Abstract: This 5-kilometer resolution raster (grid) dataset for the conterminous United States represents the average percentage of saturation overland flow in total streamflow estimated by the watershed model TOPMODEL. Saturation overland flow is simulated in TOPMODEL as precipitation that falls on saturated land-surface areas and enters the stream channel. TOPMODEL was applied to 5- by 5-kilometer areas across the conterminous United States using national climate, soils, and terrain GIS datasets. The model was run for 1,000 days for each 5- by 5-kilometer area. The average percentage of saturation overland flow in total streamflow was computed for the 1,000-day simulation in each grid cell.
Purpose: The saturation overland-flow dataset was developed to assist the U.S. Geological Survey's National Water-Quality Assessment (NAWQA) Program (Gilliom and others, 1995). The dataset was produced to help NAWQA national synthesis teams identify and quantify the watershed characteristics that theoretically affect flow paths of water through watersheds and their effects on water quality. Purpose references: Gilliom, R.J., Alley, W.M., and Gurtz, M.E., 1995, Design of the National Water-Quality Monitoring Program--Occurrence and distribution of water-quality conditions: U.S. Geological Survey Circular 1112, 33 p., available on the World Wide Web, accessed July 7, 2003, at URL http://water.usgs.gov/pubs/circ/circ1112/
Supplemental Information: TOPMODEL background: TOPMODEL (Beven and Kirkby, 1979) simulates the movement of water through a watershed from the time that it enters the watershed as precipitation to the time that it exits the watershed as streamflow. The version of TOPMODEL used to generate the saturation overland-flow dataset simulates the variable-source-area concept of streamflow generation (Dunne and Black, 1970) and the Betson (1964) partial-area modification of the Horton (1933) concept of infiltration-excess overland flow. The Horton concept states that streamflow during high flow conditions is generated by overland flow that is produced when precipitation rates exceed infiltration rates at the land-atmosphere interface. In the original concept of infiltration-excess overland flow, Horton (1933) assumed that streamflow during high flow conditions was produced by overland flow generated throughout the entire area of a watershed. Later, Betson (1964) proposed that in some watersheds streamflow during high flow conditions was generated from infiltration-excess overland flow produced on only a small part of the watershed area, an idea known as the "partial-area concept." Infiltration-excess overland flow is believed to constitute a major part of the storm hydrograph in areas where infiltration rates are less than precipitation rates; for example, in disturbed or poorly vegetated areas in subhumid and semiarid regions. In the variable-source-area concept (Dunne and Black, 1970), streamflow during high flow conditions is generated on saturated surface areas called "source areas," which occur in places where the water table rises to the land surface. The water table rises because precipitation infiltrates into the soil, moves down to the saturated subsurface zone, and then subsequently moves downslope in the saturated subsurface zone. Saturated land-surface areas commonly develop near existing stream channels and expand as more water enters the subsurface through infiltration and then moves downslope as saturated subsurface flow. Variable-source-area flow is thought to be an important streamflow-generation mechanism where infiltration rates are greater than precipitation rates; for example, in undisturbed vegetated areas in humid, temperate regions. Saturated land-surface areas are sources of streamflow during high flow conditions in several ways. Saturation overland flow (also called Dunne overland flow) is generated if subsurface hydraulic characteristics are not transmissive and if slopes are gentle and convergent. Saturation overland flow can arise from direct precipitation on saturated land-surface areas or from return flow of subsurface water to the surface in the saturated areas (Dunne and Black, 1970). The saturation overland-flow values in this dataset include only direct precipitation on saturated areas; return flow is treated as a separate flow component. Rain (or snowmelt) first is partitioned into infiltration-excess overland flow and infiltrated water. The infiltration rate and time to ponding computations are based on the Green-Ampt assumptions (see Beven, 1984). Soil permeability is assumed in the model to be spatially variable; this assumption allows infiltration-excess overland flow to be simulated for partial areas of the watershed according to Betson's (1964) concept. Water that infiltrates into the upper soil zone can evaporate or transpire at a rate dependent on the potential evapotranspiration rate (computed from air temperature and latitude) and the amount of moisture available in the upper soil zone. Some rain (or snowmelt) can bypass the unsaturated subsurface zone and move directly into the saturated subsurface zone through macropores, which are large pores in the soil that conduct water downward before the soil is completely saturated (Beven and Germann, 1982). The depth to the water table is decreased by water draining down from above or moving laterally from other parts of the watershed. If this contribution of water to the saturated subsurface zone at a particular location in the watershed is large enough, then the water table rises to the land surface, and the area becomes saturated. Saturation overland flow is produced by direct precipitation (rain or snowmelt) on saturated areas. The saturated land-surface area also can produce return flow if the water table rises above the land surface and exfiltration occurs. During all streamflow conditions, any water stored in the saturated subsurface zone is assumed to move downslope towards the stream channel. A portion of the subsurface zone water, depending on the volume stored and TOPMODEL parameters, drains into the stream. TOPMODEL assumes that the rate of subsurface flow into the stream increases exponentially as the water table moves closer to the land surface. This assumption is based on the idea that macropores can increase hydraulic transmissivity in the lateral direction and that macropores become increasingly abundant near the soil surface (Beven, 1984; Elsenbeer and others, 1992). Any drainage from the saturated subsurface zone into the stream increases the depth to the water table. The location of source areas (the saturated land-surface areas) within the watershed are affected by basin topography and soil hydraulic characteristics. This is consistent with observed spatial distributions of soil moisture and potentiometric surfaces (for example, Kirkby and Chorley, 1967; Dunne and others, 1975; Anderson and Burt, 1978). Source areas are found where subsurface water collects; these are locations where large upslope areas are drained, and where there is limited capacity for continued downslope movement. Topography, the three-dimensional configuration of gravitational effects on drainage, affects the location of source areas. As subsurface water moves downslope, it collects in topographically flatter convergent areas. The degree of convergence determines how much upslope area drains down to a given location. The slope of the flat areas affects the "ability" of water to move farther downslope. Soil hydraulic characteristics (hydraulic conductivity and soil depth) determine the transmissivity at a location and affect the ability of water to move farther downslope. Topography and soil characteristics, watershed latitude, and a time series of precipitation and air temperature must be specified to use TOPMODEL. (The methods used to compute the soil and topography parameters are described in the Process_Step sections.) The watershed latitude is used to generate a time series of day length that, along with the time series of temperature, is used to calculate potential evapotranspiration. The precipitation and temperature time series are generated in the model internally using stochastic equations and climate characteristics such as average storm intensity and mean daily temperature. Description of the climate characteristics and the stochastic approach also are given in the Process_Step sections of the metadata. TOPMODEL is run on a daily time step for days with no precipitation and on an hourly time step for days with precipitation. TOPMODEL predicts streamflow, estimates overland and subsurface flow, and estimates the depth to the water table. Infiltration-excess overland flow for a given time step is calculated from the estimated time to ponding and the precipitation intensity. Saturation overland flow is calculated from the areal extent of the saturated land-surface areas and the precipitation intensity. Subsurface flow is computed as a function of the maximum subsurface-flow rate (determined by topography and soil characteristics) and the watershed average depth to the water table. The watershed average depth to the water table is computed by water balance; that is, by tracking input (precipitation) and output (overland flow, subsurface flow, and evapotranspiration). A more detailed description of TOPMODEL can be found in Wolock (1993). TOPMODEL references: Anderson, M.G., and Burt, T.P., 1978, The role of topography in controlling throughflow generation: Earth Surface Processes, v. 3, p. 331-344. Betson, R.P., 1964, What is watershed runoff?: Journal of Geophysical Research, v. 69, p. 1541-1552. Beven, K.J., 1984, Infiltration into a class of vertically nonuniform soils: Hydrological Sciences Journal, v. 29, p. 425-434. Beven, K.J., and Germann, P., 1982, Macropores and water flow in soils: Water Resources Research, v. 18, p. 1311-1325. Beven, K.J., and Kirkby, M.J., 1979, A physically based, variable contributing area model of basin hydrology: Hydrological Sciences Bulletin, v. 24, p. 43-69. Dunne, T., and Black, R.D., 1970, Partial area contributions to storm runoff in a small New England watershed: Water Resources Research, v. 6, p. 1296-1311. Dunne, T., Moore, T.R., and Taylor, C.H., 1975, Recognition and prediction of runoff-producing zones in humid regions: Hydrological Sciences Bulletin, v. 20, p. 305-327. Elsenbeer, H., Cassel, K., and Castro, J., 1992, Spatial analysis of soil hydraulic conductivity in a tropical rain forest catchment: Water Resources Research, v. 28, p. 3201-3214. Horton, R.E., 1933, The role of infiltration in the hydrologic cycle: EOS, Transactions, American Geophysical Union, v. 14, p. 446-460. Kirkby, M.J., and Chorley, R.J., 1967, Throughflow, overland flow and erosion: Bulletin of the International Association of Scientific Hydrology, v. 12, p. 5-21. Wolock, D.M., 1993, Simulating the variable-source-area concept of streamflow generation with the watershed model TOPMODEL: U.S. Geological Survey Water-Resources Investigations Report 93-4124, 33 p. The use of firm, trade, and brand names is for identification purposes only and does not constitute endorsement by the U.S. Government.

Progress: Complete
Update Frequency: None planned

Theme Keywords: None, TOPMODEL, Dunne overland flow, Saturation overland flow, Streamflow generation
Place Keywords:

Access Constraints: None.
Usage Constraints: The values of saturation overland flow in the dataset should be viewed as highly uncertain. Uncertainty in the saturation overland-flow percentages is high because there are many sources of potential error and uncertainty in the TOPMODEL simulations. For example, the climate, soils, and terrain data required by the model are spatially coarse and possibly inaccurate. The spatial coarseness of the climate data produces a noticeable coarseness in the estimated percentages of saturation overland flow in total streamflow. In some Western States, there are very few meteorological stations for computing the climate characteristics. This data sparseness propagates into the TOPMODEL results and creates a splotchy pattern in the estimated percentage of saturation overland flow. Another important source of uncertainty in TOPMODEL (and any other model) is termed "model uncertainty." This type of uncertainty exists because the model is only a simple representation of the hydrologic processes assumed to be most important in determining how water moves through the environment. The estimated flow components from TOPMODEL will be in error to the extent that the model does not represent important complexities in the real hydrologic system.