Benefit from a sound exploratory data analysis to control and validate your input data. Identify data heterogeneities and characterize the contamination spatial variability in order to increase data value. In addition define and rationalise your sampling strategy according to your evaluation objectives.
DATA QUALITY CONTROL AND VALIDATION
Before starting any pollution assessment, it is essential to control the quality of your data. This preliminary investigation and cleaning work allows ensuring strong basis and supplying reliability to your characterization process. Exploratory data analysis allows identifying and handling data anomalies, heterogeneities and anisotropies using appropriate statistical representations: basemap, histogram, q-q plot, variogram map, logs, 3D viewer. Integrate all qualitative and quantitative data (historical information, geophysics, indirect measurements), replace them in their environmental context and optimize your understanding of the contamination.
SAMPLING DESIGN AND OPTIMIZATION
Geostatistics allows defining the most relevant sampling design according to the site characterization objectives that could be:
- identification of target objects with known size,
- delineation of the contamination extent,
- validation of residual concentration/activity levels after remediation using classical statistical tests.
IDENTIFICATION OF UNDER-SAMPLED AREAS
According to sampling density or more advanced uncertainty maps (confidence interval, probability of exceeding a threshold…), geostatistics helps in identifying missing data points that would improve significantly the characterization quality and reduce the associated uncertainties.