Development of pedotransfer functions in soil hydrology

Geoderma 136 (2006) 470 – 473 www.elsevier.com/locate/geoderma Book reviews Y. Pachepsky, W.J. Rawls, Development of pedotransfer functions in soil hydrology, Developments in Soil Science, vol. 30, Elsevier, Amsterdam, 2005, 512, p. 169. ISBN: 0-444-51705-7 Technologies to estimate soil water transport and storage in field soils from other, more easily measured soil properties emerged almost a century ago with publications by Briggs, McLane and Shantz. Subsequent publications through to the 1980s showed continued development. In 1989, with the introduction of the term pedotransfer function PTF, Johan Bouma greatly accelerated interest and progress in this topic. A PTF was initially considered a mechanism to benefit from the historical worldwide soil survey database to provide quantitative measures of soil parameters for estimating and managing land quality processes. The Editors Yakov Pachepsky and Walter Rawls defined the goal of this book to be a first-of-a-kind treatise of existing pedotransfer functions, i.e., a new book about new, emerging PTFs — their development, application and evaluation. And indeed, they met their goal! Each chapter reads easily with ideas, concepts and applications clearly presented. Readers will appreciate the presentations of each author, and the wisdom of the editors to select them. On less than five pages of the Preface, Pachepsky and Rawls succinctly annotate the major contributions in each of the 25 chapters. The style, grammar, explanations and illustrations of each chapter are extremely well done. The Foreword is optimistic, yet written with adequate prudence to be realistic. The book has six parts: I) three chapters on the methods for developing PTFs, II) seven chapters on the estimates of soil water retention and soil hydraulic conductivity, III) four chapters on the estimates of parameters describing soil erosion, solute adsorption, solute transport and soil shrinkage, IV) three chapters on the inclusion of spatial scale and terrain analysis, V) three chapters on the user-oriented techniques, software and methods of evaluation and VI) five chapters on the case studies at selected global locations. The three chapters of Part I illustrate the statistical development and evaluation of PTFs. As most of the voluminous data in soil surveys were not collected for establishing a PTF, the development of a PTF hinges on a successful sorting and identification of delimited data that allows a quantitative definition of the function descriptive of a desired soil property or process. Examples of statistical regression, artificial neural networks, group methods and regression trees together with procedures for validating PTFs and assessing the accuracy and uncertainty of PTF predictions are comprised in these first three chapters. Part II consists of an overview of research attempts to develop PTFs of soil water retention and soil hydraulic conductivity. Chapter 4 discusses how various descriptions of soil texture and soil particle size distributions have been used in different PTFs for soil hydraulic properties. Straightforward parametric methods for soil water retention and hydraulic conductivity are presented in Chapter 5. Requiring easily available data including soil texture and bulk density, the PTFs were improved with scaling methods and a few additional measurements related to soil water parameters. It is shown in Chapter 6 that the inclusion of soil organic matter affects the predictability of PTFs of soil water retention and saturated hydraulic conductivity. A review of the contributions of soil morphology to the derivation and improvement of PTFs of soil hydraulic properties is given in Chapter 7. Relationships between soil water retention and distributions of aggregate sizes and particle sizes using regression tree modeling are reported in Chapter 8. Chapter 9 is a brief reminder that the linkages between chemical and mineralogical characteristics and soil hydraulic properties have yet to be exploited sufficiently to provide PTFs for yield management models of water and salt movement. Grouping strategies based on several different criteria are discussed in Chapter 10 in order to cope with the ubiquitous spatial variability of soils. The four chapters of Part III focus on PTFs that apply to models of soil erosion, solute and gas retention and transport, soil shrinkage and relating soil water retention Book reviews to the so-called key soil water contents. The three chapters in Part IV address the derivation, modification and utility of PTFs over large spatially variable domains where the ever-present scarcity of field data is augmented with various kinds of ancillary measurements to relate soil properties to landscape position, topography and scale. Part V has three chapters. In the first chapter, the innovative concept of a soil inference system is described, and its use illustrated to predict the soil physical and chemical properties of clay. The second chapter succinctly discusses four convenient graphic user interfaces for implementing PTFs in complex models. The last chapter reviews many different methods for evaluating PTFs, and is understandably, the largest chapter of the entire book. The majority of the chapter focuses on various means to characterize the accuracy of specific PTFs, while the remainder of the chapter focuses on functional evaluation of PTFs, i.e., judging the performance of PTFs in specific applications. The last five chapters of the book, Part VI, describe the development and evaluation of PTFs for soil hydraulic properties from various regions — those of tropical soils, European and national databases, regional and national U.S. databases, the national database of Poland, and a database from a small, highly productive agricultural region of Slovakia. A bibliography at the end of each chapter provides an extensive list of relevant literature. Moreover, at the end of the book, the editors provide a bibliography containing 359 additional references related to PTF development. A comprehensive 16-page index allows the reader to easily find the location of a specific subject matter in the book. Overall, the book describes and documents the research effort of the world's community of soil scientists to develop functions of soil properties, derived primarily from historic and present-day soil survey data, capable of being theoretically applied to better manage global soil and water resources. According to this book, what and how much were accomplished? First note that most of the book focuses on details of the development of PTFs with their evaluation usually being statistically derived against measured values of difficultto-obtain soil hydraulic properties. This outcome is fully expected and desired in view of the initial dearth of technologies relating soil survey data to the analytical data of those highly spatially variable soil functions, and the general lack of experience for their development. Second, many authors spoke of the necessity to augment soil survey databases with ancillary data. Landscape measurements of slope and curvature were suggested to address the issue of scale. Needs were 471 expressed for more databases from laboratory analyses such as soil aggregate size distributions, solute breakthrough curves for unsaturated soils, micro- and macroporosities, etc. as well as remotely sensed data and various other kinds of field data. These suggestions and needs imply that a few detailed local measurements would improve the PTF being developed and also provide a basis for its calibration. Third, only about 20% of the book (primarily part V and portions of other chapters) focuses on methods for evaluating PTFs and their performance in modeling or managing soil quality. Nevertheless, it is encouraging to see concepts of functional evaluation of PTFs and soil inference systems beginning to emerge. Fourth, a few authors spoke of the future — e.g., the need to provide better data rather than better statistical methods, increasing the use of soil morphological attributes to predict soil hydraulic properties, assessing the spatial structure of soil properties and processes, and using a complete soil inference system in a GIS context. Quoting one author, “An emerging challenge is the determination of equivalent hydraulic properties using PTF estimates and information on their variations in space. We have a long way to go yet”. Yes, this reviewer agrees that “we have a long way to go yet”, but because of this book, at least two paths to unprecedented success are potentially elucidated. This documentation of research already completed and underway should 1) hasten the interest of others to join and contribute to research and its application under the recently established IUSS Working Group “hydropedology” in several national soil science societies and in the Pedometrics commission of the IUSS, and 2) provide incentive and momentum to improve the scientific content, reliability and application of present-day national and international soil classification schemes. Rationale for the first path is obvious. Rationale for the second path is given in the following three paragraphs. Paraphrasing from the World Reference Base for Soil Resources, FAO, (1998), “Soil is a continuous natural body which has three spatial and one temporal dimension. The morphological organization of the soil exists at four different scales of observation: 1) elementary organizations, 2) assemblages, 3) horizons and 4) pedological systems — the latter being the spatial distributions and relationships of horizons at the scale of the landscape. Pedologists have, until now, mainly considered only the first three scales with relatively few detailed studies having been made with regard to the three-dimensional, spatial organization of the soil and with respect to the actual dynamics of the three-dimensional organization. Such studies are needed to understand the dynamic soil 472 Book reviews entities or soil units, at the scale of the landscape and ecosystems, and to disclose the relationships between the pedosphere and other components of the earth”. From the above paragraph, it is obvious that the horizontal extent and continuity of soil properties and their uncertainty across the landscape have not been adequately quantified in three dimensions, and hence, no such information is routinely available in soil survey databases. Such a dearth of information also explains why present-day characteristics of a soil series are the modal values expected within a soil-mapping unit with no explicit procedure available to ascertain the probability of their spatial location. Measures of autocovariance and cross covariance between the state variables of soil properties and processes at different scales of space and time would allow the development of a conceptual framework to quantify and categorize the remaining fourth morphological organization of a soil landscape. Such information would signal the need to aggregate or disaggregate (upscale or downscale) data across one or several soilmapping units for any particular application. It would also provide technologies to stop the continual subdivision of the earth's landscape into smaller and smaller differently classified domains. If both paths above are chosen, I can well imagine that the contents of a sequel to this book written after the next 20-year period of research will far exceed the expectations of those who contributed to this volume as well as promote the stature and respect of soil science throughout the global community. And, perhaps one of its authors shall not have to say, “We have a long way to go yet”, but rather, “We're almost there!” I urge students and soil scientists of all ages to read, study and recommend this book and its bibliographical references to others. Donald R. Nielsen Department of Land, Air and Water Resources University of California, Davis, CA, 95616, USA E-mail address: drnielsen@ucdavis.edu. doi:10.1016/j.geoderma.2006.03.047 Mark S. Coyne, James A. Thompson, Math for Soil Scientists, Thomson Delmar Learning, Clifton Park, NY, 2006, ISBN: 0-7668-4268-1 (US$ 26.95, 285 pp.) It's always intrigued me why in American English the abbreviation is ‘math’ and not maths as in the British and Australian dialects — it sounds almost as if there is only one mathematic in the US and several mathematics in other places. (I guess it's a lot like the ‘aluminum’ aluminium divide [but that has something to do with the transient neologisms of Sir Humphrey Davy.]) This title suggests this is a textbook on maths for soil scientists. What is it really? It covers a range of topics in soil physics, chemistry, biochemistry and biology which require some basic calculations. I guess pedology — if it's covered at all (doesn't it require calculations?) — is dealt with by some fairly basic statistics. The topics completely enumerated are basic calculations, significant figures, metric units, converting units, soil texture and surface area, bulk and particle densities, soil water, soil strength and structure, water and air flow, soil temperature, chemical buffering, equilibrium concentrations, redox reactions, kinetics, isotopes, microbial growth, counting microbes, decomposition rates, respiration, mineralisation, immobilization, quantifying microbial processes, microbial ecology and diversity, fertiliser recommendations, lime requirement, application rates and nutrient availability, pH, cation and anion exchange capacity, solubility and solution concentration, descriptive statistics, error analysis, inferential statistics, and finally sampling. Phew! (Or too many?) The book is a bit limited mathematically in that it doesn't get as far as the differential or integral calculus. However even as a fairly experienced soil scientist I learned a few new things and re-learned much that I had forgotten. I was amazed to see that the acre-inch is a unit still in use. (I wonder if the foot-poundal is still out there – this used to flummox me in High School.) The sooner we all move to SI the better! Some of the chapters are perhaps too brief and therefore are of limited use. For example, I contrast the five pages on sampling with the recent 332-page treatment of the subject (de Gruijter et al., 2006). In my experience students really struggle with quantitative concepts and calculations these days — they prefer the visual to the mental. Anything that helps this situation is useful, so this book probably offers a useful response to a recent query like “I am really having problems understanding the equations in soil physics and what they are used for. I was just wondering if you knew of any really good textbooks that explain this equations and their meanings simply?” (The question is clearly an indictment on the quality of my soil physics teaching.) Thankfully, there are many examples and worked problems in the book.