journal article
LitStream Collection
Home
doi: 10.1029/WR024i010p01569pmid: N/A
Despite the importance of water use by the industrial sector, little economic research has been conducted to study the nature of industrial water demands. This paper uses a relatively simple model of input demands to test whether industrial water use is sensitive to changes to input prices or the level of production. Demands for four aspects of industrial water use (intake, treatment prior to use, recirculation, and treatment prior to discharge) are estimated as a system of simultaneous equations. The data set is a 1981 cross section of firm level observations on water use and expenditure. Empirical results are reported for four manufacturing subgroups: petrochemicals, heavy industry, forest industry, and light industry. Intake price elasticities range from −0.12 to −0.54 and intake elasticities with respect to the level of output range from 0.76 to 1.90. The cross price elasticity between intake and recirculation is positive for all subgroups (indicating substitutability) and ranges from 0.14 to 0.26.
Grygier, Jan C.; Stedinger, Jery R.
doi: 10.1029/WR024i010p01574pmid: N/A
Disaggregation models are perhaps the best available method for generating multiseasonal multsite streamflow sequences for use in water resource system simulations. They are the basis of two general packages: LAST and SPIGOT. Both packages use staged disaggregation procedures and condensed disaggregation models to avoid the excessive number of parameters required by the multisite monthly Valencia‐Schaake model. This paper examines the model size issue and what is gained and lost by the use of condensed models. In particular, when modeling the normal transforms of streamflow volumes, generated seasonal flows generally fail to sum to the specified annual total. A Monte Carlo study demonstrates that the discrepancy can be quite large and that alternative correction schemes perform quite differently.
Killey, R. W. D.; Moltyaner, G. L.
doi: 10.1029/WR024i010p01585pmid: N/A
Following evaluation of three potential sites, a fluvial sand aquifer with high (0.7–1.6 m d−1) linear average groundwater flow velocities was instrumented with a fully penetrating injection well and 82 monitoring installations over a 40‐m flow path length. Two natural gradient tracer migration experiments have employed pulse injections of a nonreactive radioisotope (131I) in a study of hydraulic conductivity structure and dispersive characteristics of the flow system. The use of a gamma‐emitting radionuclide in these experiments allowed measurement of tracer distribution by scanning fully penetrating boreholes with a scintillation detector. This technique provided many advantages in data collection over conventional tracers, including unprecedented resolution of the vertical distribution of tracer, measurements without pumping and the attendant concerns for flow system perturbation, and minimal manpower for sample analysis. Tracer breakthroughs at monitors along the axis of flow revealed the presence of heterogeneities in aquifer hydraulic conductivity at two scales. Over the 40‐m path length, six regions with hydraulic conductivities substantially different from adjacent strata were identified. Thicknesses of these regions varied from 0.2 to more than 3 m; their lateral extents (in the direction of groundwater flow) ranged from 15 m to more than 40 m. Smaller strata, with an average thickness of 0.34 m and lengths of less than 10 m, were detected within all six of the hydraulic conductivity regions, but there was no evidence of increases in local dispersivity with increased transport distance as a result of these variations. The six hydraulic conductivity regions do, however, have measurable effects on whole‐aquifer dispersion. Stratification characteristics of the tested aquifer, believed to have been deposited in a sandy braided river, are consistent with those observed in similar modern environments. Estimation of the magnitude and spatial structure of hydraulic conductivity variations in aquifers of concern would be greatly aided by studies of the stratification characteristics in well‐exposed modern or lithified sedimentary analogues.
Moltyaner, G. L.; Killey, R. W. D.
doi: 10.1029/WR024i010p01613pmid: N/A
Two natural gradient dispersion tests were conducted at the Chalk River Nuclear Laboratories, Ontario, in 1982 and 1983 in order to quantify field‐scale dispersion processes over a 20‐ and 40‐m distance of a radiotracer migration. The patterns of dispersion that were observed in the field using a dry access tube monitoring technique were compared with patterns predicted by the advection‐dispersion model with constant parameters. Analysis of the time‐concentration profiles of the tracer at observation wells located along the mean direction of the longitudinal spreading showed that the dispersion model adequately describes the dispersion patterns actually observed and averaged over the aquifer depth. The values of longitudinal dispersivity determined from continuous time‐concentration profiles were virtually equal to those obtained from laboratory‐scale columns. The dispersive properties of the aquifer were characterized at the local scale by a constant independent of travel distance dispersivity value of 0.0045 m. Averaging the test data over the aquifer depth, analogous to the information provided by a fully penetrating pumping well, yielded a full‐aquifer dispersivity varying in an irregular fashion from 0.16 to 0.06 m. The scale effect observed between local‐scale and full‐aquifer dispersivities is the result of inhomogeneity of the aquifer and the concentration averaging produced by the sampling instrumentation. The results of two tracer tests also showed that the use of the dry access tube monitoring technique for field experimentation provides detailed representation of spatial and temporal distribution of a radiotracer's activity and, consequently, provides a practical means to determine the groundwater velocity field and dispersive properties of porous materials.
Moltyaner, G. L.; Killey, R. W. D.
doi: 10.1029/WR024i010p01628pmid: N/A
Two natural‐gradient dispersion tests were performed at the Chalk River Nuclear Laboratories in Ontario in 1982 and 1983. During the field tests 131I was used to label the groundwater, and its activity was measured in situ, collecting continuous records of activity versus depth at selected times and of activity versus time at selected locations. Curves generated by a three‐dimensional analytical solution to the advection dispersion equation were fit to activity versus depth scans recorded at various locations in the test aquifer. Within the individual velocity regions the vertical dispersivities were 0.0006 and 0.0008 m, with the arithmetic average 0.0007 m as calculated from the 1982 data. The vertical dispersivity values calculated from 1983 data were 0.0014 and 0.0010 m for the high‐ and intermediate‐velocity regions, respectively, with an arithmetic average 0.0012 m. The increase in dispersivity was attributed to the error associated with the sampling and fitting procedures. The average vertical dispersivity of the tracer test aquifer was an order of magnitude less than the average longitudinal dispersivity (0.0010 m versus 0.010 m for the intermediate‐velocity region). Within individual velocity regions the vertical dispersivity of the porous medium was independent of the travel distance of the tracer and was proportional to the velocity.
Gvirtzman, Haim; Paldor, Nathan; Magaritz, Mordeckai; Bachmat, Yehuda
doi: 10.1029/WR024i010p01638pmid: N/A
A profile of tritium concentrations measured in the unsaturated zone in loessial sediments in a semiarid area is interpreted in terms of mobile and immobile water domains, according to a nonequilibrium transport model. The mobile domain is represented by percolating freshwater from both rain and irrigation, and the immobile one is represented by isolated fossil saline water pockets. The two domains are connected by partially saturated narrow passages within dispersed clay minerals. The transport of the mobile water is described by convective‐dispersive flow and by mass exchange between the two water domains. The relevant equations with the given initial and boundary conditions are solved numerically, and the simulated profile is adjusted to fit the measured one. In this study we concentrate on examination of the mass exchange law between the two domains. It was assumed that matrix characteristics vary in time due to the dispersion of clays at the interface between fresh and saline waters. Accordingly, a time‐dependent mass exchange was adopted, which made it possible to obtain an adequate reconstruction of the measured tritium profile. By using a least squares optimization procedure it was found that the best fit between the simulated and measured profiles is attained when the fraction of mobile water is 30%, and the rate of mass exchange decreases from 0.60 to 0.01 year−1 in 26 years. The proposed model implies is that it is the immobile water domain which contains the memory of the “high tritium period” (thermonuclear tests period) of the 1960s.
Zecharias, Yemane B.; Brutsaert, Wilfried
doi: 10.1029/WR024i010p01645pmid: N/A
Twenty geomorphologic parameters which constitute a well‐balanced representation of the area, length, elevation, form, and shape aspects of watersheds as well as the texture of their drainage networks were considered. On the basis of groundwater theory and the results of previous related studies, eight of these parameters were identified as being relevant to the process of groundwater discharge. Values of the selected parameters were obtained for representative watersheds located in a section of the Appalachian Plateaus. The data, which were generated from U.S. Geological Survey topographic maps, were then subjected to a factor analysis to assess the relative importance of the parameters with respect to their influence on groundwater outflow behavior. The analytical results and their interpretation showed that among the parameters that are related to groundwater outflow, total length of perennial streams, average basin slope, and drainage density are the ones that are most closely related to the process. The influences of these parameters on groundwater outflow behavior are independent of each other. Thus in relating groundwater outflow characteristics to basin morphology, only these parameters need be considered and the inclusion of additional parameters does not necessarily yield a better relationship and may result in redundancy. The findings of the study are consistent with existing concepts regarding base flow (or drought flow) and have been confirmed by an independent study that investigated the process based on a more mechanistic approach.
Zecharias, Yemane B.; Brutsaert, Wilfried
doi: 10.1029/WR024i010p01651pmid: N/A
The relationship between base flow recession characteristics in steep watersheds and geomorphologic and soil parameters is investigated. The formulation for the groundwater outflow was obtained by means of a hydraulic approach applied to a simple conceptual model for a hillslope. Long‐term flow data of 19 representative basins in the Allegheny Mountain section of the Appalachian Plateaus were analyzed on the basis of this formulation. Results showed that the reaction factor, which is a time scale of base flow recession, is dependent on the mean land slope, the drainage density, and the ratio (K/f) of the hydraulic conductivity and the drainable porosity. On account mainly of the nonuniform distribution of the physical characteristics within a basin, the reaction factor for a given watershed is somewhat variable with time, but the adoption of a constant value is useful to represent average conditions for a recession period. Analysis of the (K/f) dependency showed that macropores and other structural features may greatly affect the watershed base flow. Evaporation from groundwater appears to constitute only a minor portion of overall basin evaporation.
Howard, Alan D.; McLane, Charles F.
doi: 10.1029/WR024i010p01659pmid: N/A
Surface grains of noncohesive sediment eroded by emerging groundwater are acted upon by three forces, the tractive force of the cumulative surface flow contributed by upslope seepage, the local seepage force, and gravity. The balance of the force moments determines the mode and rate of transport. Seepage forces are strong in a narrow “sapping zone” at the upstream end of the emerging flow, where erosion occurs by mass movement and the surface gradient is determined by the balance of the seepage and gravity moments. Most of the erosion occurs in this zone, and the resultant backcutting triggers intermittent failure of overlying slopes in a “undermining zone” maintained at the angle of repose of the dry or damp sediment. In the “fluvial zone” downstream from the sapping zone the seepage force is small compared to the tractive force, and transport occurs by normal fluvial traction. The overall rate of sapping erosion in noncohesive sediments is determined by the capacity of fluvial transport to remove sediment eroded in the sapping zone. Prediction of sapping rates is complicated by the interaction between the geometry of the fluvial and sapping zones and the quantity and spatial distribution of seepage. A simulation model incorporating a groundwater flow model and sediment transport relationships closely replicates the observed evolution of sapping erosion in a two‐dimensional tank filled with noncohesive sand subjected to lateral groundwater flow.
Showing 1 to 10 of 25 Articles