Estimation of Representative Errors of Surface Observations of Air Pollutant Concentrations Based on High-Density Observation Network over Beijing-Tianjin-Hebei Region

Ground stations provide raw monitored point data of air pollutant concentration, and three-dimensional chemical models can simulate the concentration distribution on grids. When ground stations observations are used for verification of model performance or assimilated into model simulations, representative errors may occur due to the mismatch between the spatial scales of discrete monitored point data and model simulations. This study produces a high-resolution observational dataset for Beijing-Tianjin-Hebei region by combining the information obtained from China National Monitoring Center and from local monitoring centers. The combined dataset allows the computation of representative errors of ground observations of six typical air pollutants (PM_2.5, PM₁₀, SO₂, NO₂, CO and O₃) in Beijing-Tianjin-Hebei region. Results from the aforementioned method are compared with those obtained from the theory of Elbern et al. (2007). It is found that the results from the two methods agree well in terms of the representative errors of ground observations of NO₂. However, representative errors of SO₂, CO and O₃ are significantly underestimated by using the Elbern's approach. Therefore, characteristic parameters associated with the four air pollutants used in the Elbern's method are modified and new characteristic parameters relevant to PM_2.5 and PM₁₀ are introduced in the present study, which makes the method to be more applicable and can yield more accurate results when processing ground observations in Beijing-Tianjin-Hebei region.