Chapman Conference
State-of-the-art in hillslope hydrology


Larry Band, University of North Carolina, Chapel Hill, NC

Jeff McDonnell, Oregon State University, Corvallis, OR


Scientific Technical Committee:


Sponsored by:


Co-sponsored by:


Focus statement:

Watershed hydrology has evolved from a set of agricultural and engineering applications focusing on problems at distinct scales.  The agricultural approach focused on issues dealing with soil water availability and movement at the scale of the soil column. Detailed, bottom up consideration of soil physical and chemical processes could be incorporated as the soil column could be well described, and potentially extended to the level of an agricultural field as a relatively homogeneous, basic unit of production and irrigation needs.  Water resources engineering issues of flood flow and water supply, on the other hand, generally dealt with significantly larger catchments and traditionally took a top-down approach to the prediction and understanding of runoff generation and flood frequency.  Specific process observation, however, was also necessarily limited to plots, and generally borrowed concepts drawn from the agricultural community. Therefore, initial concepts used for runoff and streamflow generation in watershed hydrology were dominated by the infiltration-excess mechanisms. This is exemplified by Horton's work in the New York City watershed in his descriptions of infiltration and runoff generation. Significantly, this approach conceptually treated water movement through a hillslope as a spatially uniform process. What we commonly refer to as Hortonian runoff was incorporated into mathematical models that became widely used in the water-resources engineering community. It is interesting to note that models that are essentially based on plot scale infiltration excess concepts are still dominant for representing land surface hydrology in atmospheric circulation models.

In the 1960s and 1970s, an accumulation of field observations at the hillslope and catchment scale appeared to contradict previously held concepts of uniform precipitation excess mechanisms and the models that had come into general use.  Many of these observations were made in forested and other less disturbed catchments where infiltration capacity was much higher, and significant amounts of water entered the soil and moved laterally through the subsurface to form distinct wetness zones. The variable-source area concept of streamflow generation required a fuller treatment of soil water flowpaths at the hillslope scale than was possible with the current generation of models. As a result, a set of mathematical models were then refined or newly developed to accommodate these observations. These models first concentrated on extending the equations for matrix flow in the soil column to include a lateral flux (e.g. Freeze, 1972) often short of a full 3-dimensional model due to the limitations in computing resources and in spatial information on soil properties required. A set of models that sought to reproduce the effects of a full 3-d flow field were developed that coupled a vertical infiltration model with a method for moving soil moisture downslope either by attempting to trace topographically controlled flowpaths with a local Darcy flux or developing a conceptual redistribution scheme (e.g. Topmodel, TOPOG Steady  State.) 

During this time, a new set of scientific questions have evolved, separate in focus and intermediate in scale to traditional agricultural and water resources engineering. These dealt with problems in non-point source contaminants, acid rain, and other integrated watershed hydrologic processes and required a fuller understanding of the vertical distribution of flowpaths through hillslopes, their velocities and residence (or contact) times, and temporal dynamics on storm to seasonal time scales. Partially in response to these emerging problems, in the 1980s and 1990s, catchments (both hillslopes and streams) have been intensively studied by simultaneously monitoring the distribution and flux of water and chemicals at many locations in space and time. Technical advances in methods of measuring soil water distribution in the soil profile (TDR), soil and substrate structure (ground penetrating radar), and the use of chemical and isotopic tracers to source streamwater during this time have dramatically increased the amount of information available to study basic processes by which water moves through hillslopes. The information is not consistent in certain cases, between isotopic and chemical signals, with signals derived from more standard hydrometric methods.

As our ability to measure flowpaths and the evolution of soil moisture patterns and chemistry progressed, it was recognized that the concept of matric flow as the dominant subsurface pathway may not be universally applicable. It is now recognized that the process of macropore flow and other preferential flowpaths through hillslopes are significant at least under certain conditions.

Over the last 10-15 years, their has been a rapid advance in our ability to observe distributed processes in watersheds over a range of scales. The intensive monitoring of catchments has been made possible by automatic sampling and remote sensing techniques. The advent of isotopic methods to study the source and evolution of stormwater have added significantly to our knowledge base, sometimes yielding information that appears to conflict with previously held concepts and more traditional hydrometric data collection methods.  These observations have raised many questions and issues about all of our commonly held concepts of streamflow generation and soil water dynamics.

Current model development continues to seek to incorporate the distribution and dynamics of soil moisture and the set of possible flowpaths that become important under given conditions. However, at present there does not appear to be either a general understanding or consensus on how the dominant processes of water input, internal flow (pathway dynamics) and outflow (including evapotranspiration) interact and evolve on different hillslopes. At the same time, an understanding of these processes and their interactions have become critical to a set of pressing scientific questions regarding flood generation, water supply, water quality and land/atmosphere interactions.  The specific problems of flow path effects or soil moisture patterns have arisen in a number of conferences and workshops focused on other topics, such as the recent Chapman Conference on Nitrate and the joint Chapman/SSA Conference on GIS and solute transport, but have not been the focus of an extended workshop for some time.  In this light, it is very timely to convene a workshop dedicated to the review and critical analysis of recent research and understanding of hillslope hydrology. We expect the concepts that emerge from such a workshop should have a significant influence on the foundations and implementation of mathematical models that can be used by the scientific community to address questions about hydrological processes and water quality.


Objectives of the Meeting:

Objectives of the conference are to define a robust physical description of how water moves into, through and out of hillslopes in the context of how this information can and should be captured in model formulation, calibration and testing at small to large catchment scales.


More Information:

Visit AGU's meeting page for more information on this Chapman Conference.




Conference Program