1Import and Visualize Point Data
Step 1 — Import the Chloride Shapefile
Go to 'Analysis Tools' → 'Data Explorer' → 'Data Analysis'. In the Scatter Data Spatial Analysis interface, select 'Import Shapefile' as the Data Source. Check 'Import' and click the hyperlink to upload 'ChlorideData_OttawaCounty.zip' containing the .shp file and its auxiliary files.
Step 2 — Filter and Assign Attributes
Click 'Apply' to launch the Internal Filters interface. Select 'WellID' as the filter field and check 'Select All' to import all records. Under Select Attribute Options, assign:
WellID field → WellID attribute
Zf field → Zf attribute (vertical sampling elevation — well screen depth)
Zt field → Zt attribute (same as Zf in this dataset)
Click 'Apply'. After processing, data markers appear on the map showing sampling locations with IDs across Ottawa County.
Step 3 — Color-Code Markers by Concentration
Select 'CL' from the Parameter dropdown (chloride concentrations in mg/L). Click 'Color Ramp for Marker'. Markers update to a blue-to-red color ramp — blue for low chloride, red for high. Marker size also scales with concentration. You can instantly see the spatial pattern: where is chloride elevated, and where is it background-level?
Step 4 — Customize Colors and Sizes
Click 'Edit Size/Color'. In the Point Data Color and Size interface, set custom thresholds: 20, 100, 150, 500, 1855 mg/L with colors from dark blue → royal blue → yellow → orange → red. Click 'Save', check 'Use Customized', and click 'update size/color'. The markers now use a scientifically meaningful color scheme — with thresholds that correspond to regulatory levels and health-based standards.
2Exploratory Statistical Analysis
Step 5 — Explore Data Distributions
Select 'PDF/CDF' from the Tools dropdown and click 'Apply'. The Exploratory Data Analysis interface opens showing a histogram with data statistics (mean, median, standard deviation, min, max, percentiles). Options include adjustable bin count, theoretical normal PDF overlay, and data filters by range or time period.
Click the 'Z-spatial' tab — this plots sampling elevation versus chloride concentration. A critical pattern emerges: the highest chloride concentrations occur at lower elevations (greater depths). This is not road salt behavior — road salt concentrations peak near the surface. The depth trend suggests a deeper source: natural brine upwelling from below, exacerbated by declining water levels and increased pumping.
What the Data Reveals Before Any Model Runs
The Z-spatial pattern is diagnostic: Chloride increasing with depth rules out surface contamination (road salt, agricultural runoff) as the primary mechanism. It points toward upward migration of deeper saline water — a natural process accelerated by over-pumping. This single plot fundamentally changes the conceptual model: instead of a surface source migrating downward, you're looking at a deep source migrating upward. The modeling strategy, boundary conditions, and remediation approach all follow from this data insight.
Statistics guide modeling decisions: The histogram shows the data is heavily right-skewed — most values are low, with a long tail of high concentrations. This means a log-transform may be needed for geostatistical interpolation. The percentile statistics help define threshold zones for the model. All of this understanding comes before a single flow equation is solved.
3Spatial Interpolation
Step 6 — Define an Interpolation Domain
Go to 'Conceptual Model Tools' → 'DrawDomain' → 'DomainPoly'. Draw a polygon on the map that roughly follows the data distribution — the envelope of your sampling network. Click 'SaveShape'. This defines the area over which interpolation will produce meaningful results (interpolation outside the data envelope is extrapolation — unreliable).
Step 7 — Perform 2D Interpolation (IDW)
In the Scatter Data interface, select 'IDW' under Tools (Inverse Distance Weighting). Click 'Apply'. A color map and contours of interpolated chloride concentration appear on the map. Expand 'Interpolation Options >>>' and check 'Legend' to display the concentration scale. The 2D interpolation shows the spatial pattern — chloride hotspots and background areas — as a continuous surface across the county.
Step 8 — Extend to 3D Interpolation
Open Domain Attributes. Under Bottom Elevation, select 'DataCenter' — the bedrock top surface from the MAGNET Data Center. In Simulation Settings, check 'Number of SubLayers = 5'. This creates a 5-layer 3D interpolation grid spanning from the land surface (DEM) to the bedrock top. In the Scatter Data interface, click 'Advanced Options', check '3D', select 'Defined by model', and click 'Apply'. The 3D interpolation runs — producing chloride concentrations at every cell in the 5-layer grid. Use the layer dropdown to view results at different depths.
43D Iso-Surface Visualization
Step 9 — Create a 3D Iso-Surface
In the Scatter Data interface, enter 400 in the field next to 'isoV1'. Click 'Create Iso-Surfaces'. The MAGNET 3D Surface Plot appears showing the 400 mg/L chloride iso-surface — a 3D "shell" enclosing all locations where interpolated chloride exceeds 400 mg/L. This is the shape of the high-concentration zone in three dimensions.
Step 10 — Add Terrain and Bedrock Surfaces
Add the DEM top surface ('ISObmp_top.png' from the first dropdown) and bedrock bottom surface ('ISObmp_bot.png' from the second dropdown). The iso-surface now sits between terrain and bedrock — you see exactly where in the subsurface the high-chloride zone exists, how deep it extends, and how it relates to the geological framework.
Step 11 — Add 3D Point Data
In Other Options (>>), click 'Add 3D Point Data' and load from the current data table. Chloride measurements appear as 3D spheres at their actual sampling depth, with lines extending up to the land surface (representing the well casing above the screen). Check 'Customized' and click 'Update Size/Color' to apply the custom color scheme from Step 4. Check 'Hide Lines' for a cleaner view. Reduce opacity of the DEM and bedrock surfaces to see through them.
Key Concepts
From points to volumes: This tutorial progressed from scattered point measurements to a continuous 3D field with iso-surfaces. The raw data (points) became 2D interpolated contours, then 3D interpolated volumes, then 3D iso-surfaces — each step adding spatial understanding. The final 3D view shows the shape, depth, and extent of the chloride anomaly in a way that no table or 2D map can match.
Data analysis informs modeling: Every insight from this tutorial feeds into modeling decisions. The depth trend (Step 5) determines whether you need a surface or deep boundary condition for chloride. The spatial pattern (Step 7) guides grid refinement. The iso-surface (Step 9) shows where the model needs the most detail. Data analysis and modeling are not separate activities — they're iterative partners.
IDW vs. kriging: This tutorial used Inverse Distance Weighting (IDW) — a simple, assumption-free interpolation method. IGW-NET also supports kriging (2D and 3D), which accounts for spatial correlation structure through variograms and provides uncertainty estimates at each interpolated point. For publications and regulatory submissions, kriging's uncertainty quantification is often required. For rapid exploration, IDW is faster and gives similar spatial patterns.
Connection to the Ottawa County case study: This tutorial uses the same chloride dataset that drove Ottawa County's Groundwater Sustainability Initiative. The 3D visualization revealed that the salinity problem was regional and depth-dependent — not localized surface contamination. This insight transformed county governance from reactive crisis management to proactive, science-based water stewardship. Data-driven understanding made the difference.
5What's Next
Continue to data integration and advanced analysis:
Tutorial 27: DataNET-based Model — build a groundwater model from federated data services
Tutorial 28: Data Processing & Regression — statistical analysis, interpolation, regression tools