Which projection should my pipeline compute in?
Measured areal distortion for EASE-Grid 2.0, Mollweide, Albers and UTM, why Web Mercator is disqualifying for carbon work, and how to document the choice.
Read the guide
Climate/Carbon Accounting Spatial Pipelines
Build, automate, and validate satellite-driven carbon accounting and MRV systems — from raw imagery to audit-ready emissions inventories.
A technical resource for spatial MRV automation
Modern carbon accounting has moved from spreadsheets into distributed, spatially explicit data pipelines. Spatial Pipeline Engineering is a practitioner-focused resource for engineering those pipelines: deterministic geospatial processing, rigorous uncertainty quantification, and cryptographically verifiable audit trails that meet institutional standards.
Written for ESG engineers, climate data scientists, and Python GIS teams, every guide pairs production-ready code with the compliance context behind it — GHG Protocol scoping, ISO 14064, CSRD, and the registry workflows that turn satellite observations into defensible inventories.
The library is organized into four sections covering the full lifecycle: designing MRV architecture, processing satellite imagery for emissions tracking, modeling and validating carbon stocks, and orchestrating and governing the pipelines that tie them together — schema contracts, workflow orchestration, and registry standards. 84 pages in total, each with production Python, the compliance mapping behind it, and hand-drawn diagrams of the mechanisms that actually break. Pick a section below to dive in.
Start here
Four questions practitioners arrive with, and the page that answers each.
Why is my figure wrong when nothing failed?
Twelve silent failure modes that corrupt carbon numbers without failing a task — datum fallbacks, partial writes, categorical resampling — each with a diagnostic and the invariant that catches it.
Read the guideHow do I turn a methane plume into a defensible number?
Integrated mass enhancement and cross-sectional flux implemented end to end, with the uncertainty budget that shows why wind data beats sharper imagery.
Read the guideWhat happens after the credits are sold?
Reversal detection, leakage against matched controls, and the pinned issuance-era model that stops a software upgrade being reported as a forest loss.
Read the guideExplore the content
Four sections, each with focused topics and step-by-step Python implementation guides.
MRV Architecture
Core design patterns for MRV pipelines: equal-area spatial alignment, GHG Protocol scoping, spatial joins and topology, data lineage, and registry integration.
- Carbon Credit Registry Data Integration
- Emissions Data Quality & Validation Gates
- Geospatial Coordinate Reference Systems (CRS) Alignment
- GHG Protocol Scope 3 Spatial Mapping
- MRV Data Lineage & Provenance Tracking
Orchestration & Reference
Orchestrating MRV pipelines with Airflow/Prefect/Dagster, pipeline observability and failure modes, the canonical data-schema reference, and registry standards.
- Carbon Registry Standards & Methodologies
- MRV Data Schema Reference
- MRV Pipeline Observability & Failure Modes
- Orchestrating MRV Data Pipelines
Satellite Imagery
Cloud masking, temporal aggregation, distributed tile processing, deforestation alerts, and methane plume retrieval from Sentinel, Landsat and hyperspectral archives.
- Async Satellite Tile Processing with Dask
- Change Detection Algorithms for Land Cover
- Cloud-Optimized Geospatial Formats
- Deforestation Alert Generation Pipelines
- Methane Plume Detection from Hyperspectral Imagery
- Sentinel-2 & Landsat Cloud Masking Workflows
- Temporal Aggregation for Land-Use Change
Spatial Modeling
Carbon stock baselines, LiDAR/SAR biomass fusion, soil organic carbon, permanence and leakage monitoring, ground-truth alignment, and uncertainty mapping.
- Biomass Estimation from LiDAR & SAR Fusion
- Emission Factor Uncertainty Mapping
- Forest Carbon Baseline & Additionality Modeling
- Ground Truth Alignment for Carbon Models
- Permanence, Reversal & Leakage Monitoring
- Soil Organic Carbon Modeling & Validation
- Threshold Tuning for Carbon Stock Baselines