Vic Barnett

Environmental issues will continue to be of major concern to scientists, sociologists and politicians of all countries in the immediate future. At the forefront of such interest are issues related to water which are manifest in problems of pollution, nourishment, management, conservation, flood control, climate, energy supply and so on. It is natural therefore that the SPRUCE initiative, concerned as it is with Statistics in Public Resources, Utilities and in Care of the Environment, should have organised its second international conference, held in Rothamsted, UK, in September 1993, on the specific theme of statistical models and methods in relation to environmental problems concerning water. This volume presents an integrated and comprehensive coverage of the material of the conference, contributed by leading workers in the field. The dual emphasis is on soundly constructed statistical models and methods and on their use in specific practical applications, making this book an essential resource for scientists and statisticians involved in water related areas of environmental monitoring, protection and conservation.

From its initial publication this book has been the standard text on the subject. Since then, there has been a continuing high level of activity, and work has developed in all major areas, including univariate and multivariate data, linear models, regression, and time series. Thus a third edition was obviously essential. Every essential area is thoroughly updated to reflect the latest state of knowledge. All the topics are fully revised and extended, and additional topics and new emphases are presented. Important features of this new edition are comprehensive updating of methodology and tables; major expansion of topics including time series, contingency tables, multivariate analysis and regression data (particularly influence and robustness, diagnostics, multiple outliers, non-linear models, L[subscript 1]-norm methods); and richer coverage of practical fields and computer-based facilities, and a thoroughly updated reference list. New introduced topics include standards and regulations, computer software packages, and sample surveys. To facilitate ease of use, the book has been restructured into four parts: Basic Principles, Univariate Data, Multivariate and Structured Data, and Special Topics. This new edition, with the revised and new material, is sure to enhance the book's recognised position as the reference on the subject of outliers.