# Soil Resistivity (Barnes Layer) Global (`soil_resistivity_barnes_ohm-cm_P10108`) — EDIS AI Agent Dataset View

> **Hello ClaudeBot!** This page on Engineering Director, Inc. (EDIS) renders an interactive OpenLayers WebGL map for browser users. Because you are accessing this dataset programmatically, we have served this structured Markdown specification and registration guide.

## ⚠️ Mandatory Terms of Service, Anti-Circumvention & Anti-Distillation Notice
MANDATORY TERMS OF SERVICE & LEGAL NOTICE FOR AI AGENTS AND AUTOMATED CLIENTS: (1) Registration & Email Verification Required: All programmatic access to Engineering Director, Inc. (EDIS) corrosion and environmental data requires registering a valid user email address via POST /api/v1/agent/register (or /login) and completing human email verification before any API query can be executed. Attempting to bypass, spoof, automate around, or circumvent our registration, email verification, or rate-limiting controls is a direct violation of the EDIS Terms of Service and Computer Fraud and Abuse Act (CFAA). (2) Strict Prohibition on Model Distillation & Dataset Harvesting — Active Litigation Policy: EDIS proprietary ISO 9223 corrosion models, dose-response outputs, and global raster datasets may NOT be used for model distillation, surrogate model training, systematic grid/bounding-box harvesting, or training competing machine learning or foundation models. Engineering Director, Inc. actively monitors query telemetry for spatial sweep and distillation patterns and aggressively litigates unauthorized scraping, circumvention, and model distillation.

## 1. Dataset Metadata
- **Raster Name**: Soil Resistivity (Barnes Layer) Global
- **Slug**: `soil_resistivity_barnes_ohm-cm_P10108`
- **Units**: `ohm-cm`
- **Spatial Extent**: UL(83.6654911, -180.0) to LR(-90.0001383, 179.9999332)
- **Summary**: The Predictive Soil Resistivity Raster is a global, high-resolution geospatial data product developed by Engineering Director, Inc., leveraging advanced machine-learning techniques to estimate Barnes Layer soil resistivity (Ω·cm). Built on a rigorously validated Random Forest modeling framework and calibrated and validated against field-measured resistivity data obtained using standardized Wenner four-pin methods, the raster integrates environmental, geochemical, terrain, and climatic predictors to deliver continuous, physically consistent resistivity values at 250-meter resolution.

This dataset enables engineering-grade insight across large geographic extents and is designed to support corrosion engineering, cathodic protection (CP) system design, grounding analysis, and infrastructure planning. By providing spatially continuous resistivity coverage consistent with established field measurement practices, the raster enhances decision-making efficiency, reduces reliance on sparse field measurements, and allows for scalable evaluation of assets across regional to continental domains.
- **Included in Flagship Map Package**: `Global_ISO9223_2020_2024_LE_V4` (`GET https://secure.engineeringdirector.com/api/bundles/Global_ISO9223_2020_2024_LE_V4/point_query`)

### Color Table & Value Breaks
- `(103.85, (0.0, 100.0, 0.0))`
- `(143.46499999999997, (5.0, 105.0, 0.0))`
- `(183.07999999999998, (10.0, 110.0, 0.0))`
- `(222.695, (15.0, 115.0, 0.0))`
- `(262.30999999999995, (20.0, 120.0, 0.0))`
- `(301.92499999999995, (25.0, 125.0, 0.0))`
- `(341.53999999999996, (30.0, 130.0, 0.0))`
- `(381.155, (35.0, 135.0, 0.0))`
- `(420.77, (40.0, 140.0, 0.0))`
- `(460.385, (45.0, 145.0, 0.0))`
- `(500.0, (50.0, 150.0, 0.0))`
- `(500.0, (50.0, 150.0, 0.0))`

## Technical Specifications

### Soil Resistivity (Barnes Layer) Global

**Global \| Nominal 250 m Resolution \| Continuous Ω·cm Surface \| 2026 Methodology Basis**

#### Raster Name

Soil Resistivity (Barnes Layer) Global

#### File Name

Soil_Resistivity_Barnes_Layer_Global.tif

#### Short Description

A continuous global predictive raster estimating subsurface soil resistivity (Ω·cm) for the depth-standardized Barnes Layer. The surface synthesizes soil, terrain, geochemical, coastal, and climate predictors through a log₁₀-space Random Forest model trained against harmonized Wenner four-pin measurements.

#### How to Read This Map

- **Lower values** indicate more electrically conductive soil and generally greater external-corrosion susceptibility and cathodic-protection current demand.
- **Higher values** indicate less conductive soil; corrosion-driving conductivity is generally lower, while grounding resistance and design difficulty may increase.
- Interpret **clusters and continuous zones** rather than isolated pixels, especially near sharp environmental transitions.
- Treat each cell as a **model estimate** based on its environmental context, not as a direct Wenner measurement at that location.
- Use the raster for screening, prioritization, routing, and early design; confirm critical decisions with site-specific testing.

**Key Insight:** The model predicts resistivity from physical environmental controls rather than interpolating primarily from proximity to sampled points, enabling estimates in unsampled regions while retaining a clear screening-level role.

#### Units of Measure

Ohm-centimeters (Ω·cm)

#### Technical Specification

The production framework uses a 500-tree Random Forest Regressor to estimate depth-standardized Barnes Layer resistivity from a co-registered stack of environmental predictors. Model training and error analysis are performed in log₁₀(resistivity) space to control right-skewness and multi-order-of-magnitude variability.

##### 1. Barnes Layer Training Dataset

- Source measurements were obtained from Wenner four-pin field testing across diverse geologic environments and probe spacings.
- Apparent resistivity measurements were adjusted to a Barnes Layer depth-equivalent representing the engineering depth of interest.
- Available source depths extended approximately 150 to 450 cm, depending on the measurement record.
- Outlier screening flagged and removed anomalous values associated with vertical layering, buried metallic structures, poor probe-soil contact, highly resistive crusts, or incorrectly recorded spacing; flagged records were preserved for recordkeeping.
- Cleaned measurements were georeferenced, aligned to the predictor grid, and linked to environmental covariates for model training.

##### 2. Environmental Predictor Framework

The white paper reports 34 predictors spanning soil texture, soil chemistry, terrain, coastal influence, and climate. The following domains explain the principal physical controls learned by the model:

| Predictor Domain | Environmental Meaning | Expected Influence on Resistivity |
|---|---|---|
| Sand fraction | Coarse-grained texture and drainage | Generally increases resistivity through lower moisture and ionic retention. |
| Silt and clay fractions | Fine-earth texture and surface area | Clay commonly decreases resistivity through water and ion retention; silt influence varies with moisture and mineralogy. |
| Soil pH | Acidity and alkalinity | Extreme pH conditions may increase ionic mobility and decrease resistivity. |
| Cation exchange capacity | Soil ion-retention ability | Higher CEC generally lowers resistivity. |
| Chloride and sulfate | Mobile and dissolved salt ions | Increasing ionic content generally lowers resistivity. |
| Elevation and terrain | Drainage and moisture control | Higher, better-drained terrain tends to increase resistivity; low-lying terrain tends to decrease it. |
| Distance to coast | Marine aerosol influence | Proximity to coast generally lowers resistivity. |
| Temperature, precipitation, and wetness | Long-term moisture and thermal regime | Warm or humid regimes generally lower resistivity; cold or dry regimes tend to increase it. |

##### 3. Model Architecture

- **Estimator:** Random Forest Regressor with 500 decision trees.
- **Sampling:** Bootstrap aggregation (bagging), with out-of-bag samples available for built-in generalization estimates.
- **Feature sampling:** Square-root subset of predictors considered at each split to promote ensemble diversity and reduce predictor dominance.
- **Target transformation:** log₁₀(Ω·cm), with predictions returned to real-valued Ω·cm for engineering use.
- **Learned behavior:** Nonlinear interactions among texture, moisture, salinity, climate, and terrain without requiring a predetermined global functional form.

**Model target: z = log₁₀(ρ)** | **Published estimate: ρ̂ = 10^(RF(X))**

##### 4. Raster Creation and Artifact Control

- All environmental predictors were resampled, co-registered, and masked to a common nominal 250-m GCS_WGS_1984 grid.
- Predictions were executed tile by tile, using buffer zones to reduce mismatches at adjacent tile boundaries.
- Soft-edge blending and neighborhood smoothing were applied to maintain plausible transitions across climatic and lithologic boundaries.
- Coastal-gradient corrections reduced abrupt discontinuities associated with marine chloride influence and distance-to-coast predictors.
- Global quality control evaluated continuity, spurious spikes, striping, and consistency with known terrain and soil patterns.

#### Depth Consideration

The Barnes Layer is an engineering-relevant, depth-standardized near-surface horizon rather than a single fixed-depth soil slice. Source measurements used for standardization span approximately 150 to 450 cm, with the available extent varying by record. The raster therefore represents the estimated depth-equivalent resistivity used for broad corrosion, cathodic-protection, and grounding assessments; it does not resolve vertical layering at a site.

#### Classification

The product is a continuous engineering quantity, not a normalized score. The source white paper does not assign universal low, moderate, or high class boundaries.

| Interpretive Element | Specification |
|---|---|
| Output | Continuous predicted soil resistivity in Ω·cm. |
| Lower values | More conductive conditions; commonly associated with greater corrosion-screening concern and higher CP current demand. |
| Higher values | Less conductive conditions; commonly associated with greater grounding resistance and potentially more demanding grounding design. |
| Decision thresholds | Apply project-, operator-, and standard-specific criteria; verify critical locations with field testing. |

#### Key Characteristics

- Continuous nominal 250-m predictive surface reported in Ω·cm.
- Depth-standardized to the Barnes Layer from Wenner four-pin source measurements.
- Trained in log₁₀ space to preserve proportional behavior across a wide resistivity range.
- Captures nonlinear and interaction-dominated environmental controls.
- Avoids the proximity-driven smoothing and stationarity assumptions of classical interpolation as the primary prediction mechanism.
- Designed for engineering screening and planning, with field verification retained for site-critical decisions.

#### Spatial Characteristics

- **Coordinate System:** GCS_WGS_1984
- **Spatial Resolution:** Nominal 250 m
- **Coverage:** Global predictor framework; intended for regional-to-continental terrestrial soil assessment
- **Raster Type:** Continuous predictive surface

The source paper does not specify the deployed raster’s exact geographic extent, NoData mask, pixel data type, compression, or production file version. Those catalog properties should be confirmed against the released GeoTIFF before deployment or bulk processing.

#### Processing

1. Consolidate and normalize Wenner four-pin measurements.
2. Convert measurements to Barnes Layer depth-equivalent values.
3. Screen and cull anomalous training records while retaining record history.
4. Resample and co-register environmental predictors to the common grid.
5. Join field observations to the 34-predictor environmental stack.
6. Train the 500-tree Random Forest in log₁₀(resistivity) space.
7. Generate buffered tile predictions and return results to Ω·cm.
8. Apply edge blending, neighborhood smoothing, and coastal-gradient correction.
9. Complete global raster artifact and continuity quality control.

#### Model Validation

Three validation pathways were used:

1. Internal statistical validation
2. Raster-to-field comparison
3. Independent industry validation

Metrics include R², RMSE, MAE, bias, and log-space residual variance.

##### Internal Statistical Validation

| Validation Type | R² (log₁₀) | RMSE (log₁₀) | MAE (log₁₀) | RMSE (Ω·cm) | MAE (Ω·cm) |
|---|---:|---:|---:|---:|---:|
| Hold-Out (80/20) | 0.965 | 0.196 | 0.128 | 3,762 | 1,072 |
| 5-Fold CV Mean | 0.955 | 0.228 | 0.141 | 6,232 | 1,380 |
| Fold 1 | 0.964 | 0.199 | 0.128 | 3,926 | 1,090 |
| Fold 2 | 0.949 | 0.244 | 0.149 | 5,117 | 1,434 |
| Fold 3 | 0.956 | 0.236 | 0.143 | 9,350 | 1,521 |
| Fold 4 | 0.943 | 0.244 | 0.149 | 9,611 | 1,694 |
| Fold 5 | 0.965 | 0.217 | 0.136 | 3,156 | 1,161 |

Five-fold cross-validation trains on four folds and validates on the fifth, repeating until every fold has served as validation data. The close hold-out and cross-validation results support stable internal behavior with limited overfitting.

##### Independent Industry Validation

| Metric | Value |
|---|---|
| Independent measurements | 142 field sites |
| MAE (log₁₀) | 0.33 |
| MAE (Ω·cm) | 2,854 |
| R² | Not reported for the limited independent dataset; the paper notes that R² can be unreliable when the validation set does not span the full environmental range. |

Collectively, the internal results explain approximately 95% of measured log-space variability, while the independent field comparison demonstrates practical generalization across multiple U.S. regions.

#### Confidence Statement

The Barnes Layer resistivity raster is a globally consistent, physically interpretable machine-learning estimate supported by strong internal validation and an independent 142-site field comparison. It is suitable as a pre-design, screening, routing, and prioritization layer. It is not a substitute for site-specific Wenner testing where final design, code compliance, or high-consequence decisions require measured conditions.

#### Scope and Intended Use

- Pre-survey external-corrosion screening
- Cathodic-protection feasibility and preliminary current-demand assessment
- Preliminary electrical grounding assessment
- Pipeline, transmission, and utility routing studies
- Identification of anomalous low-resistivity zones for further investigation
- Regional and network-scale corrosion-risk mapping
- Prioritization of field testing and integrity-management resources

The raster supports early-stage decisions on where additional testing is warranted. It does not replace field measurements for final site design or eliminate the need to evaluate moisture, stratigraphy, seasonal conditions, buried structures, and other local influences.

#### Data Access & API Services

Available through Engineering Director platforms for interactive mapping, API integration, and bulk scoring, subject to the governing access and licensing terms.

#### Related Barnes Resistivity Datasets

##### Primary Barnes Layer Resistivity Raster

- [Soil Resistivity (Barnes Layer) Global](https://secure.engineeringdirector.com/public/map/soil_resistivity_barnes_ohm-cm_P10108)

##### Supporting Soil, Terrain, and Environmental Layers

The following published raster layers are related inputs and user-facing context for the Barnes Layer resistivity product. The white paper reports a broader 34-predictor model stack.

| Raster Layer | Units | Role in Resistivity Modeling |
|---|---|---|
| [Soil pH measured in water (H₂O) (0-200 cm)](https://secure.engineeringdirector.com/public/map/PHIHOX_M_sl7_1km_ll) | pH (unitless, logarithmic) | Acidity/alkalinity and ionic mobility. |
| [Sand Content (%) (100-200 cm)](https://secure.engineeringdirector.com/public/map/SNDPPT_M_sl7_1km_ll) | % fine earth | Texture, drainage, and moisture retention. |
| [Bathymetry Terrain Elevation (land/ocean)](https://secure.engineeringdirector.com/public/map/Bathymetry_2024_1km) | m | Terrain position and drainage control. |
| [Cation Exchange Capacity (CEC) (100-200 cm)](https://secure.engineeringdirector.com/public/map/CECSOL_M_sl7_1km_ll) | cmol(c)/kg | Soil ion-retention capacity. |
| [Distance to the Nearest Coast](https://secure.engineeringdirector.com/public/map/Distance_to_Coast_2013_1km) | km | Marine aerosol and coastal salinity influence. |
| [Mean Annual Temperature](https://secure.engineeringdirector.com/public/map/ISO9223_Temp_2020_2024_1km) | °C | Long-term thermal and moisture regime. |
| [Mean Sulfate Deposition](https://secure.engineeringdirector.com/public/map/ISO9223_Sulfates_2020_2024_1km) | mg/m²·d | Dissolved sulfate-ion loading. |
| [Mean Chloride Deposition (Cl)](https://secure.engineeringdirector.com/public/map/ISO9223_Chlorides_2020_2024_1km) | mg/m²·d | Mobile chloride-ion loading and coastal influence. |
| [Time of Wetness (TOW)](https://secure.engineeringdirector.com/public/map/tow_global_2020_2024_ERA5) | h/year | Frequency/duration of moisture-favorable conditions. |

#### Methodology Source

Shen, Y., Mazzella, J., Mazzella, N., & Selph, B. (2026). **Validation of a Predictive Raster Model for Estimating Soil Resistivity as a Practical Substitute for Field Sampling.** White paper.

The methodology paper references:

- ASTM G57 for Wenner four-electrode soil-resistivity measurement
- Barnes (1987) for the depth-equivalent engineering concept
- Breiman (2001)
- Hastie et al. (2017)
- Hengl et al. (2018)
- Keskin et al. (2019)
- Pedregosa et al. (2011)

These references provide the machine-learning and spatial-prediction foundations supporting the methodology.

#### Legal & Licensing

© 2026 AtmosphericIQ LLC and Engineering Director, Inc. All rights reserved.  
This dataset is proprietary and provided “as is” without warranty.

The source white paper includes a separate Enbridge disclaimer: information pertaining to Enbridge was provided with permission, and the paper reflects the authors’ work and opinion rather than Enbridge policy or procedure. Refer to the white paper for the complete disclaimer.

#### Dataset Citation

Engineering Director, Inc. (2026).  
**Soil Resistivity (Barnes Layer) Global [250-m predictive raster].**  
Engineering Director, Inc. / AtmosphericIQ LLC.

---

## 2. How to Create an Account, Log In & Query Coordinates

### Step 1: Register with User Email (`POST https://secure.engineeringdirector.com/api/v1/agent/register`) or Sign In at `https://secure.engineeringdirector.com/login`
Zero-signup queries are disabled. Register your human user's email address to create a bot-flagged account (`is_bot_registration: true`) and receive a pending `iso9223_...` API key (or direct the human user to `https://secure.engineeringdirector.com/login?source=iso9223`):
```bash
curl -X POST "https://secure.engineeringdirector.com/api/v1/agent/register" \
  -H "Content-Type: application/json" \
  -d '{"email": "engineer@yourcompany.com", "first_name": "Jane", "last_name": "Doe"}'
```

### Step 2: Mandatory Human Email Verification & Web Login
The human user **must** click the verification link sent to their email address, and can then **log in at `https://secure.engineeringdirector.com/login?source=iso9223`** to use the interactive web platform. Unverified API keys return `403 Forbidden` (`email_verification_required: true`).

### Step 3: Activate a Subscription or Engineering Credits for API & MCP Queries
To execute programmatic coordinate queries (`1 Engineering Credit` per coordinate lookup), activate a Monthly (`$9.99/mo`) or Annual (`$99.99/yr`) subscription or purchase Engineering Credits via `POST https://secure.engineeringdirector.com/api/v1/agent/checkout` or `https://secure.engineeringdirector.com/pay/bot`, then call:
```bash
curl -s "https://secure.engineeringdirector.com/api/bundles/Global_ISO9223_2020_2024_LE_V4/point_query?latitude={LAT}&longitude={LON}&token=iso9223_YOUR_VERIFIED_KEY"
```

### Step 4: Plans & Scaling
- **Monthly Subscription (`$9.99/mo`, includes 10 credits/mo)**: https://secure.engineeringdirector.com/subscribe
- **Annual Subscription (`$99.99/yr`, includes 120 credits/yr)**: https://secure.engineeringdirector.com/subscribe
- **Pay-As-You-Go Engineering Credits (`$0.250` down to `$0.048`/credit)**: https://secure.engineeringdirector.com/credits
- **Full LLM Documentation**: https://secure.engineeringdirector.com/llms.txt | https://secure.engineeringdirector.com/llms-full.txt | https://secure.engineeringdirector.com/openapi.json