1Extract Borehole Data
Step 1 — Zoom to the Area of Interest
Navigate to the Hastings greater area in southwest Michigan. This region has abundant water well borehole records in the MAGNET Data Center — the raw material for T-PROGS modeling.
Step 2 — Extract Categorized Borehole Data
Click '3D TP' under Analysis Tools. This opens the TP Model Data Options interface. There are multiple ways to add categorized borehole data: upload a shapefile or text file, enter data manually, or extract directly from the MAGNET Data Center.
Click 'Draw TP Area to extract borehole data from data center' under Borehole Well Data. Draw the extraction area on the map, then click 'SaveShape'. After a few moments, the data table populates with borehole records — each containing location, well ID, elevation, interval thickness, and material type.
Step 3 — Preview Borehole Data on Map
Click 'ShowOnMap' to display borehole markers. Click any marker to open its Borehole Information Table — showing the full lithology log with depth intervals and categorized material types. This is your raw data: the observations that T-PROGS will use to build the 3D geological model.
Data from the Data Center — Not from Scratch
The MAGNET difference: In traditional T-PROGS workflows, preparing borehole data is the most labor-intensive step — finding well logs, digitizing lithology intervals, categorizing materials, formatting files. MAGNET's Data Center has already processed millions of water well records across the USA and Canada. Categorized lithology is ready to extract for any area — click, draw, extract. The hours of data preparation become seconds.
Multiple data sources: You're not limited to the Data Center. You can upload your own borehole data (shapefiles or text files) — from site investigations, geotechnical reports, or proprietary databases. You can also combine Data Center records with your own field data for maximum coverage. The platform handles all formats through a unified interface.
2Define the 3D Domain
Step 4 — Set Top and Bottom Boundaries
Open Domain Attributes ('DomainAttr'). In the Aquifer Attributes tab:
Top Elevation: Select 'DEM Resolution' — the land surface from the Digital Elevation Model
Bottom Elevation: Select 'Data Center' — the bedrock top surface (interface between unconsolidated and consolidated materials)
This defines the 3D volume that T-PROGS will fill: from ground surface down to bedrock. The unconsolidated materials between these surfaces — sand, gravel, clay, till — are what the boreholes describe and what T-PROGS will model.
3Fit the Transition Probability Model
Step 5 — Configure and Run Pre-Simulation
In the TP Model Data Options interface, click 'Create TP Model'. In the TP Model Options interface:
Grid NX: 75 (Y adjusts automatically to maintain roughly square cells)
Zmax: 'MAX DEM' (maximum terrain elevation)
Zmin: 'Min RockTop Elevation' (minimum bedrock surface)
Click 'Pre-Simulation Only'. This calculates the observed transition probabilities from the borehole data — counting how often each lithology transition occurs at each lag distance — then fits exponential Markov Chain models to the observed data. The processing runs on the cloud; a prompt indicates when it's complete.
Step 6 — Review the Markov Chain Model
Click 'Proportion/Lens Length Results >>' to view the fitted transition probability models. For each category combination (AQ→AQ, AQ→MAQ, AQ→CM, etc.), you see the observed data (dots) and the fitted exponential model (curve). The fit quality tells you how well the mathematical model captures the geological structure. Key parameters include material proportions (how much of the subsurface is AQ vs. CM) and mean lens lengths (average thickness of each material type).
4Generate the 3D Geologic Model
Step 7 — Run the Conditional Simulation
Click 'Simulation' in the TP Model Data Options interface. The conditional sequential simulation process begins — filling every cell in the 3D grid with a lithology category. Cells containing borehole data receive the observed lithology. Cells without data are assigned categories that honor both the nearby observations and the fitted transition probability structure. Processing runs on the cloud.
Step 8 — View the Four-Panel Output
When simulation completes, four visualization panels appear automatically:
N-S Cross-Section: Centerline slice showing material variability with boreholes
W-E Cross-Section: Perpendicular centerline slice
3D Volume Chart: Interactive full or cropped 3D view with optional borehole display
3D Surface Plot: Interactive 3D surface with fence diagrams, terrain overlay, and layer import
Each material type is color-coded: AQ (bright blue) = aquifer, MAQ = marginal aquifer, PCM = partially confining, CM (red) = confining material. The boreholes are visible in the cross-sections, confirming that the simulation honors the observed data.
5Customize the 3D Visualization
Step 9 — Add Terrain Overlay to 3D Surface
Double-click in the 3D Surface Plot to enlarge it. From the first dropdown next to 'TPSurface', select 'Surface with Edges' to show colored cells with black grid lines. Then select 'TPbmp_top.png' from the top option bar to drape a terrain rendering on top of the geologic model — the DEM as a translucent overlay. Adjust transparency and vertical offset via the '>>' Other Options menu.
Step 10 — Display Wells and Bedrock Surface
Hide the TP model cells (select 'Hidden' from the TPSurface dropdown). Add the bedrock surface by selecting 'TPbmp_bot.png' from the second dropdown — the 3D bedrock topography appears. Add color-coded wells by selecting 'TP Well Only' from the last dropdown. Now you see the structural framework: terrain on top, bedrock below, and the borehole evidence in between. The geology that T-PROGS modeled fills this space.
Key Concepts
T-PROGS belongs in the conceptual model: T-PROGS generates geology — it belongs in Part I of the modeling workflow (Conceptual Model), not in the numerical simulation. The geological model defines the material distribution; the numerical model solves flow through that distribution. Getting the geology right is the foundation — everything downstream (flow, transport, calibration, uncertainty) depends on it.
One realization vs. many: This tutorial generates a single T-PROGS realization — one possible arrangement of materials that honors the data and statistics. But there are infinitely many plausible arrangements. Tutorials 15–18 showed how Monte Carlo simulation explores this uncertainty for random K fields. The same principle applies to T-PROGS: generate many realizations, simulate flow through each, and accumulate statistics. The uncertainty in geology propagates to uncertainty in predictions.
Data density matters differently: In a data-rich area (many boreholes), T-PROGS is tightly constrained — most cells are close to an observation, and the conditional simulation has little freedom. In a data-sparse area, the simulation has more freedom, and different realizations will differ more. The transition probability model provides the statistical "glue" that fills gaps consistently — but more data always reduces uncertainty.
From T-PROGS to flow model: The categorical output (AQ, MAQ, PCM, CM) converts directly to a 3D hydraulic conductivity field. Each category maps to a K value (e.g., AQ = 30 m/d, CM = 0.001 m/d). This K field feeds into IGW-NET's flow solver. The geological realism of T-PROGS — lenses, pinch-outs, connectivity patterns — produces flow behavior that uniform or simple layered models cannot capture.
6What's Next
Continue to 3D visualization and data analysis:
Tutorial 25: 3D Flow Visualization — immersive 3D flow fields, water table, and layers
Tutorial 26: 3D Point Data Analysis — borehole and well data in 3D
Tutorial 27: DataNET-based Model — build a model from federated data services