Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
D5922-18 – Standard Test Method Technical Guide
D5922‑18 offers a comprehensive framework for the analysis, interpretation, and modeling of spatial variation within geostatistical site investigations. Focused on regionalized variables in soil, rock, and contained fluids, this guide provides the critical analytical steps necessary to quantify spatial continuity before proceeding to estimation or simulation. It is specifically written for practitioners already familiar with geostatistical theory and is not intended for beginners.
🎯 Scope and Intended Application
The guide’s scope is precisely defined: covering recommendations for the analysis, interpretation, and modeling of spatial variation using measures like variograms and correlograms. It explicitly states that it does not promote a single method but rather presents a series of validated options. The user is responsible for applying professional judgment and adhering to all applicable safety and regulatory standards. Crucially, this standard is not a replacement for professional education, experience, or the standard of care, and not all aspects may be applicable in all circumstances.
D5922 is directly linked to D5923 (Guide for Selection of Kriging Methods) and D5924 (Guide for Selection of Simulation Approaches), establishing a logical decision-making workflow for comprehensive geostatistical site characterization.
📊 Core Measures of Spatial Variation
Central to the guidance in D5922 is the detailed discussion of the two primary tools for quantifying spatial continuity: variograms and correlograms. These functions describe how data similarity changes with distance and direction within the sampled domain.
Variogram: Plots the variance of differences between data pairs as a function of separation distance (lag). Key modeling parameters derived from this function include the nugget effect (micro-scale variability and measurement error), sill (total variance of the regionalized variable), and range (the distance beyond which data are spatially uncorrelated).
Correlogram: A normalized measure of spatial autocorrelation, providing a complementary view of spatial structure scaled between -1 and 1.
| 🟦 Standard | 📏 Title | 🎯 Application in Spatial Investigation |
|---|---|---|
| D653 | Terminology Relating to Soil, Rock, and Contained Fluids | Standard technical definitions for common terms |
| D5923 | Guide for Selection of Kriging Methods | Estimation of values at unsampled locations |
| D5924 | Guide for Selection of Simulation Approaches | Stochastic simulation of spatial heterogeneity |
💡 Professional Application Note: While D5922 provides the analytical framework, the standard emphasizes it “does not recommend a specific course of action.” The geostatistical modeling of spatial variation must be tailored to the project’s unique data density, heterogeneity, and objectives. Careful interpretation of the experimental variogram is essential before selecting a valid mathematical model and fitting it to the data.
❓ Frequently Asked Questions
🔍 What is the primary purpose of the variogram in D5922?
The variogram is the central tool for quantifying spatial continuity. It measures the average dissimilarity between data points as a function of their separation distance, providing the essential parameters needed for subsequent kriging and simulation.
💡 Who is the intended audience for this standard?
This guide is explicitly for those who are “already familiar with the geostatistical tools discussed herein.” It serves as a rigorous framework for experienced practitioners, not as an introductory tutorial on spatial analysis.
⚡ How does D5922 integrate with kriging and simulation standards?
D5922 provides the foundational analysis of spatial variation. The structural models developed under this guide are directly utilized in D5923 for selection of kriging methods (estimation) and D5924 for selection of simulation approaches (stochastic modeling).
📌 Does this standard prescribe a specific mathematical model for the variogram?
No. The standard focuses on the analysis, interpretation, and modeling process. It references established models such as spherical, exponential, and Gaussian, but leaves the selection and fitting to the practitioner’s professional judgment based on the experimental variogram and project context.