Areas Of Soil Science Research Articles

A National Soil Profile Database for Brazil Available to International Scientists

A comprehensive digital soil profile database of Brazil was compiled and is being made available through the Internet. Most of the soil data were obtained from the Radambrasil project and other regional surveys. The database contains information from 5086 profiles distributed over the whole Brazilian territory corresponding to data from 10034 horizons, each with 31 variables. The variables were chosen to represent different areas of soil science, embracing soil morphological, chemical, mineralogical, and physical attributes. The distribution uniformity of the data was low with sampling densities varying from one profile per 10000 km2 to one profile per 1370 km2 The access to the database is free and its design allows its use not only by soil scientists but also by those working with agricultural, environmental, and land use issues.

From pedotransfer functions to soil inference systems

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.

MEASUREMENTS OF CUMULATIVE INFILTRATION AND WETTING FRONT LOCATION BY TIME DOMAIN REFLECTOMETRY

Knowledge of wetting front location and cumulative infiltration during irrigation and rainfall events are important to many areas of soil science. We tested a new procedure using time domain reflectometry (TDR) to estimate wetting front location and cumulative infiltration. The method is based on an assumption of a uniform soil water content behind the wetting front. Measurements of wetting front location and cumulative infiltration were made with 0.144-m-long TDR probes vertically inserted into homogeneous loamy sand columns. Unlike previous procedures using TDR, the new procedure could be applied to short TDR probes (≤ 0.3 m) because the procedure did not require determination of a reflection of electromagnetic waves from the wetting front in TDR traces. The method gave satisfactory estimates of both wetting front location and cumulative infiltration.

An investigation of brain tumours and other malignancies in an agricultural research institute.

A historical cohort study was carried out in an agricultural research institute in the Republic of Ireland to investigate a perceived excess of cancer deaths among employees. The objectives of the study were (a) to confirm whether or not total or cause specific death rates among selected employees were higher than expected based on national figures, and to quantify any excess, and (b) to identify centres and employee grades associated with any excess risk. The cohort comprised all administrative, research, and technical staff employed for at least two years between 1960 and 1980, giving a final total for analysis of 1323 employees. Employment history was categorised by place of work and employment grade. Deaths among the cohort were identified and causes of deaths verified. Overall the cohort had a lower than expected mortality compared with national rates (O/E = 0.67). Female deaths were too few for analysis but, among the male employees, there was an excess of all cancers (O/E = 1.24; 95% confidence interval (95% CI) 0.8-1.9), in particular brain cancer (O/E = 4.69; 95% CI 1.2-11.4) and bladder cancer (O/E = 9.31, 95% CI 1.9-27.2). The pattern of brain tumour deaths was consistent with an occupational aetiology but it was not possible to identify specific exposures associated with the excess. All four cases occurred among research workers in the plant and soil science area. The pattern of bladder cancers was difficult to interpret. The authors are at present participating in a multicentre study to further investigate cancer risk in laboratory workers.

Improved Models of Particle‐Size Distribution: An Illustration of Model Comparison Techniques

AbstractWe investigated a relatively unexplored area of soil science: the fitting of parameterized models to particle‐size distribution (a subject more thoroughly explored in sedimentology). Comparative fitting of different models requires the use of statistical indices enabling rational selection of an optimum model, i.e., a model that balances the improvement in fit often achieved by increasing the number of parameters, p, against model simplicity retained by minimizing p. Five models were tested on cumulative mass‐size data for 71 texturally diverse New Zealand soils: a one‐parameter (p = 1) Jaky model borrowed from geotechnics; the standard lognormal model (p = 2); two modified lognormal models (each with p = 3); and the bimodal lognormal model (p = 3). The Jaky and modified lognormal models have not previously been introduced into the soil science literature. Three statistical comparators were used: the coefficient of determination, R2; the F statistic; and the Cp statistic of Mallows. The bimodal model and one modified lognormal model (denoted ORL) best fit the data. The bimodal model gave a marginally better fit, but incorporates a sub‐clay mode (untestable with the present data), so we adopted the ORL model as the physically best benchmark for comparison of other models. The simple Jaky one‐parameter model gave a good fit to data for many of the soils, better than the standard lognormal model for 23 soils. The model comparison methods described have potential utility in other areas of soil science. The Cp statistic is advocated as the best statistic for model selection.

Artificial Intelligence, Cognitive Science, and Measurement Theory Applied in Soil Classification

AbstractRecent advances in artificial intelligence and related fields have relevance to problems encountered in soil classification. Expert systems can be used to strengthen efforts to update and revise soil classification, to maintain large data bases such as Soils‐5, and to involve field personnel in technical evaluation programs. Expert systems may also be applied to other areas of soil science, where judgement and practical experience by experts can be computerized, such as soil management recommendations on irrigation, fertilizer use, and erosion control. Cognitive science research aimed at better understanding of the fundamental aspects of mental representation and categorization may provide stimuli to research on improvements in soil classification. Understanding of the theory of measurement can lead to more efficient, less arbitrary use of quantification techniques in the description and interpretation of soils. The purpose of this paper is to draw to the attention of pedologists and other soil scientists the potential usefulness of recent advances in artificial intelligence, cognitive science, and measurement theory. Following an introduction to these fields, we provide suggestions and specific examples regarding their application to soil classification. Our hope is that this introduction will encourage soil scientists to learn more about these disciplines and consider additional applications to soil science.