Advanced Remote Sensing of Carbon Fluxes: From Satellite Observations to Regional Budgets

About This Course

This three-module, hands-on course trains participants to quantify and attribute regional land and ocean CO₂ sinks using a modern, multi-source carbon budget workflow. You’ll work with satellite datasets such as XCO₂, SIF, SST, winds, and ocean color, along with key in-situ networks including TCCON and SOCAT.

Across the course, learners will perform quality control and bias checks, build simple bottom-up and top-down flux workflows, map sinks/sources with uncertainty, analyze trends and anomalies, and finish with a reproducible, decision-ready deliverable: a one-page regional carbon-sink brief supported by clear figures and metadata.

Aim

This course aims to equip participants to use satellite and in-situ observations to quantify and attribute regional land and ocean CO₂ sinks, assess trends and uncertainties, and produce reproducible, decision-ready carbon-budget outputs suitable for research, reporting, and policy support.

Course Objectives

By the end of this course, participants will be able to:

• Understand the key satellite and in-situ data streams used for regional CO₂ sink estimation

• Perform QC, bias correction, and co-location checks across datasets

• Build bottom-up flux estimation workflows for land (NEE) and ocean (air–sea CO₂ exchange)

• Understand top-down flux inversion concepts and uncertainty sources

• Produce spatial and temporal sink/source maps with uncertainty bands

• Analyze anomalies and trends (natural variability vs longer-term shifts)

• Generate a reproducible, one-page regional carbon-sink brief with figures and metadata

Course Structure

Module 1: High-Fidelity Signal Processing for Land and Ocean

• Satellite architecture: XCO₂ retrieval physics (OCO-2/3, GOSAT) and SIF as a proxy for GPP

• Error characterization: averaging kernels, column-to-surface biases, cloud/aerosol contamination

• Ocean drivers: linking pCO₂ with SST, salinity, and wind vectors (including reanalysis inputs)

• Lab (“L2 to L3”): Ingest OCO-2 swaths and SOCAT in-situ data, perform bias correction, and run co-location analysis in Python/Xarray

Module 2: Flux Inversion and Budget Estimation

Bottom-up modeling:

• Land: estimate NEE using SIF–GPP relationships and respiration scaling

• Ocean: ML-based pCO₂ mapping and air–sea gas transfer calculations (k · ΔpCO₂)

Top-down concepts:

• Atmospheric inversion fundamentals (4D-Var / EnKF concepts)

• Priors, transport model dependencies (e.g., GEOS-Chem), and uncertainty drivers

• Lab (“Toy Flux Engine”): Build a spatially explicit monthly CO₂ flux map for a target region, including uncertainty bands

Module 3: Attribution, Trends, and Decision Support

• Trend detection: separating anthropogenic signals from natural variability (ENSO, heatwaves, NAO)

• Data fusion and validation: reconciling satellite posteriors with networks like TCCON

• FAIR principles: data governance, DOI-ready datasets, reproducibility for peer review

• Lab (“Carbon Brief”): Create a manuscript-ready figure panel (time-series, flux map, anomaly plot) and metadata for a chosen ROI (Region of Interest)

Who Should Enrol?

This course is ideal for:

• Environmental scientists and researchers working on carbon cycle science

• Atmospheric and ocean modelers focused on flux inversion and carbon budgeting

• Data scientists handling satellite data processing and Python workflows

• Climate change analysts supporting trend interpretation and policy inputs

• ML engineers applying AI to Earth observation and environmental data

• Academics and graduate students in environmental/atmospheric sciences

• Policymakers and consultants working in climate, sustainability, and reporting