MethaneEye – Methane Plumes Detection and Quantification

Overview

Methane is a potent greenhouse gas and a major contributor to global warming. MethaneEye addresses the need for more accessible and systematic methane monitoring by integrating heterogeneous Earth observation datasets, automated data ingestion pipelines, geospatial processing, emission quantification methods, and interactive web-based visualization.

The platform combines data from Sentinel-5P, Sentinel-2, and NASA’s EMIT mission, alongside meteorological datasets such as ERA5 and GEOS-FP. This project was realized with the Africa Space Works team.

My Contributions

My main contributions focused on developing the EMIT methane plume detection and quantification module and implementing the platform’s backend and frontend.

1. EMIT Methane Plume Detection and Quantification

Developed a processing workflow using NASA’s EMIT hyperspectral satellite data to analyze and quantify methane plume emissions. The code of this part can be found here.

Key tasks included:

• Accessing and retrieving EMIT products programmatically through NASA Earthdata using the earthaccess Python API.
• Processing the EMIT L2B Estimated Methane Plume Complexes product, which provides methane enhancement measurements at approximately 60 m spatial resolution.
• Extracting methane plume geometries, concentration information, acquisition metadata, and associated uncertainty values.
• Implementing the Integrated Mass Enhancement (IME) method to estimate methane emission rates.
• Integrating meteorological wind fields from ERA5 and GEOS-FP to calculate emission fluxes and investigate the influence of wind data on quantification results.
• Comparing estimated emission rates with reference detections from the Carbon Mapper API to assess the consistency of plume locations and emission estimates.

The resulting workflow connects satellite-derived methane enhancement measurements with atmospheric wind information to produce spatially explicit plume and emission estimates.

2. Backend and API Development

Designed and implemented the backend architecture supporting data ingestion, geospatial processing, database management, and communication with the web interface.

Key contributions included:

• Developing RESTful APIs using FastAPI.
• Designing a centralized PostgreSQL/PostGIS database to store methane detections, satellite raster products, plume geometries, and infrastructure layers.
• Implementing automated data ingestion workflows for satellite observations, meteorological products, and external reference datasets.
• Developing scheduled background tasks with APScheduler to retrieve and update data from external APIs.
• Building geospatial processing services for raster and vector data.
• Implementing spatial queries and filtering based on geographic boundaries, detection types, source categories, and temporal intervals.
• Developing API endpoints to retrieve raster data, methane detections, infrastructure layers, and plume visualization metadata.
• Integrating calculations of estimated emission-related costs and penalties into the detection search workflow.

3. Interactive Geospatial Web Application

Implemented the frontend interface to make methane detection information accessible through interactive maps, analytical visualizations, and filtering tools.

Main functionalities included:

• Developing an interactive mapping interface using OpenLayers.
• Displaying methane plume geometries and raster concentration layers.
• Integrating satellite and meteorological raster products into the map interface.
• Implementing spatial and temporal filtering of methane detections.
• Developing interactive charts and statistics using Chart.js.
• Providing access to historical detection information and associated emission estimates.
• Supporting the visualization of oil and gas infrastructure alongside methane plumes to help investigate potential emission sources.

Data Sources

The platform integrates several complementary Earth observation and geospatial datasets:

Data Source	Purpose
Sentinel-5P / TROPOMI	Methane hotspot identification through CAMS products
Sentinel-2	High-resolution satellite observations for methane plume detection
NASA EMIT	Hyperspectral methane plume detection and concentration analysis
ERA5	Meteorological wind fields for emission quantification
GEOS-FP	Atmospheric wind data for emission estimation and comparison
Carbon Mapper	Reference methane plume detections and emission estimates
OGIM	Oil and gas infrastructure mapping for potential source identification

Methodology

The platform follows an integrated workflow:

1. Data Acquisition: Retrieve satellite observations, meteorological data, and infrastructure datasets from external APIs and repositories.
2. Methane Detection: Identify and integrate methane hotspots and plume geometries from Sentinel-5P, Sentinel-2, and EMIT products.
3. Emission Quantification: Apply the Integrated Mass Enhancement (IME) approach, combining methane enhancement measurements with wind information.
4. Validation: Compare EMIT-based estimates with corresponding Carbon Mapper detections and reported emission rates.
5. Infrastructure Analysis: Cross-reference detected plumes with nearby oil and gas facilities and infrastructure.
6. Data Management: Store and organize spatial detections, raster products, and associated metadata in a PostgreSQL/PostGIS database.
7. Visualization and Analytics: Provide interactive maps, filtering capabilities, historical analysis, and emission-related statistics through the web application.

Key Outcomes

• Developed an EMIT-based methane plume processing and quantification workflow using hyperspectral satellite observations.
• Integrated ERA5 and GEOS-FP wind datasets to estimate methane emission rates and examine sensitivity to meteorological inputs.
• Implemented a validation workflow using Carbon Mapper reference detections.
• Built a centralized geospatial database integrating methane detections, satellite raster products, meteorological information, and infrastructure layers.
• Developed backend APIs and an interactive web mapping interface for exploring methane emissions and associated statistics.
• Integrated emission-related cost and penalty calculations to support the assessment of the potential economic impact of methane leaks.