GEOGRAPHICA BERNENSIA G 88 Preferential Flow Stokes Approach to Infiltration and Drainage Peter Germann University of Bern, Switzerland • Institute of Geography Preferential Flow - Stokes Approach to Inltration and Drainage GEOGRAPHICA BERNENSIA Published by Lectureres of the Institute of Geography University of Bern, Switzerland Series A African Studies Series B Field Guides (German only) Series G Basic Research (mainly in German) Series P Applied Geography (German only) Series U Teaching Materials (German only) G 88 Arbeitsgemeinschaft GEOGRAPHICA BERNENSIA in cooperation with Geographical Society of Bern Hallerstrasse 12 CH-3012 Bern (Switzerland) GEOGRAPHICA BERNENSIA G88 ___________________________________________________ Preferential Flow - Stokes Approach to Inltration and Drainage Peter F. Germann Professor Emeritus of Soil Science University of Bern Institute of Geography University of Bern The book draws from projects financially supported by the following agencies: Swiss National Science Foundation Swiss Federal Office for the Environment US-National Science Foundation US-Environmental Protection Agency US-Agricultural Research Service US-Forest Service New Jersey Agricultural Experiment Station Prof. em. Peter F. Germann Haldenackerweg 21 CH-3065 Bolligen (Switzerland) [email protected] Cover Stains in a paddy soil under rice, Pakistan. Courtesy by Anna Fehlmann and Bettina Wolfgramm Design by Alexander Hermann Published by Geographica Bernensia, University of Bern Graphics by Alfred Bretscher, Peter Germann, Alexander Hermann Institute of Geography, University of Bern Printed by Publikation Digital AG Bolacker 3, 4564 Obergerlafingen, Switzerland Geographica Bernensia 2014 Print version: ISBN 978-3-905835-34-2 Geographica Bernensia 2018 Corrected online version: doi: 10.4480/GB2018.G88 © GEOGRAPHICA BERNENSIA 2018 Institute of Geography, University of Bern, Switzerland Creative Commons Licences To Dorothee, who had different views on our first five years of retirement Preface About the book There are two ways of irrigating a potted plant. Water is either added to the plate at the pot’s bottom from where the soil sucks it up gradually or it is sprinkled on the surface from where it infiltrates and drains rapidly. Preferen- tial flow deals with the second way of irrigation. Preferential flow in permeable media concerns various disciplines like the sciences of soils and ecosystems at large, the hydrology of soils, catchments, and aquifers as well as various areas of geology and geophysics. Preferential flow also relates to numerous applications such as irrigation and drainage, and it raises issues of water quality and sanitation. Equally diverse are the backgrounds of experts dealing with preferential flow. This has produced a wide number of approaches to preferential flow of both, the experimental and the modeling kinds. This book offers a unifying concept based on first principles of hydrody- namics, leading to viscous flow in permeable media. This starting point not- withstanding, the issues are easy to understand, requiring but skills in math- ematics and physics offered to BSc-students attending university science or engineering courses. Einstein’s recommendation of Everything should be made as simple as possible, but not simpler is highly regarded. Part I develops a track of concepts from Newton (1729) to Stokes (1845, 1851) and then Lamb (1932). It presents the principles of viscous flow in permeable media, supplemented with the mathematical tool of the kinematic wave theory of Lighthill and Witham (1955). Accordingly, flow in partial- ly saturated permeable media is only dependent on time-variable boundary conditions and time-invariant properties of the medium. Thus, flow is treated with ordinary differential equations similar to Darcy‘s law describing flow in water-saturated permeable media. Part II applies these principles to data recorded in systems of various com- plexity. It also explores the theory’s limitations in space and time. In addition, it provides information detailed enough that the readers can calculate the ex- amples for themselves as personal exercises. Preferential Flow – Stokes Approach to Infiltration and Drainage may serve students, teachers, researchers, and practitioners alike as a hand- book that presents an old phenomenon in a fresh and comprehensive manner. i About the context It is a privilege to look back at the trials, errors and occasional successes over the last four decades presented in numerous articles, chapters, proceed- ings, proposals, theses and dissertations. The enviable position of a professor emeritus includes release from obligations whereas the forced cessation of experimentation offers an opportunity to reshape accumulated knowledge in a manner similar to sculpting: chisel all unnecessary material away from the chunk in the expectation that the remainder will catch attention. All began when Felix Richard, the first Soil-Physics Professor at the ETH in Zürich, sparked my interest in his subject during my education as a forestry engineer. He lectured about drainage of forest soils aimed at improving root aeration in soils of lesser quality at sites foreseen for reforestation. At that time the expansion of the forested area in the Swiss Pre-Alps was still a na- tional mandate based on, among other reasons, the Forest Hydrological Par- adigm that forests are the canopies that reduce flood-flows most efficiently. My PhD-project with Felix Richard at the Swiss Federal Forest Research sta- tion (EAFV, today‘s WSL) concentrated on the weekly soil-water balances of forest and grass-land patches, based on the Darcy-Richards relationships. The data base was two years of periodic in-situ suction measurements with hun- dreds of tensiometers and the determination of soil-water functions on dozens of soil samples. However, closing the water balance occasionally produced apparently negative evapo-transpiration values. The environmental physicist Paul Schmid-Haas at the EAFV had already in the early 1970s begun to ques- tion the general applicability of the underlying relationships. Difficulties with the Richards-approach continued during my Post-Doc with Herbert Lang‘s team at the Laboratories of Hydraulics, Hydrology and Glaci- ology (VAW) at the ETH-Zürich, where our group investigated hydro-mete- orological relationships in the first-order watershed Rietholzbach. It was sus- pected that occasional fast flows along root channels were severely disturbing the closure of the water balance. A major breakthrough came in 1979 when the Swiss National Science Foun- dation granted me a one-year study leave at the Institute of Hydrology (IH) in Wallingford (UK). Keith Beven and I collected reported deviations from Darcy-Richards type flow. Not surprisingly, the notion surfaced of kinemat- ic-wave theory as a feasible alternative, merely because Dave Woolhiser, then introducing the theory to hydrology, had visited the IH the preceding year. The continuing exchange with Keith was particularly intense from 1980 to 1982 when he lured me to the faculty of the Environmental-Sciences Depart- ment at the University of Virginia, where he had been offered a position as Assistant Professor in late 1979. We recorded flow in blocks of undisturbed soils with the university hospital‘s X-ray facility. We soon concluded that the ii integration periods of measuring soil water contents with gamma-ray- and neutron-probes were far too long to test any of our hypotheses during pre- sumed macropore flows, and that we should turn to direct flow measurements instead. The still cited paper of 1982 in Water Resources Research about Macropores and Water Flow in Soils expresses the climax of our co-oper- ation. We are now wondering about the fate of its 2013-revisit in the same journal. Projects with Bill Edwards and Bob Pierce during the summer breaks at the USDA-watershed at Coshocton (OH) (1981) and at Hubbard Brook (NH) watersheds, operated by the US-Forest Service (1982) added experimental evidence of fast flows in soils. At Coshocton we investigated in columns of undisturbed soils the spatial distribution of bromide in the vicinity of stained macropores after infiltrating a solution of dye and salt. We demonstrated the kinematic wave’s applicability at Hubbard Brook by measuring with a buck- et, stop watch and spring balance drainage flows from an inclined glacial till upon sprinkling on the isolated soil block above it. Technical advances in fast in-situ water-pressure logging switched our attention to the rise of a shallow water table as a reaction to ponded infiltration. The methodology was imple- mented at Adelphia Research Farm, operated by Cook College of Agriculture at Rutgers University, where I served as an Associate Research Professor of Soil Physics from 1986 to 1989 in the Soils-and-Crops department. However, the attraction of becoming the first chaired Professor of Soil Sci- ence at the Institute of Geography, University of Bern was stronger. Theses and dissertations applied the TDR-technique to the fast in-situ recordings of water-content waves due to controlled sprinkler infiltration. With the uncor- rupted mind of the practitioner André Chervet of the soil protection agency of the Canton of Bern (Switzerland) recognized immediately fast drainage as independent flow between saturation and field capacity when he first glanced at a graph of such a wave. Discussions with Liliana Di Pietro from INRA in Avignon (France) and Vijay P. Singh, then professor at the Louisiana State University and