits resolution. These maps do not adequately express the complexity of soils across a landscape in an easily understandable way.

Spatial prediction of soil properties from landform attributes derived from a digital elevation model

Abstract Digital elevation models (DEMs) provide a good way of deriving landform attributes that may be used for soil prediction. The geostatistical techniques of kriging and cokriging are increasingly being applied to predicting soil properties. Whereas ordinary kriging (and universal kriging) utilise spatial correlation to determine the coefficients of the linear predictor, cokriging involves both inter-variable correlation and spatial covariation among variables. Multi-linear regression modelling also offers an alternative to predicting a soil variable by means of covariation. The performance of predicting four soil variables by these methods and two regression-kriging models are compared. The precision and bias of prediction of the six methods were dependent on the soil variable predicted. The mean error of prediction indicates reasonably small bias of prediction for all the soil variables by almost all of the methods. With the exception of topsoil gravel, for which multi-linear regression performed best, the root mean square error showed the two regression-kriging procedures to be best. Further analysis based on the mean ranks of performance by the methods confirmed this. All the kriging methods involving covariables (landform attributes) have a more smoothing effect on the predicted values, thus minimising the influence of outliers on prediction performance. Both the methods of regression-kriging show promise for predicting sparsely located soil properties from dense observations of landform attributes derived from the DEM. Histograms of subsoil clay residuals show outliers in the data set. These outliers are more evident in multi-linear regression, ordinary kriging and universal kriging than regression-kriging. There was a clear advantage in using the regression-kriging methods on those variables which had a small correlation with the landform attributes: root mean square errors for all the soil variables are much smaller than those resulting from any of the multi-linear regression, ordinary kriging, universal kriging or cokriging methods.

Future Directions of Precision Agriculture

Precision Agriculture is advancing but not as fast as predicted 5 years ago. The development of proper decision-support systems for implementing precision decisions remains a major stumbling block to adoption. Other critical research issues are discussed, namely, insufficient recognition of temporal variation, lack of whole-farm focus, crop quality assessment methods, product tracking and environmental auditing. A generic research programme for precision agriculture is presented. A typology of agriculture countries is introduced and the potential of each type for precision agriculture discussed.

Comparison of different approaches to the development of pedotransfer functions for water-retention curves

Abstract Pedotransfer functions (PTFs) for estimating water-retention from particle-size and bulk density are presented for Australian soil. The water-retention data sets contain 733 samples for prediction and 109 samples for validation. We present both parametric and point estimation PTFs using different approaches: multiple linear regression (MLR), extended nonlinear regression (ENR) and artificial neural network (ANN). ENR was found to be the most adequate for parametric PTFs. Multiple linear regression cannot be used to predict van Genuchten parameters as no linear relationship was found between soil properties and the curve shape parameters. Using the prediction set, ANN performance was similar to the ENR performance for the prediction dataset, but ENR performed better on the validation set. Since ANN is still considered as a black-box approach, the ENR approach which has a more physical basis is preferred. Point estimation PTFs were estimated for water contents at −10, −33 and −1500 kPa. Multiple linear regression performed better for point estimation. An exponential increase trend was found between particles

Application of fuzzy sets in soil science: fuzzy logic, fuzzy measurements and fuzzy decisions

Abstract Fuzzy systems, including fuzzy set theory and fuzzy logic, provide a rich and meaningful improvement, or extension of conventional logic. The mathematics generated by this theory is consistent, and fuzzy set theory may be seen as a generalisation of classic set theory. Applications in soil science, which may be generated from, or adapted to fuzzy set theory and fuzzy logic, are wide-ranging: numerical classification of soil and mapping, land evaluation, modelling and simulation of soil physical processes, fuzzy soil geostatistics, soil quality indices and fuzzy measures of imprecisely defined soil phenomena. Many other soil concepts or systems may be modelled, simulated, and even replicated with the help of fuzzy systems, not the least of which is human reasoning itself.

From Pedotransfer functions to soil inference systems

Abstract Pedotransfer functions (PTFs) have become a ‘white-hot’ topic in the area of soil science and environmental research. Most current PTF research focuses only on the development of new functions for predicting soil physical and chemical properties for different geographical areas or soil types while there are also efforts to collate and use the available PTFs. This paper reviews the brief history of the use of pedotransfer functions and discusses types of PTFs that exist. Different approaches to developing PTFs are considered and we suggest some principles for developing and using PTFs. We propose the concept of the soil inference systems (SINFERS), where pedotransfer functions are the knowledge rules for inference engines. A soil inference system takes measurements we more-or-less know with a given level of (un)certainty, and infers data that we do not know with minimal inaccuracy, by means of properly and logically conjoined pedotransfer functions. The soil inference system has a source, an organiser and a predictor. The sources of knowledge to predict soil properties are collections of pedotransfer functions and soil databases. The organiser arranges and categorises the PTFs with respect to their required inputs and the soil types from which they were generated. The inference engine is a collection of logical rules selecting the PTFs with the minimum variance. Uncertainty of the prediction can be assessed using Monte Carlo simulations. The inference system will return the predictions of soil physical and chemical properties with their uncertainties based on the information provided. Uncertainty in the prediction can be quantified in terms of the model uncertainty and input data uncertainty. In order to avoid extrapolation, a method was developed to quantify the degree of belonging of a soil sample within the training set of a PTF. With the first approach to a soil inference system, we can optimally predict various important physical and chemical properties from the information we have utilising PTFs as the knowledge rules.

Simultaneous estimation of several soil properties by ultra-violet, visible, and near-infrared reflectance spectroscopy

Fast and convenient soil analytical techniques are needed for soil quality assessment and precision soil management. Spectroscopy in the ultraviolet (UV, 250–400 nm), visible (VIS, 400–700 nm), and near-infrared (NIR, 700–2500 nm) ranges allows rapid acquisition of soil information at quantitative, and qualitative or indicator, levels for use in agriculture and environmental monitoring. The main objective of this study was to evaluate the ability of reflectance spectroscopy in the UV, VIS, and NIR ranges to predict several soil properties simultaneously. Soil samples (161 surface and subsurface) were used for simultaneous estimation of pH, electrical conductivity (EC), air-dry gravimetric water content, organic carbon (OC), free iron, clay, sand, and silt contents, cation exchange capacity (CEC), and exchangeable calcium (Ca), magnesium (Mg), potassium (K), and sodium (Na). Principal component regression analyses (PCA) were used to develop calibration equations between the reflectance spectral data and measured values for the above soil properties obtained by traditional laboratory methods. By using randomly selected calibration and validation sets of samples, PCA models were able to successfully predict pH, OC, air-dry gravimetric water content, clay, CEC, exchangeable Ca, and exchangeable Mg of soil samples. The predictions, however, were poor for EC, free iron, sand, silt, exchangeable K, and exchangeable Na. The study shows that reflectance spectroscopy in the UV–VIS–NIR range has the potential for the rapid simultaneous prediction of several soil properties.