ChinaHighO3: Daily 0.1 Degree Ground-Level Ozone Data for China, 2013-2020

ChinaHighO3 is a long-term, high-resolution dataset of ground-level maximum 8-hour average (MDA8) ozone concentrations across China. It was generated from big data sources using artificial intelligence by Jing Wei at the University of Maryland, College Park, and covers the period from 2013 to 2020. The dataset has a reported daily cross-validation R² of 0.87 and an RMSE of 17.10 µg m⁻³.

Use Cases

• Modeling long-term ozone pollution trends based on the 2013-2020 daily time series.
• Analyzing spatial patterns of ground-level ozone across China based on the 0.1-degree gridded data.
• Validating satellite-derived air quality estimates using the ground-measurement-informed dataset.
• Assessing population exposure to ozone for public health studies based on the full-coverage product.

Strengths

• High reported accuracy with a daily cross-validation R² of 0.87 and RMSE of 17.10 µg m⁻³.
• Full spatial coverage across China at a high resolution of 0.1 degrees (approximately 10 km).
• Long temporal coverage spanning 8 years from 2013 to 2020.
• Integrates multiple data sources including ground measurements, satellite data, and atmospheric reanalysis.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.
• Row count is unknown, which may limit suitability assessment for large-scale ML tasks.

Provenance

Source

Jing Wei, University of Maryland, College Park.

Collection Method

Generated from big data (ground measurements, satellite products, reanalysis, emission inventory) using artificial intelligence, considering spatiotemporal heterogeneity.

Time Range

2013 to 2020.

Freshness

The dataset is described as continuously updated, but the last update date is unknown.

Geography

China.