Premium
Automatic Generation of Flow Nets with Conventional Ground‐Water Modeling Algorithms a
Author(s) -
Fogg Graham E.,
Senger R. K.
Publication year - 1985
Publication title -
groundwater
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.84
H-Index - 94
eISSN - 1745-6584
pISSN - 0017-467X
DOI - 10.1111/j.1745-6584.1985.tb00778.x
Subject(s) - streamlines, streaklines, and pathlines , stream function , hydraulic head , flow (mathematics) , computer science , algorithm , head (geology) , hydraulic conductivity , groundwater flow , potential flow , geometry , mechanics , geology , mathematics , groundwater , geotechnical engineering , physics , geomorphology , soil science , vorticity , vortex , aquifer , soil water
Though flow nets offer one of the best means of visualizing ground‐water flow fields, accurate construction of flow nets can be very difficult in heterogeneous, anisotropic media or in regional cross sections where vertical exaggeration of scale is great. A method was developed for automatically generating the streamlines of a flow net using conventional ground‐water flow modeling algorithms. The method can be implemented using most any of the existing computer programs for flow modeling, and no modification of the programs is necessary. One need only replace the hydraulic head and conductivity terms in the steady‐state ground‐water flow equation with the stream function and an inverse form of hydraulic conductivity, respectively. Possible boundary conditions include prescribed stream function (Dirichlet type) and prescribed stream function gradient (Neumann type). The former can be determined from prior knowledge of recharge or discharge rates or from boundary fluxes computed in previous simulations of hydraulic head distribution. The latter can be determined from data on hydraulic head along the boundaries, raising the possibility of solving for the streamlines directly without first solving for the head distribution. The method is generally not appropriate when sources or sinks occur inside the flow region, and it will not handle transient conditions. Accuracy of the method is demonstrated in three hypothetical test problems and a vertical‐plane model through the Palo Duro Basin of Texas and New Mexico.