Subsurface Stormflow Research Articles

Mathematical simulation of interdependent surface and subsurface hydrologic processes

It has become popular with some hydrologists to consider runoff produced by rainfall intensities in excess of soil infiltration capacity (‘Horton’ excess) and runoff produced by rain falling on saturated soil zones (saturation excess) to be alternative or competitive models for describing surface runoff. On the contrary, we contend with Freeze (1972) that these two mechanisms, along with subsurface stormflow, are three parts of a continuum of hydrologic response that are controlled by the relative properties of the soil and the rainfall. We introduce a model which trades some of the generality of the Freeze (1972) hillslope model for increased efficiency, by selecting efficient approximations for key elements in hillslope hydrodynamics: infiltration and movement in the unsaturated flow zone. The hillslope is considered to consist of two soil layers with the lower soil capable of restricting vertical flow at the interface to create a perched aquifer and subsurface stormflow. The slope of the interface is arbitrary. A kinematic method for routing unsaturated vertical flow is described and is linked consistently with a modern analytically derived infiltration model, which operates when rainfall exceeds surface saturated hydraulic conductivity Ks. The model is used to demonstrate several of the relationships between rainfall flux, soil hydraulic properties, hillslope geometry, and runoff characteristics. The model is also applied to simulate experimental data from a hillslope in Western Australia. Subsurface stormflow recessions are demonstrated to be affected by the ratio Ks/ql, anisotropy, and spatial distribution of saturated hydraulic conductivity Ks, where ql is the rate of loss from the perched aquifer. Vertical growth of the saturated zone depth is demonstrated to have a potentially greater effect on runoff than horizontal movement in cases where anisotropy and/or slope is not severe.

Vertical soil water movement in a tropical rainforest catchment in northeast queensland

AbstractIn a tropical rainforest catchment, shallow piezometers respond almost instantaneously to rainfall, but the dominant ground water recharge mechanisms are not well understood. To improve understanding, the downward movement of soil water on a runoff plot was traced using tritiated water injected at 0·20 m below the surface which marks the lower boundary of active subsurface storm flow. The tritium pulse was translated slowly down the profile, apparently dominated by interstitial piston flow on the lines described by Zimmermann's theoretical model. This recharge mechanism accounted for about 35 per cent of rainfall or 50 per cent of throughfall. The pulse's advance may have also been delayed by the upward movement of soil water indicated by the distribution of hydraulic potential under different hydrological conditions. The result was an increase in soil water transit time particularly below 1·0 m. There was also evidence in the tracer profiles for rapid by‐pass flow but the volumes concerned could not be quantified in this experiment.

Linear theory of subsurface storm flow

A solution to the subsurface storm flow problem is developed through linearization of the complete one‐dimensional governing equation. Water surface profiles and outflow hydrographs obtained from this approximate theory are compared with the exact solution and with the kinematic approximation to this problem for a number of cases, characterized by the parameter λ = 4 i cos θ/K sin2θ. The accuracy and range of applicability of the proposed theory are greater than that of the kinematic theory. Its results may be considered satisfactory for the cases with λ ≤ 1; however, the kinematic theory is a preferable substitute for λ ≤ 0.25.

On subsurface stormflow: an analysis of response times

A simple theory for predicting the response times of unsaturated and saturated flows on hillslopes, based on kinematic wave equations, is presented. The sensitivity of unsaturated zone lag and time of concentration of the saturated zone to hillslope and soil parameters, antecedent moisture and input rates is examined. It is shown that for many combinations of parameters subsurface response times are too long to be considered as stormflow; but that the combinations of high hydraulic conductivities, shallow soils and wet initial conditions that lead to fast responses are quite reasonable when compared with field studies of subsurface stormflows. It is suggested that the kinematic theory provides a basis for a practically useful predictive model.

On subsurface stormflow: Predictions with simple kinematic theory for saturated and unsaturated flows

This paper provides analytical solutions for some simple cases of subsurface stormflow. The solutions are based on kinematic approximations in both unsaturated and saturated zones. Solutions for rising, falling, and partial equilibrium hydrographs are given. The analytical model has been applied to data collected on a hillslope of the East Twin Brook catchment, Mendip, United Kingdom. Measured and predicted hillslope hydrographs were in good agreement.

Macropores and water flow in soils

This paper reviews the importance of large continuous openings (macropores) on water flow in soils. The presence of macropores may lead to spatial concentrations of water flow through unsaturated soil that will not be described well by a Darcy approach to flow through porous media. This has important implications for the rapid movement of solutes and pollutants through soils. Difficulties in defining what constitutes a macropore and the limitations of current nomenclature are reviewed. The influence of macropores on infiltration and subsurface storm flow is discussed on the basis of both experimental evidence and theoretical studies. The limitations of models that treat macropores and matrix porosity as separate flow domains is stressed. Little‐understood areas are discussed as promising lines for future research. In particular, there is a need for a coherent theory of flow through structured soils that would make the macropore domain concept redundant.

Storm runoff processes and subcatchment characteristics in a new zealand hill country catchment

AbstractThe quickflow responses of six subcatchment areas in a small hill country catchment in the Craigieburn Range, South Island, New Zealand, were compared for a range of storm sizes, rainfall intensities and antecedent wetness conditions. Topography and soil characteristics suggested that all subcatchments would receive subsurface stormflow input, but that some would receive larger saturation overland flow inputs than others. Quickflow yields and response ratios were positively correlated with storm size and antecedent wetness conditions in the subcatchment most suited to producing saturation overland flow. In subcatchments more likely to be dominated by subsurface flow, quickflow yields and response ratios were positively correlated with storm size, but were either not correlated, or negatively correlated, with antecedent wetness. Quickflow responses were either not significantly or negatively correlated with rainfall intensity variables. Quickflow from the subcatchment most suited to produce saturation overland flow providing an increasing proportion of total catchment quickflow in larger storms and as antecedent conditions became wetter. Subcatchment responses varied greatly in space and time and there was less pattern to the variation than had been expected. Where topographic and pedologic conditions permit substantial responses to storm rainfall by both saturation overland flow and subsurface stormflow, simple topographic and soil indicators may not be useful guides to the relative importance of runoff mechanisms, or to the identification of runoff‐source areas.

Kinematic subsurface stormflow

This paper builds on the analysis of Henderson and Wooding (1964) in comparing simplified models of subsurface stormflow for the case of a sloping soil mantle in which the hydraulic conductivity is constant throughout. It is shown that models based on a kinematic wave formulation may be good approximations for those based on the extended Dupuit‐Forchheimier assumptions for values of the nondimensional parameter λ = 4i cos θ/(K sin2θ) less than about 0.75 in terms of predicting both water table profiles and subsurface stormflow hydrographs. By using this critical value of λ the range of values of slope angle and saturated hydraulic conductivity for which the kinematic approximation is valid can be specified. A comparison is made with slope angles and hydraulic conductivities at field sites where subsurface stormflow has been shown to be an important component of catchment storm response. It is concluded that the kinematic approximation may be useful for cases of practical interest.

Use of specific conductance and contact time relations for separating flow components in storm runoff

The difference between the dissolved‐solids concentration in base flow and in storm flow has often been used as the basis for separating components of flow. However, an analysis that explicitly relates the amount of time that runoff water has been in contact with watershed soils to the resulting dissolved‐solids concentration shows that simple mass balance chemistry methods for hydrograph separation are misleading. Field studies of surface and subsurface storm flow, when coupled with laboratory determination of the relationship between contact time and dissolved solids content of a soil water mixture, suggest that the residence time of infiltrated water is as short as a few hours in the cases studied. In those cases, hydrograph separation methods based on the simple mass balance equation for the dissolved solids will yield considerable overestimates of the base flow component.

Hydrologic behavior of a forested mountain soil in coastal British Columbia

Inferences with respect to the hydrologic behavior of a West Coast forested mountain soil are based on measurement of the water potential field during the wetting and drainage phases of simulated rainfall events. The field was measured as a continuous function of time with an array of 13 tensiometers in combination with pressure transducers and a data collection system. Measurements were carried out with the forest floor intact, partly disturbed, and totally removed. Inferences were made by interpreting the water potential fields as ‘fingerprints of hydrologic behavior.’ It was concluded that during a simulated rainfall event, water flow through the profile was partitioned between root channels and the soil matrix which is transversed by the channels. The proportion of the flow conducted by the channels was at its maximum during the non‐steady state phase of the event, decreasing to a minimum as steady state was approached. During the drainage phase of the event the channels did not contribute to downward flow. The process of internal infiltration is discussed. It causes the soil matrix to wet up both from the surface of the H horizon of the forest floor and from the wetted periphery of the root channels. Results indicated that simulated disturbance of the forest floor down to the mineral soil caused a shift in the water flow pathway from the channels to the soil matrix. It was speculated that this shift was due to closure of the channel openings. It was speculated that the flashy response of streams to rainfall events is related to rapid subsurface storm flow through root channels. Also, because water flow through the soil matrix is much slower than that through root channels, it is likely that forest floor disturbance on a watershed‐wide scale would result in an increase in the time lag between the rainfall event and the corresponding streamflow event.

A field evaluation of subsurface and surface runoff: I. Tracer studies

Recent studies have demonstrated the occurrence of a variety of storm runoff processes, including those associated with variable source areas. However, there is a need for experimental methods for studying the spatial variability of these processes in the field. One promising method is the use of tracers. When two radiotracers, 51Cr and 82Br, were used in a pilot study of surface and subsurface storm runoff, irregular spatial flow patterns were observed. These patterns were primarily related to the thickness of the surface soil horizon and showed the value of tracers for quantifying horizontal and vertical profiles of flow pathways. Several improvements are possible in measurement techniques, and should provide a means for obtaining detailed spatial definition of surface and subsurface storm flow pathways in the field.

Streamflow generation

During the past 10 years under the impetus of the International Hydrological Decade there has been a burst of research activity aimed at obtaining better insight into the mechanisms of streamflow generation. The research has focused on an examination of the ways in which water moves from hillslopes into small stream channels during and between rainfall events in upstream tributary basins. The research has been of two types: field measurements in representative experimental drainage basins and theoretical studies using mathematical models of hydrologic processes, carried out with the aid of a digital computer. The field studies have identified the sources of lateral inflow to streams as overland flow, subsurface storm flow, and groundwater flow. They have also proven that most overland flow is generated either on small upland partial areas that are more or less fixed in size and that are controlled by the distribution of soil types or on expanding and contracting ‘variable source areas’ that are adjacent to streams and that are controlled by the topographic and hydrogeologic configuration of the hillslopes. They have shown qualitatively that the relative importance of the different processes depends on such controls as climate, geology, topography, soil characteristics, vegetation, and land use. This variety of field conditions coupled with the complexity of the processes themselves has hindered the growth of generalized interpretations based solely on the field evidence, and hydrologists have increasingly turned to mathematical modeling to isolate the various components of the system. Physically based mathematical models, in the form of boundary value problems based on the partial differential equations of flow, are now available for all the processes involved in streamflow generation. The concepts of open channel hydraulics form the basis for models of flow in a stream channel and for overland flow. The concept of saturated and unsaturated porous media flow has been brought to bear on models of infiltration, subsurface storm flow, and groundwater flow. Applications of these models at field sites and for series of hypothetical cases have identified the physical parameters that control the various observed mechanisms of downslope water movement and for some environments have determined the necessary conditions for the dominance of one mechanism over the others.

Role of subsurface flow in generating surface runoff: 2. Upstream source areas

Runoff simulation for rainfall events on hypothetical upstream source areas, carried out with a deterministic mathematical model that couples channel flow and saturated‐unsaturated subsurface flow, provides theoretical support for the runoff‐generating mechanisms observed in the field by Ragan and Dunne. The simulations show that there are stringent limitations on the occurrence of subsurface storm flow as a quantitatively significant runoff component. Only on convex hillslopes that feed deeply incised channels, and then only when saturated soil conductivities are very large, is subsurface storm flow a feasible mechanism. On concave slopes with lower permeabilities, and on all convex slopes, hydrographs are dominated by direct runoff through very short overland flow paths from precipitation on transient near‐channel wetlands. On these wetlands surface saturation occurs from below because of rising water tables that are fed by vertical infiltration rather than by lateral subsurface flow. These conclusions, when coupled with field observations that show classic Hortonian overland flow to be a rare occurrence in vegetated humid environments, have implications in the planning of field instrumentation networks, and in the designing of hydrologic response models.

On factors governing subsurface storm flow in volcanic ash soils, New Zealand

The possible occurrence of subsurface storm flow in volcanic‐ash soils was investigated by means of small lysimeters with and without window openings to measure lateral flow. The information obtained from the lysimeters led to a further investigation of apparently aberrant results by a number of experimental methods largely devised under field conditions. The results obtained from these experiments show that (1) a resistance to wetting exhibited by the surface soil at a low moisture content, (2) a low permeability of the B horizon compared with that of the A horizon, apparently related to an initial anisotropy in the ash beds, and (3) the transition from a fine to a coarse‐textured layer cause a lateral flow through the surface soil on sloping ground to take place during rain storms. The magnitude of this flow largely depends on the intensity of a rain storm.

Appendix B—Report of the subcommittee on subsurface flow

Because of war programs, no attempt has been made to expand the membership or the activities of this subcommittee during the past year. A statement of research‐needs,to comprehend subsurface stormflow better, was presented at the last Annual Meeting and appears in Part V, Transactions of 1944, pages 743–746.Exchange of correspondence has brought to light a growing interest in the phenomena of dynamic subsurface water‐movements during storm periods. During the year a valuable paper by Parsons [see 1 of “References” at end of report] has emphasized the importance of interpreting the flood hydrographs in terms of “the various elements of basin storage” which he recognizes as outflows from surface, subsurface, and ground‐water storage. Following the general method of Barnes [2], he has presented a graphical recession analysis indicating that storm runoff appears as three approximately straight lines when plotted on semiloganthmic paper—representing the surface, the subsurface, and the ground‐water fractions. This concept is then used in the calculation flood hydrographs from precipitation records. Stafford and Troxell [3] have likewise recognized runoff an being of three types: (a) surface storm‐runoff; (b) subsurface storm‐runoff; and (c) ground‐water seepage, and have associated these throe types with basin characteristics in the San Bernardino and Eastern San Gabriel mountain drainages in southern California.