[About Sandia] [Unique Solutions] [Working With Us] [Contacting Us] [News Center] [Search] [Home] [navigation panel]
SmartSampling at Mound

SmartSampling Home Page

Bibliography

Technical Papers

Related Websites


SmartSampling [Sandia National Laboratories]

SmartSampling Training Syllabus

DRAFT training materials for individual sections may be opened
or downloaded in pdf format by clicking on the section title.
Introduction and Overview: Decision Makers and Technical Staff
Concepts and principals behind SmartSampling
Sorts of problems addressed by this process
Review of tools required and terms used (glossary provided)
Ways to use this process
Objective Function
False positive, false negative
Conceptual process flow
Section 1 - Illustrated case study: Technical Staff
In-Depth Review of outputs of the process
Section 2 - Exploratory Data Analysis: Technical Staff
Mapping of the data set
Histogram techniques
Probability-plotting techniques
Correlation between multivariate data
Data transformations
logarithmic
indicator
rank-order
normal-score
uniform-score
Trend Analysis / Removal
Section 3 - Quantification of Spatial Continuity: Technical Staff
Calculation of experimental variograms
Fitting models to experimental variograms
Bi-gaussian check
Concepts of anisotropy and nested structures in variograms
Other techniques for defining spatial variability
indicator
co-variance (relationship to semi-variance)
Hands-on exercise using software for calculation & modeling of variograms
Incorporation of subjective knowledge into modeling of variograms
Section 4 - Spatial Estimation: Technical Staff
Review of techniques for spatial estimation
nearest neighbor polygons
inverse distance to a power
trend surface
splines
kriging
Concept of a “best” linear unbiased estimate
Hand calculations to solve the kriging system on a small example data set
Hands-on exercise using kriging software on example data set
Indicator kriging, probability kriging
Explanation of kriging variance
Checking of kriging output
Section 5 - Spatial Simulation: Technical Staff
Difference between estimation and simulation
Basics of probabilistic risk assessment
Transfer of uncertainty in spatial distribution to uncertainty in decision making
Review an example spatial simulation
Hands-on exercise using software to create simulations
Process realizations to test the model - is the input reproduced?
Section 6 - Probability and Excavation Mapping: Technical Staff
Concept of probability of exceedence
Concept of probability mapping
Incorporation of spatial uncertainty in remediation maps
Cross validation (additional data)
Estimation vs. Simulation in probability mapping
Hands-on exercise using software to create probability maps
Introduce Independent Work
Problem and data set for independent solution
Initiate analysis
Break for Independent Analysis of site-specific data set
Section 7 - Economic Analyses: Technical Staff
Cost elements, Cost assumptions
Basis of cost/benefit analyses in decisions involving spatial uncertainty
Objective function
Loss functions and their effects on decisions (how cost of failure is calculated)
Linear loss function
Squared loss function
Data worth and determination of # and location of additional samples
Ranking of potential additional sampling sites
Probabilistic calculations in re regulatory action levels
Cost calculations in re probability of exceedence levels
Economic risk / Human health risk
Section 8 - Scaling Issues: Technical Staff
Measurement and remediation scales
Case study - review of effects of scale discrepancy
Analytic techniques to address scale discrepancies
Numerical techniques to address scale discrepancies
Block kriging
Averaging
Hands-on exercise to apply scaling, create cost and inventory curves
Section 9 - Independent Analysis Results: Technical Staff
Review of independent analyses
Discussion of model variation
Model Validation Exercises
Generation of Cost and Inventory curves
Section 10 - Summary/Wrap-Up: Decision Makers and Technical Staff
Solutions presented to managers and stakeholders from opening session
Results of process
review output
defensibility of plans generated by process
Next Steps

Smart Sampling was developed with major support from the U. S. Department of Energy's Office of Science and Technology.
Contact: Sean McKenna Sandia National Laboratories, 505-284-2450
Contact: Anthony Armstrong Oak Ridge National Laboratories, 423-576-1555



Back to top of page ||Back to SmartSampling Home Page || Questions and Comments || Acknowledgment and Disclaimer