Computational methods for water resource assessments – an exercise kit

Motivation and intended use

Computational methods for water resource assessments – an exercise kit is an exercise book that attempts to (i) develop skills, and (ii) impart an understanding of, the use of computational methods in water resource assessments. The motivation for this book lies in the firm belief of the value in learning by doing. The exercises are developed in a way that encourages readers to explore what happens ‘behind the scenes’ in environmental simulation models. Hence in many exercises the computational procedures are built almost from scratch before applying the model to a particular problem. It is believed that gaining understanding of how computational methods used in environmental assessments actually work not only promotes readers’ potential to develop their own applications, but also increases their skill to assess critically existing environmental simulation models. Many of the exercises are also built on real data measured in the field. This approach hopefully increases the readers’ interest in the subject as well as their appreciation of uncertainties, and even sheer errors, appearing in measurements.

The intent of the present book is not to provide a detailed coverage of the various methods and topics. Instead, it is presumed that, either before embarking on the exercises or in parallel, readers learn the relevant theory by attending lectures or by studying more comprehensive text books. But as readers attain a sound knowledge of the basic theory, the present book is intended to be self-sufficient in giving a brief, concentrated dose of theory that is necessary for tackling the exercises.

Contents

The book contains instructions for the software required in solving the exercises. Most of the exercises are based on Microsoft Excel, which is the only proprietary software required for the exercises of this book and typically available to most students. The Geoinformatica GIS software is used in exercises involving geospatial computations. It is free and open source software and its installation package is bundled with the distribution package for this book. Subsequent chapters provide a concise description of theory on common statistical methods. First readers are introduced to several statistical concepts (such as mean, variance, covariance, bias, efficiency, consistency, confidence intervals). Then follows a discussion on linear and non-linear parameter estimation techniques where linear ordinary and generalised least squares estimation as well as the Gauss-Newton-Levenberg-Marguardt method are presented. In exercises the readers are requested to construct estimation methods in supplied Excel templates and study how they work. Concepts of biasedness, efficiency and consistency are also studied in the exercises, as well as the implications of violating assumptions for ordinary least squares estimates to be of a good quality.

The remaining chapters deal with various topics in water resource assessment. In each chapter a brief discussion of theory is given first, which is then followed by a description of exercises where readers can apply what they just have read. In most of the exercises readers need to construct and apply mathematical simulation models for finding answers to the given questions. For example, in one of the exercises the readers assess how predicted climate change could affect the runoff regime in southern Finland. A table of contents of the book can be found in Appendix 1

Distribution package and availability

A majority of the exercises come as Excel worksheet templates where the solutions can be in-filled according to the instructions given in the book. In addition, Excel worksheets containing technical model solutions are also included in the book distribution package. The term ‘technical’ means here that, while a model solution shows one way of building the computational procedure in question, it does not include answers to questions that address applying the procedure. Technical solutions are provided to help readers deal with situations where progression with the exercise is threatened such as when a certain step in solving the exercise remains unclear. This should diminish frustrations that easily surface when not understanding a single detail prevents from completion of the exercise. The access that is provided to model solutions is also intended to facilitate use of the book for independent learning as readers can compare their solutions to the suggested ones. Hence some confidence in the learning results can be attained without having a teacher available to comment on the solutions.

The distribution package includes the text as a pdf file, Excel templates for the exercises along with the technical solutions, Geoinformatica software installation package, and data needed in the geospatial computation exercises and can be downloaded free of charge from here. The pdf of a book alone can be downloaded from here.