Mercury Emission and Concentration Modeling

Technical Overview

This master's thesis project at ETH Zurich involved developing a comprehensive computational framework for modeling global mercury emissions from power plants and predicting their atmospheric distribution. The project required integrating existing geospatial models, implementing custom visualization algorithms, and designing mitigation scenario analysis tools.

Technical Challenges & Solutions

Challenge 1: Reverse Engineering Visualization Standards

The project began with a significant technical challenge: reproducing reference visualizations without access to the original code or detailed specifications. I needed to reverse-engineer the color mapping, projection system, and data interpolation methods used in published research.

Solution: Developed a custom R visualization pipeline that could: - Parse geospatial mercury concentration data from multiple formats - Implement flexible color scaling algorithms to match reference standards - Create reproducible plotting functions for consistent visual comparison across datasets

Challenge 2: Multi-Model Integration

Working with two distinct computational models - one for emission calculations and another for atmospheric distribution - required careful data pipeline design to ensure compatibility and accuracy.

Solution: Built a robust data processing workflow that: - Standardized coordinate systems between different model outputs - Implemented data validation checks to ensure model consistency - Created automated comparison tools to verify cross-model accuracy

Challenge 3: Scalable Mitigation Analysis

Analyzing the impact of technology upgrades across thousands of global power plants required an efficient computational approach to handle multiple scenarios and large datasets.

Solution: Designed a modular analysis framework that: - Optimized computational efficiency for large-scale scenario modeling - Implemented parallel processing for mitigation strategy comparisons - Created automated reporting tools for policy recommendation generation

Implementation Results

Visualization Pipeline Validation

The first major milestone was successfully reproducing the reference visualization standards. This validation step was crucial for ensuring the accuracy of all subsequent analysis.

Reference Standard:

Reproduced Visualization:

Technical Achievement: Successfully reverse-engineered the visualization methodology, achieving visual consistency despite using different computational approaches (R vs. original MATLAB implementation). The color mapping algorithm accurately represents the same data ranges and geographic projections, validating the reliability of our modeling pipeline.

Scenario Visualization & Analysis

Using the existing temporal modeling system developed by the lab, I created visualizations to compare baseline conditions with technology intervention scenarios.

Baseline Mercury Concentrations (2000-level emissions):

Post-Mitigation Projections (2060 with technology upgrades):

My Contribution: Applied the lab's temporal modeling system to generate comparative visualizations, translating complex atmospheric transport data into clear, publication-ready graphics that effectively communicate the impact of different emission scenarios across multiple decades.

Geospatial Data Visualization

Working with existing geospatial databases, I created visualizations showing the global distribution of mercury emission sources to support targeted mitigation analysis.

Global Power Plant Distribution Analysis:

Visualization Challenge: Processed complex geospatial datasets containing power plant locations, emission characteristics, and technology classifications to create clear, informative plots that highlight high-impact intervention opportunities across different geographical regions.

Mitigation Technology Comparison

Using the lab's existing models, I generated comparative visualizations showing the effectiveness of different mercury control technologies across global power plant deployments.

Technology A Implementation Results:

Technology B Implementation Results:

Technical Accomplishment: Created percentage change visualizations that enable direct comparison between different technology deployment strategies, translating complex model outputs into actionable policy insights.

Technical Lessons & Methodology

Visualization Standardization Challenge

The core technical challenge of this 1.5-month project was creating a consistent visualization framework for comparing different datasets and scenarios without access to the original methodologies.

Key Technical Solutions: 1. Color Scale Reverse Engineering: Developed custom R algorithms to match reference color schemes by analyzing pixel values and interpolating color transitions 2. Consistent Scaling: Implemented standardized scaling across all visualizations to ensure linear mapping between numerical changes and color representation 3. Cross-Platform Compatibility: Created R-based solutions that could reproduce MATLAB-style outputs while maintaining data integrity

Project Impact: The visualization consistency enabled direct quantitative comparison across different scenarios, supporting evidence-based policy recommendations for global mercury emission reduction strategies.

Technical Trade-offs Identified: - Color accessibility considerations vs. reference standard matching - Visual clarity vs. detailed data representation in global-scale visualizations - Processing efficiency vs. visualization quality for large geospatial datasets