Please note: Contents and the Preface for the book are shown below. ALL Demonstration software for the book (DTM, TOPMODEL, GLUE and TFM) can be found in the main Freeware menu section (see left)

From The Preface

There is one fundamental problem in studying hydrological systems – that most of the action takes place underground. Despite all the advances that have been made in remote sensing, ground probing radar and other techniques for exploring the subsurface, our knowledge of what goes on underground is still very limited. What we do know from detailed studies of water movement through soils and rocks in both the laboratory and small scale field studies is that the patterns of water movement are very complex indeed and change in very nonlinear ways with flow rates and degree of wetness. From the point of view of rainfall-runoff modelling at scales of practical interest (generally medium to large catchments), this complexity means that we cannot hope to reproduce all the details of the flow processes that give rise to the stream hydrograph: too much of the complexity involved is simply unknowable with current measurement techniques. At one level therefore, rainfall-runoff modelling is an impossible problem!

That has not, of course, stopped many different hydrologists and teams of hydrologists and collaborating hydrological Institutes from developing rainfall-runoff models. The word plethora springs to mind, despite the fact that at different times in my career I have been involved in the development and testing of many different models trying, in different ways, to do the job properly. This book cannot hope to enumerate all the different rainfall-runoff models currently available and I must apologise in advance to all the modellers whose models get only a brief mention or none at all. It is now virtually impossible for any one person to be aware of all the models that are reported in the literature, let alone know something of the historical framework of the different initiatives. I gave up making a list of available models when I reached a count of 100 more than 20 years ago (and that was before I too reported on another one in my PhD thesis).

Therefore I have attempted instead to try to outline the most interesting themes that are currently being explored in trying to improve our predictive capability in hydrology, while still trying to reflect the historical context of the subject. That means that the reader will find very little on the type of conceptual storage models that started the whole process in the 1960s and which are still widely used today. Nor will you find much on monthly water balance models. These represent the past, I hope the reader will find this book looking to the future. Even so, such a book can only represent a snapshot of the progress that has been made, and even then one that is only sharply in focus in some places. To provide a complete review of the literature in a primer such as this would be impossible but I have tried to provide references to both recent and classic papers to allow the reader to explore more of the research literature as needed.

There remains a continuing and important need for rainfall-runoff modelling for very practical problems in water resource assessment, flood forecasting, design of engineered channels, assessing the impact of effluents on water quality, predicting pollution incidents, and many other purposes. Fortunately, the situation is not quite as dire as suggested by the impossibility of predicting the flow pathways of water in detail. For many practical purposes we do not need to include all the detail in developing a predictive model. Indeed many successful rainfall-runoff models are essentially very simple indeed. This book is intended as an introduction to recent developments in rainfall-modelling that allow such practical predictions to be made. This will be done, however, in the context of an understanding that the impossibility of making detailed predictions of the flow processes involved necessarily means that all rainfall-runoff models can only be very approximate descriptions of the rainfall-runoff processes and, as such, must be considered to be uncertain in their predictions.

Thus, a whole section of this book is devoted to considerations of predictive uncertainty. This might normally be considered to be an "advanced" topic. My own view is that it is essential to understand the uncertainties in any environmental modelling exercise and that, using modern computer intensive Monte Carlo techniques, the estimation of uncertainty can be introduced in a conceptually very simple way. Such uncertainty estimation feeds directly into the assessment of risk in decision making and, in most practical cases, rainfall-runoff modelling is carried out precisely to make decisions. Should a flood forecast be issued given forecasts of river stage over the next 6 hours? What should be the reservoir overspill channel capacity to cope with the flood expected once in 50 years? The limitations of our hydrological knowledge are such that these decisions should be taken within a risk assessment framework, recognising the inherent uncertainty in our predictions.

It is hoped that understanding of the material set out in the text of this book, including the section on uncertainty estimation, will be enhanced by the demonstration software provided. The software is primarily based on methods that have been developed at Lancaster University over the last decade and many colleagues, research assistants and students have stimulated ideas or made direct contributions to that development. I would particularly like to mention Peter Young, whose idea that the data (rather than theory alone) may suggest an appropriate model structure have been an important influence. Andrew Binley, Kathy Bashford, David Cameron, James Fisher, Stewart Franks, Jim Freer, Rob Lamb, Matthew Lees, Paul Quinn, Renata Romanowicz, Karsten Schulz and Jonathon Tawn, all at, or formerly at, Lancaster have made important contributions to modelling projects. Collaborations with other groups has also been very important, particularly George Hornberger (Charlottesville), Bruno Ambroise (Strasbourg), Charles Obled and Georges-Marie Saulnier (Grenoble), Eric Wood (Princeton), Peter Germann (Bern), Sarka Blazkova (Prague) and Philippe Merot (Rennes). Parts of this book were written while on sabbatical in Santa Barbara with support from Tom Dunne and Jeff Dozier, in Lausanne with support from André Musy, and in Leuven with support from Jan Feyen and the Francqui Foundation.

I would also like to acknowledge my lasting debt to Mike Kirkby. A long while ago, his lectures while I was an undergraduate at Bristol University made me realise that it was possible to model geomorphological and hydrological systems in a thoughtful and insightful way, while the origins of TOPMODEL lie in his wealth of ideas during in my post-doctoral work with him at Leeds. His skill in capturing the essence of a problem in a set of relatively simple assumptions has been a lasting inspiration, even while I was struggling to understand what he was getting at!! I hope he will recognise some of that influence in what follows.

Keith Beven