Engineers and applied geophysicists routinely encounter interpolation and estimation problems when analysing data from field observations. Introduction to Geostatistics presents practical techniques for the estimation of spatial functions from sparse data. The author's unique approach is a synthesis of classic and geostatistical methods with a focus on the most practical linear minimum-variance estimation methods, and includes suggestions on how to test and extend the applicability of such methods. The author includes many useful methods (often not covered in other geostatistics books) such as estimating variogram parameters, evaluating the need for a variable mean, parameter estimation and model testing in complex cases (e.g. anisotropy, variable mean, and multiple variables), and using information from deterministic mathematical models. Well illustrated with exercises and worked examples taken from hydrogeology, Introduction to Geostatistics assumes no background in statistics and is suitable for graduate-level courses in earth sciences, hydrology, and environmental engineering, and also for self-study.

"Whether a novice or an experienced geostatistical analyst, this book is a welcome addition to one's library because of its emphasis on combining sound statistical practice with external scientific knowledge of field characteristics, soil chemistry, or flow dynamics." Richard F. Gunst, Bulletin of American Meteorological Society "The author sets an enviable standard with this exposition of spatial statistical methods in a hydrogeological setting." Bulletin of the American Meteorological Society

"...Kitanidis often produces clear and insightful explanations. I would welcome a second edition...eleborating on those topics in which Kitanidis is a leading researcher, particularly in the incorporation of geostatistical concepts into the groundwater inverse problem." Computers & Geosciences

"This book is a good reference, especially for those with limited knowledge in the application of linear estimation methods in statistical modeling." The Leading Edge