Mesh And Spatial Support

This section groups the theory notes for mesh and spatial support choices in HydroModPy.

Mesh choice is not only a preprocessing detail. It controls:

• which basin, river, geology, and boundary constraints are honoured;

• how fields are projected to solver cells;

• which numerical method is appropriate;

• whether solver comparisons isolate backend behaviour or mix backend and support effects.

Use this section for the concepts. Use the capability-gallery mesh pages for concrete, versioned examples and figures.

Scientific Role

HydroModPy separates the modelling problem into several layers:

geographic domain
-> planar support or mesh
-> vertical representation
-> field-to-cell transfer
-> solver-specific discretization
-> validation or comparison evidence

That separation matters because the same physical question can be interpreted on different supports:

• a structured sgrid support for legacy MODFLOW-NWT-style workflows;

• a MODFLOW 6 support that may remain structured or consume a DISV-style irregular mesh;

• a triangular finite-volume support used by the Boussinesq backend;

• a saved mesh bundle reused by comparison and gallery workflows.

Representative gallery evidence

Mesh concepts are easier to read with concrete artifacts next to the text. The full case pages remain in Mesh Gallery; the figures below are the minimum evidence to keep in mind while reading this section.

Mesh quality diagnostics are not decorative. They are the checks that decide whether a discretization is acceptable before solver results are trusted.

Constraint-balance figures show why mesh choices should be discussed as modelling assumptions, especially when hydrography, geology, and basin boundary constraints compete.

Main Questions

Question	Scientific page	Gallery evidence
Which mesh families exist and why do they matter?	Mesh And Discretization Strategies	Mesh Gallery
How do regular and irregular meshes compare for the same cell budget?	Regular And Irregular Mesh Cell Budgets	Resolution Sensitivity Across Scales
How are geology, hydraulic properties, or forcing mapped to cells?	Field-To-Cell Parameter Transfer	Hydraulic Properties
Which mesh quality checks should be read before trusting a solve?	Mesh Quality And Acceptance Criteria	Mesh Quality Diagnostics (10 km2)
Why does MODFLOW 6 use XT3D on irregular meshes?	XT3D On Irregular DISV Meshes	MODFLOW 6 Irregular Triangles: Why XT3D Is The Default
How does vertical representation affect storage and thickness?	Vertical Representation And Storage Assumptions	Validation and comparison cases that reuse the same support.

Capability Gallery Contract

The capability gallery should not replace this section. Its role is different:

• it shows versioned mesh bundles and static figures;

• it exposes node, cell, edge, river, geology, and quality summaries;

• it provides reproducible commands and artifact paths;

• it documents what is already available, not the full method rationale.

The scientific mesh pages should explain how to interpret those gallery artifacts. The gallery pages should link back here when the reader needs the conceptual background.

Recommended Reading Order

For a new mesh-related question, read in this order:

1. Mesh And Discretization Strategies

2. Regular And Irregular Mesh Cell Budgets

3. Field-To-Cell Parameter Transfer

4. Mesh Quality And Acceptance Criteria

5. Mesh Gallery

6. One concrete case page, such as Mesh Quality Diagnostics (10 km2)

• Mesh and discretization strategies
  • Purpose
  • Common Discretization Chain
  • Conceptual Diagram
  • Result Illustration: Nancon Basin Support To Solver Result
  • What Nancon Shows, And What It Does Not
  • Runtime Mesh Contrast
  • Planar Support Families
  • Solver-Side Interpretations
  • Why Mesh Choice Matters Scientifically
  • Field-To-Cell Transfer
  • Why Dedicated Mesh Execution Exists
  • What To Check Before Solving
  • Practical Choice Guide
  • Related Pages
• Regular and irregular mesh cell budgets
  • Purpose
  • Same Cell Count, Different Topology
  • River-Only Constraint Figures
  • Geology And Rivers Constraint Figures
  • Versioned HydroModPy Mesh Budgets
  • Controlled Size-Floor Pair
  • Existing Gallery Pages To Reuse
  • What Is Still Missing For The Ideal Comparison
  • Current Generation Scripts
• Field-to-cell parameter transfer
  • Purpose
  • The Three Objects To Keep Separate
  • Common Mapping Chain
  • Homogeneous Versus Heterogeneous Parameters
  • Why This Is Not Exact Geometry Preservation
  • Structured Grid Path
  • Irregular MODFLOW 6 Path
  • Boussinesq Path
  • Depth Dependence
  • What Should Be Reported
  • Where To Go Deeper
  • Related Scientific Pages
  • Current Limitation
• Mesh quality and acceptance criteria
  • Purpose
  • Two Different QA Families
  • Minimum Checks Before Solving
  • Structured Grid Acceptance Questions
  • Irregular Mesh Acceptance Questions
  • Why Comparison Workflows Need Extra Care
  • Evidence Already Available In The Repository
  • What Should Be Reported
  • Related Pages
  • Current Limitation
• Vertical representation and storage assumptions
  • Purpose
  • Conceptual Cross-Section
  • Common Vertical Construction In The MODFLOW Family Path
  • Layer Proportions
  • What This Means Physically
  • Validity Checks
  • Storage Interpretation
  • Implications For Property Mapping
  • Relation To Initial Conditions
  • Current Limits
  • Relation To The Boussinesq Backend
  • Related Pages
• XT3D on irregular DISV meshes
  • Purpose
  • Current HydroModPy Policy
  • Why This Policy Exists
  • Trade-Offs
  • When To Keep It Enabled
  • When To Disable It Deliberately
  • How It Should Be Reported
  • Same Case, Structured Versus Irregular Triangles
  • Current Evidence Anchors
  • Related Pages

Related Gallery Pages

• Mesh Gallery

• Mesh Sample Bundle

• Mesh Quality Diagnostics (10 km2)

• Constraint Balance Across Scales

• Resolution Sensitivity Across Scales

Future Evolution

The current mesh evidence in the documentation is built from static figures generated alongside each gallery case. Browsing a real triangulation, an irregular DISV mesh, or a mesh-quality field today still requires opening the underlying mesh bundle in a desktop tool such as ParaView or Gmsh.

A future iteration could embed an in-browser 3D mesh viewer based on vtk-js directly in the gallery and architecture pages. vtk-js can stream a small VTU/VTP serialized mesh into a WebGL canvas with picking, slicing, and field-by-cell coloring without any server-side rendering.

This is a deliberate non-goal for the current refactor: the dependency surface (WebGL bundle, larger asset payloads, longer page-load budget) is out of scope until analytics confirm that the static figures are no longer enough. If a prototype lands later, it should remain opt-in (loaded only on the case pages that explicitly opt into the viewer) and ship a small, redacted mesh per case rather than the full bundle.