The Application Of Local Grid Refinement Technique Based On Modflow

Shared Node Local Grid Refinement Technique (LGR) is a two-way iterativetechnique coupled grid refinement technique based on Modflow. This technique as anadvanced grid refinement technique can save much time but ensure accuracy.However, the basic principle of this technique is not yet deeply understood. Therefore,it is seldom used in groundwater flow modeling in our country. On the basis ofprofound understanding of its basic principle, in this study we applied this techniqueto the numerical simulation of groundwater flow for alluvial-proluvial fan of Hutuoheand tested if this technique could improve the accuracy of simulating in local area.Because the LGR is only fit for grid refinement of rectangular inerratic area, werecomposed its source code to verify that grid refinement technique could also beused in irregular area according to the results of numerical examples.This study is composed of three parts as follows:(1) Firstly, this part describes the basic principle of the local grid refinementtechnique, which includes how to deal with the source/sink of the contact boundaryof child model and parent model, the format of LGR Control File, and the processingof other relevant files. Finally, we operated the established groundwater flow model ofalluvial-proluvial fan of Hutuohe by using the LGR source program (local refinement),and got the flow field of the model. By comparing the results of the model with that ofgeneral groundwater model, rationality and accuracy of the modeling of irregular areabuilt up by LGR were tested.(2) In this part, we preliminarily discussed the feasibility that if the LGR basedon Modflow could be used in irregular area. We also mainly introduced thedevelopment of groundwater flow modeling built up by LGR in irregular area, andobtained the water budget and the flow field of the model in combination with anumeric example. Finally, we compared the results of the model with that of generalgroundwater model to examine the rationality and accuracy of the modeling of irregular area built up.In this study, the outcomes showed that all the models built up by LGR couldembody the depression cone in a more detailed way. Meanwhile, the model ofirregular area built up by LGR also could be run by revising the source code.Therefore, we could concludes that (1) the model built up by LGR could improve thesimulation accuracy of local area;(2) the LGR technique could fit for the irregulararea by revising the source code of the subprogram.