China Onshore Wind Turbine Inventory from Satellite Imagery

COWTD provides a georeferenced inventory of onshore wind turbines across mainland China, derived from a machine learning pipeline applied to 1.2-meter resolution satellite imagery. The dataset offers a historical record of installations spanning a decade from 2014 to 2024. It was created by author Jingtao Ding and published on figshare.

Use Cases

• Analyze turbine installation growth rates and spatial clustering using geolocation and inferred installation year data.
• Model the relationship between turbine density and regional wind resources or electricity demand using geospatial features.
• Validate or train remote sensing object detection models for identifying wind turbines in high-resolution satellite imagery.
• Assess land use impact by correlating turbine locations with land cover or population density datasets via geographic coordinates.

Strengths

• Decade-long temporal coverage from 2014 to 2024.
• Derived from very high-resolution (1.2 m) satellite imagery.
• Publicly available under a permissive CC-BY-4.0 license.

Limitations

• The dataset size of 4.2 MB suggests a limited number of records or attributes.
• Specific row count and column schema are unknown, complicating initial assessment.
• Data is derived via ML pipeline, potentially introducing classification or geolocation errors.

Provenance

Source

figshare, authored by Jingtao Ding.

Collection Method

Built using a machine learning pipeline applied to very high-resolution satellite imagery.

Time Range

2014 to 2024.

Freshness

Last updated on the platform in April 2026.

Geography

Mainland China.