This section is experimental. It provides the comparison for the prediction for several individual forecast models

For a full list of all Air Quality Forecast models being analyser, check the forecast models page: https://aqicn.org/forecast/models/cn/

How accurate are the forecasts?

Check the correlation graphs below for PM_2.5, PM₁₀ and Ozone for the past 30 days.

(Note that all values are based on the AQI, and that a maximum threshold of AQI 300 is used).

In the above graph, the forecast is data is the forecast one day in advance (i.e. the forecast computed today for tomorrow).

We need your help

The above prediction is based on composite meta model, computed using several Air Quality Forecasting Systems (AQFS):

• The Spectral Radiation-Transport Model for Aerosol Species (Sprintars) model from Kyushu University.
• The China National Air Quality forecast by Zhejiang Experts in Experimental Sciences (PM2.5科学实验专家小组).
• The Air Korea Air Quality forecast, by the Korean Integrated Air Quality forecast center(대기질 통합예보센터).
• The Copernicus Atmosphere Monitoring Service (CAMS) near-realtime surface ensemble forecast , by the ECMWF.
• The System for Integrated modeLling of Atmospheric coMposition (SILAM) by the Finnish Meteorological Institute (FMI).
• The Max Planck Institut fur Meteorologie (MPIM) model.
• The Ozone Air Quality Model (AQM) by the NOAA/NCEP.
• The Regional Air Quality Deterministic Prediction System (RAQDPS) model from the Government of Canada.
• The irCELine (Belgian Interregional Environment Agency) for the Belgian Regions.
• The LCSQA (Laboratoire Central de Surveillance de la Qualité de l’Air) for France.
• The RIVM (Rijksinstituut voor Volksgezondheid en Milieu) for the Netherlands.
• The SILAM (System for Integrated modeLling of Atmospheric coMposition) for Northen Europe.
• The EURAD (EURopean Air Pollution Dispersion) for Central Europe.

The above map is based on the PM_2.5 surface level modeling, and colors are following the US EPA AQI standard.

This forecasting model, and all AQFS which it is based on, are research products intended to provide information related to Air Quality forecast. All reasonable measures have been taken to ensure its quality and accuracy. However:

• We do not make warranty, express or implied, nor assume any legal liability or responsibility for the accuracy, correctness, completeness of the information.
• We do not assume any legal liability or responsibility for any damage or loss that may directly or indirectly result from any information contained on this website or any actions taken as a result of the content of this website;
• We may change, delete, add to, or otherwise amend information contained on this website without notice

For more information about the underlying concepts of Air Quality forecasting (or Atmospheric Dispersion Modeling), check the article on a visual study of wind impact of PM_2.5 concentration.