PHD dataset

About PHD >

PHD (PM2.5 Hindcast Database) is a database that provides historical PM2.5 estimates across China, during 2000-2016. Using a machine learning approach, PHD assembled datasets from multiple sources, including MODIS satellite measurements of aerosol, CMAQ modeling outputs based on MEIC historical emission inventories and many other spatiotemporal variables, and hindcasted the daily PM2.5 concentrations from 2000 to 2016 in China. PHD is developed and maintained by a team from Tsinghua University, Beijing, China.

PHD v1.0 provides annual concentrations of PM2.5 and counts of polluted days in a regular grid of 0.1° × 0.1°, across the mainland of China, from 2000 to 2010, and other results derived from the product.

For further details about PHD, please contact Professor Qiang Zhang（qiangzhang@tsinghua.edu.cn）or Dr. Tao Xue（xuetaogk_9032@126.com）.

Usage criteria >

  • PHD should not be utilized for commercial purposes.
  • For any published articles / materials or unpublished reports / products that related to PHD, please cite the following paper.

Download >

1. PHD v1.0 PM2.5 annual concentrations (μg/m3)：
PHD_v1_annual_concentraion.csv

2. PHD v1.0 Polluted-and-above days (PM2.5 > 75 μg/m3, day)：
PHD_v1_polluted_days.csv

3. PHD v1.0 Heavily-polluted-and-above days (PM2.5 > 150 μg/m3, day)：
PHD_v1_heavily-polluted_days.csv

4.Population-weighted mean concentration of PHD v1.0 in PM2.5 months(μg/m3)
PHD_v1_heavily-polluted_days.csv

Citation >

Xue T, Zheng Y, Tong D, Zheng B, Li X, Zhu T, Zhang Q. (2018). Spatiotemporal continuous estimates of PM2.5 concentrations in China, 2000-2016: a machine learning method with inputs from satellites, chemical transport model, and ground observations, Environment International. (Accepted).