Glossary — Terminology for IGW-NET and Groundwater Modeling

28.1 How to Use This Glossary

Entries are alphabetical, grouped into five letter ranges for faster navigation. Each entry has the form:

Term — Concise definition. See Ch. X §X.X for in-depth coverage.

Related terms cross-reference each other (e.g., "Sequential Gaussian Simulation" links to "spatially correlated random field"). Acronyms appear both under their expanded name and as an "see also" entry at the acronym letter (e.g., SGS entry redirects to Sequential Gaussian Simulation).

When a term is platform-specific (IGW-NET invented or uses a term differently from standard groundwater modeling), it is marked accordingly. When a term has a parser-backed exact form in the model file, the parser's field name is noted in monospace.

28.2 A – D

Advection

Transport of dissolved solutes by bulk groundwater flow. The dominant transport process in most contaminant problems. See Ch. 12.

Anisotropy

Directional dependence of aquifer properties. Horizontal-to-vertical K anisotropy ratios of 10:1 or more are typical in layered sedimentary systems.

AQ / MAQ / PCM / CM

The four-class material typology used by T-PROGS in IGW-NET. AQ = aquifer (high K); MAQ = marginal aquifer (moderate K); PCM = partial confining material (low K); CM = confining material (very low K). See Ch. 17 §17.1.3.

Aquifer

A geological formation capable of transmitting significant quantities of groundwater under normal hydraulic gradients. Confined aquifers have an impermeable overlying unit; unconfined aquifers have a water table as their upper surface.

Automatic Calibration

IGW-NET workflow for parameter estimation using UCODE. Enabled via a Calib checkbox on multipliers. The solver runs UCODE wrapped around flow iterations to find multipliers that minimize residuals against observed data. See Ch. 18 §18.7.

ASTER

Advanced Spaceborne Thermal Emission and Reflection Radiometer. A satellite DEM product providing global 90m resolution topographic data. Used by IGW-NET Data Center as a DEM source for global coverage. See Ch. 23 §23.1.

Backward particle tracking

Running particle paths backward in time from a specified location (e.g., a water-supply well) to trace where the water originated. Used for capture zone delineation. See Ch. 11 §11.4.

Base flow separation

Decomposition of streamflow into groundwater-sourced baseflow and surface-runoff components. Provides calibration targets for the groundwater discharge portion of the flow.

Bottom elevation

The elevation of the aquifer's lower boundary. Combined with top elevation (DEM), defines aquifer thickness. Parser field: .bottom_elevation. Data Center datasets available via BOTE_DATASETS. See Ch. 5, Ch. 23 §23.3.

Boundary condition

Specification of head, flux, or a relationship between them along the model boundary or at internal features. Types include prescribed head (CHD), prescribed flux, general head (GHB), river (RIV), drain (DRN). See Ch. 4, Ch. 14.

Breakthrough curve

Time series of solute concentration at a specified location. In transport modeling, used to analyze plume arrival times and peak concentrations at monitoring wells. See Ch. 12.

Brute-force NX trap

The anti-pattern of pushing NX (grid columns) to very high values to resolve fine detail without using IGW-NET's nested modeling (Ch. 13). Produces slow simulations without improving accuracy at feature scales where nested modeling would do better. See Ch. 7.

Calibration

Adjusting model parameters to match observed field data (typically static water levels, sometimes fluxes or concentrations). In IGW-NET, done manually through Calib checkboxes with visual charts, or automatically via UCODE. See Ch. 18.

Capture zone

The region of aquifer where water eventually reaches a specified discharge point (a well, stream, or other sink). Deterministic capture zones are hard polygons; probabilistic capture zones (from stochastic analysis) are probability fields. See Ch. 11, Ch. 19.

Cell drying

In unconfined simulations, cells where the head drops below the cell bottom become "dry" — no flow can pass. Can cause numerical problems in classical MODFLOW; MODFLOW-NWT handles wetting/drying robustly. See Ch. 20 §20.3.3.

Compute-render-discard architecture

IGW-NET's design pattern for both transient and stochastic visualization. Each time step (or realization) is computed, incorporated into the running display, and discarded — so memory cost is independent of the number of steps or realizations. See Ch. 19 §19.6.

Conceptual model

The spatial representation of the real-world aquifer in map-readable form — domain boundary, zones, features, attributes. In IGW-NET, this is what users draw; the "xyz" side of the xyz-vs-ijk distinction. See Ch. 1 §1.5.

Convergence tolerance

The maximum head change between successive solver iterations below which the iteration is considered converged. Default 0.0001 m. Position 2 of gmSolverID. See Ch. 24 §24.3.2.

Correlation scale (Lambda)

Spatial distance over which random-field values are correlated. Values closer than Lambda are correlated; values farther apart are effectively independent. Parser fields LambdaX, LambdaY, LambdaZ for the three directions. See Ch. 19 §19.3.

Coupled lake-aquifer model

Two-way coupled simulation of lake stage and surrounding groundwater flow. The lake water balance (inflows, outflows, precipitation, evaporation, seepage) and aquifer flow are solved together. Level-3 surface-water representation. See Ch. 15.

Damping factor

Scaling factor applied to head changes between iterations. Damping = 1.0 means full step; less damping slows convergence but stabilizes oscillating iterations. Position 3 of gmSolverID, default 1.0. See Ch. 24 §24.3.3.

Data Center

IGW-NET's integrated repository of curated spatial datasets — DEMs, K datasets, recharge datasets, borehole databases. Provides the Global Base Model defaults. See Ch. 23.

DataNET

The fifth MAGNET4WATER platform — a data services platform for transferring specialized datasets into modeling workflows. IGW-NET detects DataNET-sourced raster imports via filename prefix (dataNetWCS or dataNetmWCS). See Ch. 23 §23.7.

DEM

Digital Elevation Model. Raster of land-surface elevations. Used in IGW-NET for top elevation of the aquifer and for DEM-as-drain boundary condition. Multiple resolutions available via Data Center. See Ch. 23 §23.1.

DEM-as-drain

IGW-NET's Level-1 surface-water representation. The DEM is used directly as a drain elevation — cells where simulated head exceeds DEM discharge to the surface. Captures regional seepage, springs, wetlands without explicit stream placement. Parser field: DrnL. See Ch. 14 §14.2.

DIS

MODFLOW's structured-grid discretization package (rectangular row-column-layer grid). Contrasts with DISV (unstructured by vertices) and DISU (fully unstructured). See Ch. 20 §20.4.

DISV

MODFLOW 6's Discretization by Vertices — unstructured grid where cells can have variable shapes. Required for IGW-NET's refined-grid nested-modeling pattern. Triggers automatic activation of MODFLOW 6. See Ch. 13, Ch. 20 §20.3.1.

Dispersion

Spreading of dissolved solutes beyond what advection alone would produce. Mechanical dispersion (from velocity variations at the pore scale) plus molecular diffusion. See Ch. 12.

DomainAttr

IGW-NET toolbar button that opens the Default Model Parameters dialog — the central hub for aquifer attributes, grid resolution, layers, boundary conditions, simulation settings. See Ch. 2, Ch. 21 §21.2.1.

DrnL

Parser field for surface-drainage (DEM-as-drain) configuration. Current format: 4 comma-separated values (leakance, Sy, Ss, porosity). Legacy format was a single value; parser auto-upgrades 1→4 values. See Ch. 5 §5.8, Ch. 22 §22.4.2.

28.3 E – K

Ensemble simulation

Monte Carlo stochastic simulation running many realizations and aggregating into ensemble statistics (mean, variance, probability distributions). IGW-NET's recursive-statistics architecture makes this tractable at high realization counts. See Ch. 19 §19.5.

Evapotranspiration (EVT)

Water loss from groundwater and soil to the atmosphere via plant transpiration and surface evaporation. MODFLOW package EVT. See Ch. 20 §20.4.1.

Extended Tóth flow

A generalization of Tóth's regional-flow circulation to include the effect of structural geology — horizontal layers of contrasting K that redirect the natural recharge-to-discharge pathways. Where classical Tóth theory assumes a uniform aquifer beneath a prescribed water table and produces smoothly-nested local/intermediate/regional cells, extended Tóth systems include high-K basal units that channel deep flow, internal aquitards that deflect shallow flow, and heterogeneous stratigraphy that produces the complex vertical circulation patterns characteristic of real glaciated aquifers. IGW-NET reproduces these patterns directly in synthetic mode with a prescribed water table and one or more zones of contrasting K. See Ch. 1 §1.3.2, Ch. 10 §10.2.

Explicit stream

A stream represented explicitly as a polyline with stage, leakance, and bottom elevation. IGW-NET's Level-2 surface-water representation. Parser: .polylines. See Ch. 14.

Field (model file field)

Named data element in an IGW-NET model file — for example, gmSolverID, DrnL, ZoneValues. The parser decodes fields into structured representations. Platform Reference catalogs every field. See Ch. 22.

Five-platform ecosystem

MAGNET4WATER's architecture of five coordinated platforms: IGW-NET (groundwater), SwaNET (watershed), StormNET (stormwater), ConduitNET (piped flow), DataNET (data services). See Ch. 1, Ch. 26 §26.5.3.

Flow solver

The numerical algorithm solving the groundwater flow equations. IGW-NET's default is Hybrid SOR (ID 15); alternatives include PCG, SIP, SOR, and MODFLOW 6 IMS. Position 0 of gmSolverID. See Ch. 24 §24.2.

Four-tier tool organization

IGW-NET's interface structure: Conceptual Model Tools (build), Simulation Tools (run), Analysis Tools (interpret), Other Tools (manage workspace). Every toolbar button belongs to one of these four tiers. See Ch. 2, Ch. 21.

Georeferenced mode

One of IGW-NET's two canonical operating modes. The model domain is tied to real-world geographic coordinates on a map, and the Global Base Model automatically loads real DEMs, K rasters, recharge, streams, and lakes for that location from the Data Center. Used whenever the modeling question concerns a specific site, watershed, or region. Contrasts with synthetic mode, where the canvas is blank and the domain has no real-world reference. Both modes use the same solver, features, and visualization — what differs is only whether real-world data is attached. See Ch. 1 §1.3.2, Ch. 4.

GLHYMPS 2.0

GLobal HYdrogeology MaPS version 2.0 — a global surficial hydraulic conductivity dataset. Available as K dataset ID 3 in IGW-NET's K_DATASETS. See Ch. 23 §23.2.

Global Base Model

IGW-NET's concept of a default, data-backed model state. Defining a domain automatically loads DEM, regional K, recharge, streams, lakes from the Data Center — you start with a working model and refine rather than building from scratch. See Ch. 1 §1.4, Ch. 5 §5.1.2.

Global Database

The set of curated datasets IGW-NET ships with — DEMs, K rasters, recharge datasets, streams, lakes. Distinguished from DataNET-transferred data. Parser flag: isGlobal=0. See Ch. 23 §23.7.

Global Model Network

IGW-NET's repository of published models. Users can publish their models for discovery and reuse; the network supports collaborative modeling workflows. See Ch. 29.

gmSolverID

The primary solver configuration field in IGW-NET model files. Comma-separated positional string with different semantics at different positions (0 = solver ID, 1 = max iterations, 2 = tolerance, 3 = damping, 17 = density flow, 18 = auto-cal, 19 = Monte Carlo count, 20 = MFsolverStr). See Ch. 24 §24.1.

Head

Groundwater hydraulic head — the energy per unit weight of water, expressed as an elevation. The primary unknown in groundwater flow simulation. See Ch. 7, Ch. 8.

Head variance

Ensemble variance of simulated head at each cell across Monte Carlo realizations. Head variance maps visualize spatial uncertainty — low near boundaries, higher in the interior. See Ch. 19 §19.7.2.

Heterogeneity

Spatial variation in aquifer properties (most commonly K). Real aquifers are heterogeneous at every scale; IGW-NET supports T-PROGS (categorical) and SGS (continuous) random fields to represent realistic heterogeneity. See Ch. 17, Ch. 19.

Hierarchy (modeling hierarchy)

IGW-NET's foundational design principle — a regional parent model passes boundary conditions to refined submodels, supporting multi-scale modeling without brute-force fine grids. The 10-cell buffer rule (Ch. 13 §13.3) governs submodel placement. See Ch. 13.

Hybrid SOR

IGW-NET's default native flow solver. Tuned SOR-based solver combining successive over-relaxation with adaptive strategies. Solver ID 15 in SOLVER_NAMES. See Ch. 24 §24.2.1.

Hydraulic conductivity (K)

Aquifer property quantifying how easily water flows through it, in units of length/time. Typically varies over many orders of magnitude across materials; represented heterogeneously via T-PROGS or SGS random fields. See Ch. 5, Ch. 17, Ch. 23 §23.2.

Hydrostratigraphy

The vertical layering of an aquifer system based on hydrologic behavior rather than strict geological classification. In IGW-NET, hydrostratigraphic layers are represented as distinct vertical layers with their own K, storage, etc. See Ch. 10.

IMS (Iterative Model Solution)

MODFLOW 6's general-purpose iterative solver package. Activates automatically when MODFLOW 6 is invoked (typically for unstructured-grid nested models with DISV). See Ch. 20 §20.3.1, Ch. 24 §24.2.2.

Initial conditions

Starting values of the simulation state — head fields at the start of a transient run, concentrations at the start of transport. Parser field: .starting_head. See Ch. 9.

isGlobal flag

Parser flag distinguishing data sources: isGlobal=0 means Global Database (IGW-NET bundled); isGlobal=1 means user data or DataNET-transferred. See Ch. 23 §23.7.

K

See Hydraulic conductivity.

K-recharge non-uniqueness

The fundamental inverse-problem fact that K and recharge are confounded by head-only calibration — multiplying both by the same factor gives identical heads. Resolved by multi-target calibration (adding flux observations). See Ch. 18 §18.8.

K_DATASETS

Parser catalog of hydraulic conductivity datasets available from the IGW-NET Data Center: 1 Michigan Glacial-Entire Drift, 2 Michigan Glacial-Screen to 1st Confining, 3 Global Surficial K (GLHYMPS 2.0), 4 Michigan Bedrock T, 5 USA Glacial K (USGS), 6 USA/Canada T. See Ch. 23 §23.2.

Kd (partitioning coefficient)

Solute retardation factor controlling how contaminants partition between water and solid phases. Higher Kd = more retardation, slower plume migration. Can be specified as a random field in IGW-NET. See Ch. 12.

28.4 L – P

Lake coupling

See Coupled lake-aquifer model.

Layer vs. sublayer

IGW-NET distinguishes true hydrostratigraphic layers (user-defined, with distinct properties) from sublayers (numerical subdivisions used by the solver for vertical resolution, especially in water-table-as-top scheme). See Ch. 10.

Leakance

Streambed or drain conductance per unit area. Determines the rate at which head gradient across a boundary drives flow exchange. See Ch. 14.

Linearize-within / iterate-between

IGW-NET's scheme for water-table-as-top problems. Within a single time step, treat the saturated thickness as fixed (linear problem); between time steps, update saturated thickness based on the new head (nonlinear outer iteration). See Ch. 7 §7.4.3.

Macrodispersion

Excess spreading of the ensemble-mean plume in Monte Carlo transport, beyond what local dispersion alone would produce. Arises from variability in flow paths across realizations. Captured by stochastic modeling; not capturable by uniform-K deterministic models. See Ch. 19 §19.7.4.

MAQ

See AQ/MAQ/PCM/CM.

Manning's n

Roughness coefficient for open-channel flow. In IGW-NET, stored for every polyline feature but active only when the Watershed Solver is on. Platform gotcha: stored universally, conditionally used. See Ch. 16 §16.4, Ch. 22 §22.4.3.

Mass balance

Verification that water in = water out (steady-state) or water in − water out = storage change (transient). IGW-NET's mass balance chart displays this in real time. Failure to close usually indicates convergence problems. See Ch. 8, Ch. 25 §25.4.

Metadata gap

The fact that Data Center dataset names are not stored in IGW-NET model files; only numeric IDs are stored, which the parser's catalogs resolve to human-readable names. Motivates parser-backed documentation. See Ch. 23 §23.7.3.

MFsolverStr

JSON-encoded MODFLOW variant configuration stored at position 20 of gmSolverID. "!" replaces internal commas to avoid collision with the outer comma-separated format. Typical values: {"version":["mf6"]} for MODFLOW 6. See Ch. 20 §20.1.3, Ch. 24 §24.5.4.

MODFLOW

USGS's widely-used groundwater flow modeling engine family. IGW-NET integrates MODFLOW 6 (modern unified; supports DISV), MODFLOW 2005 (widely-used legacy), MODFLOW-NWT (Newton-Raphson for unconfined), and MODFLOW-USG (legacy unstructured). See Ch. 20.

Monitoring Probability

Checkbox on wells in IGW-NET. When checked, the well's head (and concentration, if transport active) is recorded every realization in Monte Carlo runs, building up a probability distribution at that location. See Ch. 19 §19.7.3.

Monitoring well

A well where head (and sometimes concentration) is measured for calibration or observation purposes, typically without pumping. In IGW-NET, flagged via isMonitoringWell. See Ch. 6, Ch. 18.

Monte Carlo simulation

See Stochastic modeling. IGW-NET's Monte Carlo workflow uses the recursive-statistics architecture to compute ensemble statistics without storing the ensemble. See Ch. 19 §19.5-§19.6.

MT3DMS

USGS/US Army solute transport engine. IGW-NET activates MT3DMS automatically when transport is enabled — the flow engine (native IGW or MODFLOW) and MT3DMS work together for coupled flow-and-transport. See Ch. 12, Ch. 20 §20.4.2.

Multiplying factors

Calibration multipliers applied to K and recharge fields. Parser field: gmZNfactor with format "Kflag,Rflag|Kfactor,Rfactor". Calib checkbox activates each for Automatic Calibration. See Ch. 18 §18.2.

Native IGW solver

IGW-NET's own flow solver, used by default for structured-grid problems. Tightly integrated with the real-time interactive framework. Contrasts with MODFLOW engines which are invoked when their capabilities are needed. See Ch. 20 §20.2.

Nested model

A submodel receiving boundary conditions from a parent model, typically for refined local detail within a regional context. IGW-NET's hierarchy principle treats nesting as a first-class pattern. See Ch. 13.

Non-uniqueness

Multiple parameter sets producing equivalent fit to observations. The K-recharge case is the most famous example; multi-target calibration resolves it. See Ch. 18 §18.8.

Numerical model

The discrete grid-based representation used by the solver — cells indexed by (i, j, k). The "ijk" side of the xyz-vs-ijk distinction. See Ch. 1 §1.5.

NX

Number of grid columns. One of the key grid-resolution parameters. Pushing NX to very high values is the "brute-force NX trap" that nested modeling is designed to avoid. See Ch. 7.

Observatory

The MAGNET4WATER Observatory — the integrated information layer covering approximately 43 countries and approximately 143 cities with datasets, models, and workflow examples. See Ch. 26 §26.5.1.

Particle tracking

Advective-only tracing of water-parcel trajectories through the flow field. Forward tracking traces where water is going; backward tracking traces where water came from (used for capture zones). See Ch. 11.

Parser

The Python code (igw_net_parser_v3.py) that reads IGW-NET model files and produces a structured representation. Authoritative source for what a model file actually contains. See Ch. 22 §22.1.

Parser-backed truth

The principle that documentation should derive claims from what the parser extracts from real model files, not from interface appearance or developer memory. Grounds the Platform Reference and this manual's technical specifics. See Ch. 22 §22.1.

PCG

Preconditioned Conjugate Gradient solver. Solver ID 1 in IGW-NET's SOLVER_NAMES. Robust for well-conditioned linear problems. See Ch. 24 §24.2.1.

PCM

See AQ/MAQ/PCM/CM.

Platform Reference

The parser-backed catalog of IGW-NET model file fields — the second layer of the three-layer reference system. Documents every field with format, valid values, gotchas, and cross-references. See Ch. 22.

Polyline

Line feature in IGW-NET representing streams, drains, prescribed-head boundaries. Parser field: .polylines. See Ch. 6, Ch. 14.

Porosity (effective)

Fraction of aquifer volume available for water flow. Used in the relationship velocity = Darcy flux / effective porosity for transport and particle tracking. See Ch. 5.

Prescribed water table

A vertical-profile modeling approach in which the user draws the water-table shape and the solver is constrained to satisfy head = y (head equals elevation) along that line — physically correct because the water table is by definition where pressure equals zero. Used in synthetic mode to reproduce Tóth regional-flow patterns and extended Tóth behavior with structural geology. Limitation: the approach does not constrain fluxes to rainfall — a drawn water table and an assigned K field that are not mutually compatible will produce simulated fluxes that violate water balance. Useful for flow-geometry visualization; for flux-realistic vertical profiles, use the 3D slice model instead. See Ch. 10 §10.2.

Probabilistic capture zone

Capture zone expressed as a probability field rather than a hard boundary. Each location has probability P of being captured by the well; P varies continuously from 1 (near the well) to 0 (far away). Supports risk-based wellhead protection. See Ch. 19 §19.7.5.

28.5 Q – T

Quick Tutorials

IGW-NET's collection of 28 step-by-step walkthroughs, each covering one specific workflow. Accessible at /magnet/docs/igwnet/quick-tutorials/. See Ch. 26 §26.3.1.

Random field

Spatial field with values drawn from a probability distribution. IGW-NET supports random fields for K, porosity, density, recharge, retardation, top/bottom elevations, and thickness. Configured per zone per variable. See Ch. 19 §19.3.

Realization

One specific random-field instance from a stochastic model setup. A single realization shows what heterogeneity does; many realizations (an ensemble) quantify uncertainty. See Ch. 19.

Realtime help

IGW-NET's collection of approximately 179 operational reference pages keyed to specific dialogs, buttons, and features. First-line resource for operational questions. See Ch. 26 §26.2.2.

Recharge

Water entering the aquifer from the surface (precipitation minus evapotranspiration minus runoff). Units matter — the 10×-rainfall pitfall is the most common error. Parser field: .recharge. See Ch. 5, Ch. 23 §23.4.

Recursive statistics

Running statistics (mean, variance, probability) updated per realization using online algorithms like Welford's. Lets IGW-NET compute ensemble statistics without storing the ensemble — the memory breakthrough that makes high-resolution stochastic modeling tractable. See Ch. 19 §19.6.

Refined grid

Grid with locally finer cells around specific features (wells, zones, polylines). In IGW-NET, refined grids require MODFLOW 6 DISV discretization; the native IGW solver handles only structured grids. Parser property: .refined_grid. See Ch. 13, Ch. 20 §20.3.1.

Regional zonation

Approach to heterogeneous K where a domain is divided into zones, each with its own set of AQ/MAQ/PCM/CM parameters. N zones × 4 parameters = 4N total K descriptors. See Ch. 17 §17.5.

Retardation factor (R)

Solute transport slowdown from partitioning to solids. R = 1 for conservative (non-sorbing) tracers; R > 1 for sorbing solutes. See Ch. 12.

SEAWAT

USGS variable-density flow and transport engine. Couples MODFLOW flow with MT3DMS transport plus density coupling (VDF package). Activated by IGW-NET when the density-flow flag is set (gmSolverID position 17). See Ch. 20 §20.4.3.

Seed

Integer value controlling the pseudo-random number generator in random-field generation. Same seed = same realization; different seed = different realization with the same statistical properties. See Ch. 19 §19.3.2.

Sequential Gaussian Simulation (SGS)

Geostatistical method for generating continuous spatially-correlated random fields with specified mean, variance, and correlation scale. IGW-NET's method for continuous-variable heterogeneity; complements T-PROGS for categorical. See Ch. 19 §19.3.

SGS

See Sequential Gaussian Simulation.

SIP

Strongly Implicit Procedure. Older iterative solver; solver ID 2 in SOLVER_NAMES. Largely superseded by PCG. See Ch. 24 §24.2.1.

Solver

The iterative algorithm that solves the matrix equations for simulated head. IGW-NET supports native Hybrid SOR (default), PCG, SIP, SOR, and MODFLOW 6 IMS. See Ch. 24.

SOR

Successive Over-Relaxation solver. Solver ID 3 in SOLVER_NAMES. Hybrid SOR (ID 15) is IGW-NET's tuned variant and the platform default. See Ch. 24 §24.2.1.

Spatially correlated random field

Random field where neighboring values are correlated through a specified structure (variogram, correlation length). Both SGS (continuous) and T-PROGS (categorical) produce spatially correlated random fields. See Ch. 19 §19.2.

Static Water Level (SWL)

Groundwater elevation measured at a well under non-pumping conditions. The primary calibration target in most IGW-NET workflows. Collected in Wellogic (Michigan) and equivalent regional databases. See Ch. 18, Ch. 23 §23.5.

Stochastic modeling

Modeling under explicit treatment of uncertainty. In IGW-NET, includes both single-realization analysis (what does heterogeneity do?) and ensemble Monte Carlo (how uncertain is the answer?). See Ch. 19.

Stress period

In transient simulations, a time interval over which stresses (boundary conditions, pumping rates, recharge) are held constant. A simulation typically has multiple stress periods, each subdivided into time steps. See Ch. 9.

Structure-first principle

IGW-NET's foundational principle: the conceptual structure of the model (BCs, geometry, layering, features) is a prerequisite, not an output, of calibration or stochastic analysis. A Monte Carlo ensemble on a structurally-flawed model gives confident confidence intervals on a wrong answer. See Ch. 18 §18.1.3.

Structured grid

Rectangular cell grid with uniform row and column spacing. IGW-NET's native solver handles structured grids; unstructured grids (refined cells) require MODFLOW 6 DISV. See Ch. 20 §20.2.1.

Submodel

A refined model nested inside a parent regional model, receiving BCs from the parent. The 10-cell buffer rule governs submodel boundary placement. See Ch. 13.

SwaNET handshake

Coordinate-system interoperability between IGW-NET and SwaNET (the watershed platform). Requires UTM projection on both sides — the UTM-Only checkbox enforces this. See Ch. 16 §16.6.3.

Synthetic mode

One of IGW-NET's two canonical operating modes. An empty canvas with no real-world geographic reference — you draw a concept (domain, boundary conditions, wells, zones, heterogeneity, sources) and immediately see the physics respond. How to enter: from the main interface, click Utilities → Go to Synthetic Case Area. Used by students exploring concepts, faculty teaching groundwater principles as a "blackboard with physics," consultants scoping problems before committing to site-specific work, and researchers testing hypotheses about heterogeneity, transport, or coupled dynamics. Uses the same solver, same features, and same visualization as georeferenced mode — what differs is only whether the domain is tied to real coordinates, and that Data Center auto-defaults are absent (you assign properties directly via DomainAttr). Particularly powerful for visualizing macrodispersion from heterogeneity, Tóth regional circulation from sloping water tables, and capture-zone behavior under pumping. Canonical tutorial: Tutorial 9 (Synthetic Heterogeneous Aquifer Model). Related tutorials: 3 (Particle Tracking), 5 (Contaminant Transport), 6 (Vertical Flow Dynamics), 12 (Profile Modeling), 15 (Stochastic Flow), 17 (MC Transport), 20 (Theis Solution). Contrasts with georeferenced mode, where the platform loads real DEMs, K, recharge, streams, and lakes from the Global Base Model for a specific location. See Ch. 1 §1.3.2.

Three-layer reference system

IGW-NET's documentation architecture: Users' Manual (concepts) → Platform Reference (field-by-field) → parser code (executable truth). Each layer serves different questions. See Ch. 22 §22.5.4.

Time step

In transient simulations, the numerical increment within a stress period. Smaller time steps capture faster physics; larger steps improve runtime. See Ch. 9.

Tolerance (convergence)

See Convergence tolerance.

Top elevation (TopE)

Elevation of the aquifer's top surface. Usually the DEM. In Level-2 stream representation, used as the "DEM-as-stage" convention where stream stage defaults to DEM elevation unless explicitly overridden. See Ch. 5, Ch. 14.

Transient simulation

Time-varying simulation with changing stresses, storage, or boundary conditions. Contrasts with steady-state simulations at a single time. See Ch. 9.

Transition probability

In T-PROGS geostatistics, the probability of transitioning from one material category to another over a specified distance. Estimated from borehole data and used to generate realistic layered random fields. See Ch. 17.

Transport (solute)

Simulation of dissolved substances moving through groundwater — advection, dispersion, reactions, retardation. Activates automatically in IGW-NET when concentration sources exist; handled by MT3DMS (or Magnet native). See Ch. 12.

T-PROGS

Transition PRObability Geostatistical Simulation. Method for generating categorical (material-class) heterogeneous K fields from borehole lithology data. Produces realistic layered structures. See Ch. 17.

28.6 U – Z, and Numeric

UCODE

USGS Universal inverse code for parameter estimation. Wrapped around the flow solve by IGW-NET to implement Automatic Calibration. Activates via gmSolverID position 18. See Ch. 18 §18.7.

Unconfined aquifer

Aquifer with a water table (free surface) as its upper boundary, allowing direct recharge from above. Produces the nonlinear flow problem solved via water-table-as-top / linearize-within / iterate-between scheme. See Ch. 7 §7.4.

Unstructured grid

Grid with cells of non-uniform shape or size, typically produced by local refinement around features. Requires MODFLOW 6 DISV discretization in IGW-NET. See Ch. 13, Ch. 20.

UTM handshake

See SwaNET handshake. The coordinate-system agreement that lets IGW-NET (groundwater) and SwaNET (watershed) interoperate reliably — both working in UTM projection. See Ch. 16 §16.6.3.

Variance

Statistical measure of spread. In IGW-NET stochastic modeling, head variance (Ch. 19 §19.7.2) and concentration variance (Ch. 19 §19.7.4) are key output maps. Computed via Welford's algorithm.

Water-table-as-top

IGW-NET's treatment of unconfined aquifers: the water table is the top of the computational domain (rather than the DEM). The saturated thickness becomes the head — making the flow problem nonlinear. Linearize-within / iterate-between scheme handles the nonlinearity. See Ch. 7 §7.4.3.

Watershed Solver

IGW-NET's in-platform subsystem for computing overland flow, infiltration, and runoff — producing a transient recharge field that the groundwater solver uses. Does not transfer to standard MODFLOW. See Ch. 16.

Welford's algorithm

Numerically stable online (single-pass) algorithm for computing running mean and variance without accumulating round-off errors. Used by IGW-NET's recursive ensemble statistics. See Ch. 19 §19.6.2.

Wellogic

Michigan's state water well database. The prototype regional borehole database that IGW-NET integrates for T-PROGS and SWL calibration. Similar integrations exist for approximately 10 US states, plus all Canadian provinces. See Ch. 17 §17.6, Ch. 23 §23.5.

WMTS (Web Map Tile Service)

OGC standard for serving map tiles. IGW-NET's Data Center delivers raster data via WMTS. Dataset IDs in the model file correspond to WMTS indices on the server side. See Ch. 23 §23.7.3.

xyz vs. ijk

IGW-NET's fundamental distinction between the conceptual model (spatial representation in xyz coordinates as drawn on the map) and the numerical model (discrete grid cells in ijk indices used by the solver). Users work in xyz; the platform handles the ijk translation. See Ch. 1 §1.5.

Zone

Polygon feature that overrides aquifer properties within its spatial extent. Each zone can have up to 15 property overrides including K source, random field configuration, lake coupling, T-PROGS borehole simulation. Parser field: .zones. See Ch. 5, Ch. 13, Ch. 15, Ch. 17, Ch. 19.

10-cell buffer rule

Nested-modeling guideline: submodel boundaries should be at least 10 cells away from any feature of interest inside the submodel. Prevents boundary effects from contaminating the results at the features you care about. See Ch. 13 §13.3.

10×-rainfall pitfall

The most common IGW-NET error: recharge unit mismatch producing recharge values 10× (or 100× or 365×) too high, leading to simulated water tables above the ground surface. See Ch. 14 §14.7.

3D slice model

A vertical-profile modeling approach in which a 3D model is built with the domain only one cell wide in plan view, but with full vertical resolution (one or more geological layers, each subdivided into many computational sublayers). The top is the DEM (or user-defined top); recharge is applied as an infiltration flux; the water table emerges as part of the solution rather than being prescribed. Contrasts with the prescribed-water-table approach: the 3D slice guarantees water-balance consistency between flux and rainfall by construction, at the cost of not being able to impose an arbitrary water-table shape. The preferred approach when flux realism matters — for water-budget analysis, recharge estimation, or site-specific interpretation. Particularly revealing in glaciated terrain where real topography and hydrostratigraphy produce complex vertical circulation that the simplified prescribed-water-table case cannot reproduce quantitatively. See Ch. 10 §10.2, Tutorial 6, Tutorial 12.