An automatic quadrilateral mesh generation algorithm applied to 2-D overland flow simulations

Lu Sun, Guoqun Zhao, Gour Tsyh Yeh

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


This paper presents an automatic quadrilateral mesh generation algorithm to provide 2-D meshes for simulating overland flow and transport problems in watershed systems. The conformal mapping method is studied deeply and specific treatment strategies are established according to the width of rivers and the area of other water storage zones. A conformity treatment method is proposed to deal with the incompatibility problems during discretizing the complex geometry with a plurality of sub-domains. This method ensures grid conformity through removing certain overlapped nodes and creating new appropriate nodes on common boundaries of the multi-domain mesh. Aiming at the undesired deviation of mesh boundaries from geometric contours, a relative position-percent interpolation method is proposed and a supplementary match method related to dead ends is presented. And, the double effects of projection and smooth for boundary nodes are achieved. An objective function method is proposed to locally optimize the degenerated quadrilaterals located on concave or large-curvature curves. In order to accommodate the need for overland simulations, the 1-D/2-D correspondence on river reaches, junctions with and without storage, ponds, lakes, dead ends and control structures is established. Finally, practical applications are provided to demonstrate the accuracy and reliability of the quadrilateral meshing algorithms proposed in this paper.

Original languageEnglish
Pages (from-to)1283-1303
Number of pages21
JournalComputational Geosciences
Issue number5
StatePublished - 1 Oct 2018


  • Boundary match
  • Conformal mapping
  • Quadrilateral mesh
  • Quality optimization
  • Scaled Jacobian
  • Water flow


Dive into the research topics of 'An automatic quadrilateral mesh generation algorithm applied to 2-D overland flow simulations'. Together they form a unique fingerprint.

Cite this