Source Apportionment of POPs in the Indus River Basin via PCA/MLR

Last Updated: February 19, 2026
Estimated reading time: ~7 minutes

Source Apportionment of POPs is a critical environmental forensic technique used to identify and quantify the origins of pollution in complex ecosystems. By applying statistical models like Principal Component Analysis (PCA) and Multiple Linear Regression (MLR) to chemical data from the Indus River, researchers can distinguish between local agricultural runoff, industrial emissions, and transboundary atmospheric transport. This study provides a masterclass for students in deciphering the “chemical fingerprints” left by Persistent Organic Pollutants (POPs) in water, air, and soil.

Key Takeaways

• Statistical Power: PCA/MLR models successfully categorized pollution sources into agricultural pesticides, industrial emissions, and mixed sources.
• Agricultural Dominance: In the Alluvial Riverine Zone (ARZ), DDT metabolites were the primary markers, confirming heavy reliance on agrochemicals.
• Industrial Signatures: Low-lying urban areas (LLZ) showed distinct PCB signatures linked to e-waste, transformers, and industrial thermal processes.
• Mountain Sinks: The Frozen Mountain Zone (FMZ) revealed that “pristine” areas are polluted primarily through Long Range Atmospheric Transport (LRAT).
• Diagnostic Ratios: Ratios of parent compounds to metabolites (e.g., DDT/DDE) served as crucial indicators of fresh versus aged pollution.

---

The PCA/MLR Analytical Approach

To conduct a robust Source Apportionment of POPs, mere concentration measurements are insufficient; scientists must understand the relationships between different chemical contaminants. The study utilized Principal Component Analysis (PCA) to group co-occurring chemicals into “factors” or “components,” which imply a shared origin. Multiple Linear Regression (MLR) was then used to estimate how much each source contributed to the total pollution load.

“Principal component analysis/multiple linear regression (PCA/MLR) describes and identifies between the related and the unrelated source tracer and also percentage contribution from the various sites” (Sohail, 2018, p. 71).

In the Indus River study, this method helped untangle complex data sets. For example, if DDTs and Endosulfan load heavily on the same PCA component, it suggests a shared agricultural source. Conversely, if PCBs load on a separate component, it points to a distinct industrial source. This statistical separation is vital for environmental management, as it tells regulators whether to target farmers or factories to reduce pollution.

Student Note / Exam Tip: PCA reduces a large dataset into smaller “components” explaining the variance (grouping), while MLR quantifies the contribution of those components to the total concentration.

Professor’s Insight: Think of PCA as sorting trash into “plastic,” “paper,” and “glass” bins based on material properties; MLR then weighs each bin to see which contributes most to the landfill.

---

Agricultural Fingerprints in the Riverine Zone

The application of Source Apportionment of POPs in the Alluvial Riverine Zone (ARZ)—the agricultural heartland of Punjab—revealed clear evidence of pesticide inputs. The statistical models identified specific “fingerprints” dominated by DDTs and their metabolites. By analyzing the ratios of parent compounds (like p,p’-DDT) to their breakdown products (like p,p’-DDE), the study could determine whether the pollution was from historical residue or recent, illegal application.

“PC-1 factor accounted 54.18 % of the total variance and was dominated by p,p′-DDE… and highlighted the current and/or aged usage of DDTs in these well-known agricultural areas” (Sohail, 2018, p. 72).

The data indicated that while some contamination was due to the “legacy” of old pesticides (aerobic degradation leading to DDE), there were also signals of fresh inputs. This is particularly concerning as it suggests ongoing use of banned substances for crop protection or vector control (malaria). The correlation of these chemical factors with agricultural land use confirmed that surface runoff from farms is a primary driver of riverine pollution in this zone.

Student Note / Exam Tip: A ratio of (DDE + DDD) / DDT > 0.5 typically indicates “aged” or historical weathering, whereas a ratio < 0.5 suggests recent pollution inputs.

PCA Component	Variance Explained	Dominant Chemicals	Interpretation of Source
PC-1 (FMZ)	33.33%	DDTs + Hepta/Octa-PCBs	Long Range Transport (Dust)
PC-2 (FMZ)	26.18%	HCB, Hexa-CBs	Mixed Atmospheric Deposition
PC-1 (ARZ)	54.18%	p,p’-DDE, p,p’-DDT	Agricultural Runoff (Agrochemicals)
PC-1 (LLZ)	>90%	Mixed OCPs + PCBs	Industrial & Agricultural Mix

Fig: Summary of Source Apportionment results derived from PCA loadings across different ecological zones (reformatted from Sohail, 2018, pp. 71-72).

Professor’s Insight: In environmental forensics, identifying the specific isomer (e.g., o,p’ vs p,p’) is crucial; technical grade DDT has a specific ratio, while Dicofol (a miticide) has a different one, allowing you to pinpoint the exact product used.

---

Industrial Sources and Urban Fractionation

In the Low Lying Zone (LLZ) and urban centers, Source Apportionment of POPs shifted from agricultural signatures to industrial ones. The PCA results here highlighted a “mixed source” heavily influenced by Polychlorinated Biphenyls (PCBs) and Hexachlorobenzene (HCB). These chemicals are associated with industrial activities, including electronics manufacturing, transformer maintenance, and waste incineration.

“For the low-lying zone (LLZ), PC-1… highlighted a mixed source containing pesticides and electronics. Agricultural applications of pesticides as well as industrial activities are the main reasons” (Sohail, 2018, p. 73).

The study noted high levels of heavier PCB congeners (like Hepta-CBs) in urban dust and soil. This points to local emissions from “urban fractionation”—where industrial activities create a “halo” of pollution around cities. Sources include open burning of e-waste (to recover metals), leakage from old transformers, and unintentional release from thermal processes like steel manufacturing. The statistical association of PCBs with population density further cemented the link between urbanization and industrial pollution.

Student Note / Exam Tip: Urban Fractionation refers to the tendency of varying chemical volatility to cause pollutants to distribute differently as they move away from a city center source.

Professor’s Insight: The presence of specific PCB congeners (like PCB-11) can sometimes track back to pigment and dye manufacturing rather than electrical equipment, serving as a specific industrial tracer.

---

Long Range Atmospheric Transport (LRAT)

Perhaps the most intriguing application of Source Apportionment of POPs was in the Frozen Mountain Zone (FMZ). Despite being a remote area with minimal local industry, the PCA revealed significant loads of POPs. The source apportionment models attributed this to Long Range Atmospheric Transport (LRAT), where air masses carry pollutants from distant regions (like India, China, or southern Pakistan) to the high-altitude cryosphere.

“These trends indicated that in the colder regions (FMZ), POPs burden was dominated by transboundary movement via LRAT process which may act as secondary sources of POPs” (Sohail, 2018, p. 94).

The mechanism at play is “cold trapping,” where semi-volatile chemicals vaporize in warm regions and condense in cold mountains. The PCA loaded heavy PCBs and DDTs onto the same component in these areas, suggesting they traveled together on dust particles via air currents. Back-trajectory analysis of wind patterns confirmed that air masses originating from Central Asia and neighboring industrial hubs were depositing these contaminants onto glaciers, which then release them into the Indus River during melt season.

Student Note / Exam Tip: HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) is the standard computer model used to trace the origin of air masses in environmental studies.

Professor’s Insight: Glaciers are not just frozen water; they are “chemical time capsules” that trap pollutants from decades ago and release them back into the ecosystem as they melt due to climate change.

---

Reviewed and edited by the Professor of Zoology editorial team. Except for direct thesis quotes, all content is original work prepared for educational purposes.

Real-Life Applications

1. Regulatory Enforcement: By distinguishing between “fresh” and “aged” DDT, authorities can determine if farmers are currently breaking the ban on organochlorine pesticides or if the residue is from legal historical use.
2. Transboundary Pollution Disputes: Source apportionment data proving LRAT allows countries like Pakistan to present evidence in international forums regarding pollution drifting in from neighboring nations.
3. Waste Management Policy: Identifying PCBs from e-waste in urban zones highlights the need for specific regulations on the recycling and disposal of electronic goods to prevent soil contamination.
4. Remediation Targeting: Knowing that ARZ pollution is driven by surface runoff while FMZ pollution is atmospheric helps engineers design specific intervention strategies (e.g., buffer zones for farms vs. air filters).

This matters because allocated funds for cleanup must target the actual root cause—factories, farms, or wind currents—to be effective.

Key Takeaways

• Forensic Statistics: PCA and MLR are not just math; they are forensic tools that solve the “whodunit” of environmental pollution.
• Zonal Differences: The source of pollution changes geographically: Air transport dominates in mountains, agriculture in plains, and industry in coastal/urban zones.
• Climate Connection: Global warming accelerates the release of “trapped” sources in glaciers, complicating source apportionment by mixing old and new pollution.
• E-Waste Legacy: The presence of PCBs in non-industrial zones often traces back to informal e-waste recycling, a growing source in developing nations.
• Diagnostic Power: Ratios of chemical isomers provide a timeline of contamination, acting as a chemical clock for researchers.

MCQs

1. What does the detection of a p,p’-DDE/p,p’-DDT ratio < 1 indicate in Source Apportionment?
A) Historical/Aged usage of DDT
B) Fresh/Recent usage of DDT
C) Long Range Atmospheric Transport
D) Microbial degradation
Correct: B
Explanation: A ratio less than 1 indicates that the parent compound (DDT) is still abundant relative to its breakdown product (DDE), suggesting recent input into the environment (Sohail, 2018, p. 58).

2. In the Frozen Mountain Zone (FMZ), what was identified as the primary source of POPs?
A) Local pesticide manufacturing
B) Industrial effluent discharge
C) Long Range Atmospheric Transport (LRAT)
D) Agricultural surface runoff
Correct: C
Explanation: The study notes that in colder regions like FMZ, the pollution burden was dominated by transboundary movement via the LRAT process (Sohail, 2018, p. 94).

3. Which statistical method was used to group co-occurring chemicals into source factors?
A) ANOVA (Analysis of Variance)
B) PCA (Principal Component Analysis)
C) T-test
D) HYSPLIT modeling
Correct: B
Explanation: Principal Component Analysis (PCA) was the method used to extract components and identify related source tracers across the sites (Sohail, 2018, p. 71).

FAQs

Q: What is Source Apportionment of POPs?
A: It is the scientific process of identifying where pollutants come from (sources) and quantifying how much each source contributes to the total pollution in a specific area.

Q: Why use PCA/MLR instead of just measuring concentrations?
A: Concentrations only tell you how much pollution is there. PCA/MLR reveals patterns and correlations that tell you where it came from (e.g., a specific type of industry vs. farming).

Q: Can these methods detect illegal pesticide use?
A: Yes. By analyzing the ratios of parent compounds to metabolites, scientists can prove if a banned pesticide like DDT was recently applied to crops.

Q: What is “Cold Trapping”?
A: It is a phenomenon where volatile pollutants evaporate from warm areas, travel through the atmosphere, and condense/deposit in cold areas (like high mountains or poles), effectively getting “trapped” there.

Lab / Practical Note

Data Integrity: When performing Source Apportionment using PCA, ensure your sample size is sufficient (usually >5 samples per variable). Always normalize or standardize your chemical data (e.g., log-transformation) before running the software to prevent high-concentration chemicals from skewing the results (Sohail, 2018, p. 54).

External Resources

• US EPA: Receptor Models for Source Apportionment
• ScienceDirect: Principal Component Analysis in Environmental Sciences

Sources & Citations

Distribution of Persistent Organic Pollutants (POPs) among Different Environmental Media (Air, Soil, Water, Biota) from Indus River Flood-Plain, Pakistan, Muhammad Sohail, Supervisor: Dr. Syed Ali Musstjab Akber Shah Eqani, COMSATS University Islamabad, Pakistan, 2018, pp. 54, 71-74, 94-96.

• PDF Correction/Note: Statistical tables were summarized into descriptive text and simplified Markdown tables for clarity.
• Correction Invite: If you are the author of this thesis and wish to provide updates or corrections, please contact us at contact@professorofzoology.com.

Author Box
Muhammad Sohail is a doctoral researcher at COMSATS University Islamabad. His work specializes in using advanced statistical models like PCA and MLR to track the environmental fate and transport of persistent organic pollutants across Pakistan’s diverse topography.

Disclaimer: The summary provided here is for educational purposes only and is based on a specific academic thesis. It does not constitute professional environmental consulting advice.

Reviewer: Abubakar Siddiq

Note: This summary was assisted by AI and verified by a human editor.