Articles • potentiomap

Start here

Five-minute quick start: A copy-paste workflow from released synthetic groundwater observations to a surface, contours, quicklook, and hydraulic-gradient arrows.

Prepare observations

Preparing groundwater observations: Direct head measurements and every released depth-to-water pathway, with data-frame, sf, and terra inputs.
Input formats and coordinate systems: Prepare data-frame, sf, and terra point inputs; distinguish CRS assignment from transformation; and diagnose common spatial mistakes.

Build surfaces

Comparing TPS, IDW, ordinary kriging, and universal kriging: Four released interpolation methods on one projected grid, with common scales, contours, and difference rasters.
Interpolation parameters and common grid geometry: See how resolution, padding, masks, IDW controls, TPS smoothing, kriging lags, and templates affect the modeled product.
Custom interpolation functions: Implement and validate the released custom-method signature without using package internals.
Contours and surface smoothing: Regular and explicit contours plus mean, median, one-pass, and multi-pass focal smoothing with scientific cautions.
Comparing contour-support thresholds: See how tighter, broader, and network-relative distance criteria divide the same modeled contours into supported, approximate, and unsupported sections.

Create products

Hydraulic-gradient arrows: Inspect gradient rasters, arrow lines, bases and tips, cartographic controls, filtering, log options, and a programmatic downgradient check.
Exporting GIS-ready products: Write, inventory, read back, and validate GeoTIFF, contour, quicklook, arrow, tip, and base products.
Repeated monitoring events: Use ordinary R iteration around potentiomap for three synthetic monitoring rounds on one common grid.

Applied examples

Public USGS groundwater-monitoring example: A reproducible TPS, contour, and inferred-gradient workflow for 36 Stanley Shale wells near Hot Springs, Arkansas, using official USGS data.
Output gallery: Actual potentiomap outputs from synthetic data and the attributed USGS public-data example.

0.2.0 guides

Getting started with potentiomap: A concise end-to-end introduction to observations, interpolation, contours, diagnostics, and support.
Interpolation diagnostics and prediction support: Review retained interpolation diagnostics and classify where a mapped prediction has limited spatial support.
Contours and hydraulic-gradient arrows: Create contours and checked downgradient display symbols while keeping support and interpretation limits explicit.
Units, vertical references, and grouped groundwater observations: Keep units, vertical datums, screened intervals, and water-bearing groups explicit before interpolation.
Preparing and checking groundwater observations: Prepare measured heads or depth-to-water records and inspect observation-level QA findings before mapping.
Validating and comparing interpolation methods: Define prediction tasks, compare validation designs, and select methods without overstating map-wide accuracy.
Variograms, anisotropy, and external drift: Inspect spatial dependence, directional structure, trend assumptions, and covariate-supported kriging models.
Prediction support, surface uncertainty, and contour uncertainty: Distinguish geometric support, model-conditional uncertainty, method spread, and pointwise contour bands.
Temporal head change and vertical hydraulic gradients: Compare monitoring events and calculate explicitly signed vertical gradients from compatible paired intervals.
Monitoring-network sensitivity and candidate locations: Assess leave-one-well influence, thinning consequences, and constrained candidate-location rankings.
Surface profiles, depth to water, and cross-sections: Derive depth surfaces and extract transect profiles and plot-ready cross-sections from modeled heads.
Exporting GIS products and technical reports: Export auditable rasters, vectors, open GIS styles, manifests, and concise technical reports.

Expanded analysis articles

Observation QA, event selection, and screen grouping: Check records, select coherent monitoring events, and separate screened intervals before surface modeling.
Leave-one-out, spatial-block, and independent validation: Match validation design to the intended prediction task and interpret the resulting errors within that scope.
Nested tuning without information leakage: Tune interpolation settings inside training folds so held-out predictions remain an honest comparison.
Comparing TPS, IDW, OK, and UK: Compare deterministic and geostatistical surfaces, diagnostics, assumptions, and validation results.
Ensembles versus method disagreement: Separate an ensemble prediction from a descriptive map of spread among interpolation methods.
Conditional simulation and model-conditional uncertainty: Summarize conditional simulations while keeping model assumptions and uncertainty scope visible.
Pointwise contour uncertainty versus approximate contours: Compare pointwise contour-position bands with support-classified approximate contour segments.
Comparing monitoring events: Select comparable monitoring rounds and map modeled head differences without interpreting them as storage change.
Vertical-gradient sign conventions: Calculate paired vertical gradients with an explicit numerator, denominator, and sign interpretation.
Well influence and network thinning: Quantify leave-one-well surface change and evaluate proposed monitoring-network reductions.
Candidate monitoring locations and design constraints: Rank feasible candidate wells using explicit objectives, exclusions, costs, and user priorities.
Grid, boundary, and parameter sensitivity: Compare surfaces across grid resolution, domain, and interpolation settings without treating spread as uncertainty.
Directional variograms and anisotropy: Inspect directional semivariance and evaluate whether anisotropy is supported by the observation network.
Universal kriging with hydrogeologic covariates: Use aligned covariates as an explicit external drift and review the resulting trend assumptions.
Interpolation within user-defined hydrogeologic regions: Fit separate surfaces inside defensible hydrogeologic regions and retain region-level diagnostics.
Depth to water and depth to a potentiometric surface: Subtract modeled head from a compatible land-surface raster and review datum and unit requirements.
Profiles and cross-sections: Sample a modeled surface along a transect and assemble plot-ready hydrogeologic section data.
QGIS and SLD style exports: Create portable open GIS style files for exported potentiometric rasters and contour products.
Building a technical analysis report: Record methods, conditions, QA, support, and exported artifacts in a reproducible technical report.
Complete synthetic aquifer example: A complete fixed-seed 0.2.0 workflow from observation QA through validation, uncertainty, network review, profiles, and export.

Scientific guidance

Interpretation, assumptions, and limitations: A hydrogeologic review checklist for observations, interpolation, smoothing, boundaries, and inferred hydraulic-gradient products.
Troubleshooting and common errors: Observed potentiomap errors and practical checks for CRS, missing values, duplicates, kriging, grids, arrows, contours, and exports.
Citing potentiomap: CRAN availability, DOI, package version, license, citation, and BibTeX.