Site Soil Pollution Source Apportionment Analysis: Methods, Testing, and Applications

Site soil pollution source apportionment analysis is a specialized technical service that identifies, characterizes, and quantifies the contribution rates of different contamination sources to soil at industrial, agricultural, and redevelopment sites. By integrating chemical fingerprinting, isotopic tracing, and receptor modeling techniques such as Positive Matrix Factorization (PMF), this analysis provides the scientific basis for remediation planning, risk assessment, and environmental liability determination. This article explains the sampling strategies, testing parameters, analytical methods, quality assurance, and typical applications of professional soil source apportionment studies conducted by accredited third-party laboratories.

Technical Overview

Source apportionment is the process of qualitatively identifying the types of pollution sources affecting a site and quantitatively estimating the contribution of each source to the observed contaminant concentrations in soil. It answers two fundamental questions in site investigation: which sources caused the contamination, and to what extent each source contributed. The analysis is essential when multiple potential sources overlap spatially, when remediation scope must be prioritized by risk, or when legal responsibility must be allocated among historical operators.

A complete source apportionment program combines three elements: systematic field sampling in accordance with ISO 18400 series soil sampling standards, high-precision laboratory characterization of contaminants and isotopic fingerprints, and mathematical source-receptor modeling. Results are typically benchmarked against soil screening values such as those in GB 36600-2018 for development land, and the entire workflow is embedded in a conceptual site model (CSM) consistent with phased environmental site assessment practice under ASTM E1527 and ASTM E1903 guidelines.

The deliverable of the analysis is a quantified source contribution table, a source profile library, spatial distribution maps of each resolved source, and an uncertainty evaluation. Qualitative apportionment identifies source categories such as atmospheric deposition, industrial discharge, agricultural inputs, traffic emissions, or natural geogenic background, while quantitative apportionment expresses each source contribution as a percentage of measured contamination.

Testing Samples

Surface Soil Samples

Surface samples, generally collected from 0 to 20 cm or 0 to 50 cm depth, capture contamination from atmospheric deposition, dust fallout, spills, and agricultural inputs. They are collected on a systematic grid or judgmental layout across the site and surrounding receptor areas to characterize horizontal variability and identify hot spots.

Subsurface Soil and Core Samples

Continuous soil cores obtained by drilling provide depth-profile samples at defined intervals, typically 0.5 to 2 m. Vertical concentration profiles distinguish surface-deposited pollution from deep sources such as leaking underground storage tanks, buried waste, or lateral migration of contaminated groundwater, and they support three-dimensional source delineation.

Source and Reference Material Samples

Candidate source materials are collected to build local source profiles. These may include industrial sludge, fly ash and bottom ash, coal and fuel residues, metal-bearing waste, paint chips, fertilizer and manure, roadside dust, and sediment from drainage channels. Matched analysis of receptor soil and candidate source samples greatly improves the reliability of quantitative apportionment.

Background Soil Samples

Unimpacted background samples from geologically similar, upwind or upstream locations establish the natural geochemical baseline. These samples allow statistical separation of geogenic element concentrations from anthropogenic enrichment and are indispensable for heavy metal apportionment.

Associated Environmental Media

Groundwater, sediment, and soil gas samples collected during the same campaign extend the source interpretation. Co-located groundwater plume chemistry and soil gas signatures help verify vadose-zone sources and confirm migration pathways identified in the soil data.

Testing Items

The analyte suite is tailored to site history but typically includes the following parameter groups:

• Heavy metals and metalloids: As, Cd, Cr, Cu, Hg, Ni, Pb, Zn, and extended elements such as Sb, Be, Co, Mo, Se, Tl, V, Mn, and Ag.
• Volatile organic compounds (VOCs): benzene, toluene, ethylbenzene, xylenes, chlorinated ethenes and ethanes, and petroleum volatiles.
• Semi-volatile organic compounds (SVOCs): 16 priority polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls (PCBs), organochlorine pesticides (OCPs), and phthalate esters.
• Petroleum hydrocarbons: total petroleum hydrocarbon fractions in the C10–C40 range and alkylated PAH series for petroleum fingerprinting.
• Isotopic fingerprints: lead isotope ratios (206Pb/207Pb, 208Pb/206Pb, 204Pb-based ratios), strontium isotopes, carbon and nitrogen isotopes, and compound-specific isotope signatures for chlorinated solvents and petroleum hydrocarbons.
• Rare earth elements (REEs) and conservative tracers: La, Ce, Nd series plus Ti, Zr, Al, and Sc used for geochemical normalization and provenance discrimination.
• Soil physicochemical properties: pH, soil organic carbon, cation exchange capacity, particle size distribution, and iron and manganese oxide contents, which control contaminant retention and support statistical normalization.

Testing Methods

Laboratory Instrumental Analysis

Metals are extracted by aqua regia digestion following ISO 11466 or dissolved totally with hydrofluoric acid under ISO 14869-1, then quantified by ICP-MS or ICP-OES in accordance with ISO 17294-2 and ISO 22036. Mercury and arsenic are determined by atomic fluorescence spectrometry, and field-portable X-ray fluorescence is used for on-site screening and dense grid pre-characterization. PAHs are analyzed by GC-MS under ISO 18287, petroleum hydrocarbons by GC-FID under ISO 16703, VOCs by purge-and-trap GC-MS under ISO 22155, and organochlorine pesticides by GC-ECD under ISO 10382. Lead isotope ratios are measured by multi-collector ICP-MS or thermal ionization mass spectrometry, and compound-specific isotope analysis (CSIA) is performed by GC coupled to isotope ratio mass spectrometry.

Background Evaluation and Contamination Assessment

Enrichment factors, geoaccumulation indices, and pollution load indices normalized to conservative elements such as Ti, Zr, or Al are calculated to separate anthropogenic enrichment from natural background. Concentrations are compared against regional background values and risk screening values in GB 36600-2018 to prioritize pollutants entering the apportionment model.

Multivariate Statistical Analysis

Pearson or Spearman correlation analysis, hierarchical cluster analysis, and principal component analysis (PCA) or factor analysis are applied to the validated data matrix to reveal element associations, group sampling points with similar signatures, and generate preliminary hypotheses about source categories before quantitative modeling.

Receptor Model Quantification

Positive Matrix Factorization (PMF) is the principal quantitative tool. It decomposes the sample-by-species data matrix into factor profiles and factor contributions while enforcing non-negativity and incorporating individual measurement uncertainty. Model robustness is verified through residual diagnostics, Q-value evaluation, bootstrap resampling, and displacement tests. Complementary receptor approaches include the chemical mass balance (CMB) model when local source profiles are well characterized, the UNMIX model, and absolute principal component score regression (APCS/MLR).

Isotope and Molecular Fingerprint Tracing

Lead isotope fingerprints distinguish gasoline-derived lead, coal combustion, smelting emissions, and ore-derived or paint-derived lead. PAH diagnostic molecular ratios differentiate pyrogenic sources such as coal and biomass combustion from petrogenic petroleum inputs. Petroleum biomarkers including steranes, terpanes, and alkylated PAH patterns support refined oil source matching, and CSIA of chlorinated ethenes separates multiple solvent releases and indicates the degree of in-situ degradation.

Spatial and Geostatistical Analysis

Kriging interpolation, hot spot analysis, and spatial autocorrelation statistics convert point data into continuous distribution maps for each contaminant and each resolved source factor. Horizontal maps combined with vertical concentration profiles are cross-checked against the site layout, historical operations, and dominant wind or drainage directions to confirm the physical plausibility of modeled sources.

Quality Assurance and Quality Control

Accredited third-party laboratories apply full analytical QA/QC including method blanks, duplicate samples, matrix spikes, calibration verification, and certified reference materials with recovery and precision acceptance criteria. Field QA/QC covers equipment decontamination blanks and co-located replicate samples. Only datasets passing data validation are entered into receptor models, ensuring apportionment results are legally defensible.

Applications

Source apportionment analysis serves a wide range of regulatory, engineering, and forensic purposes:

• Redevelopment of industrial legacy sites: identifying contributions from former chemical plants, smelters, steelworks, gasworks, and electroplating facilities to define remediation scope and cost allocation before land reuse.
• Environmental liability and forensics disputes: providing quantitative evidence on responsibility among multiple historical operators, supporting insurance claims, and withstanding scrutiny in environmental litigation.
• Remediation design and prioritization: ranking source contributions to focus remediation budgets on dominant pathways such as surface deposition versus deep subsurface leakage.
• Agricultural soil protection: separating fertilizer, manure, pesticide, irrigation water, and atmospheric deposition contributions to farmland heavy metal and pesticide accumulation.
• Mining and smelting impact assessment: delineating the footprint of ore processing and stack emission influences relative to natural mineralization background.
• Petroleum and fuel release investigation: fingerprinting multiple fuel types, ages, and release events at oil depots, filling stations, and pipeline corridors.
• Waste disposal site evaluation: characterizing landfill leachate and informal dumping impacts on surrounding soils and groundwater recharge areas.
• Regulatory compliance reporting: supporting detailed investigation reports, risk assessments, and remediation feasibility studies required under soil pollution prevention and land management regulations.

FAQ

What is the difference between source identification and source apportionment?

Source identification is qualitative; it determines which source types, such as industrial emissions, agricultural inputs, traffic, or natural background, are present. Source apportionment goes further by quantitatively estimating the percentage contribution of each identified source to measured soil concentrations, typically through receptor models such as PMF combined with isotopic fingerprints.

How many soil samples are required for reliable quantitative apportionment?

For stable PMF results, datasets of at least 60 to 100 valid samples are recommended, with more than 100 preferred for robust factor resolution. Preliminary qualitative analysis using correlation and cluster methods can be performed with smaller datasets, but contribution percentages from very small datasets carry high uncertainty.

Which pollutants can be quantitatively apportioned?

Heavy metals and PAHs are the most commonly and reliably apportioned contaminant groups because multiple marker species are available. Petroleum hydrocarbons can be apportioned using biomarker and alkylated PAH fingerprints, and chlorinated solvent sources can be differentiated using compound-specific isotope analysis. Single-species contaminants require isotope or spatial evidence instead of multi-species receptor models.

How is natural geogenic background distinguished from anthropogenic pollution?

Background soil samples, geochemical normalization to conservative elements, enrichment factor and geoaccumulation index calculations, and comparison of vertical profiles are used together. Deep soil samples unaffected by surface activity and local geochemical background databases provide the primary geogenic reference.

What is Positive Matrix Factorization and why is it preferred?

PMF is a multivariate receptor model that factorizes the data matrix into a set of source profiles and corresponding contributions. It weights each data point by its analytical uncertainty and constrains all factors to non-negative values, which produces physically interpretable source profiles. Its built-in uncertainty handling and diagnostic tools make it the international standard tool for quantitative source apportionment.

Can the results be used in environmental disputes and litigation?

Yes, when sampling and analysis follow recognized ISO, ASTM, and national standard methods, the laboratory is accredited, QA/QC documentation is complete, and the modeling includes uncertainty analysis. Complete chain-of-custody records and validated data are essential for evidentiary use.

What is the typical project timeline and deliverable set?

A standard project including sampling, laboratory analysis, modeling, and reporting typically requires 6 to 12 weeks depending on site size, sample count, and isotope testing scope. Deliverables include the sampling and analysis report, validated database, source profile library, quantified contribution results with uncertainty, spatial distribution maps, updated conceptual site model, and remediation recommendations.