Estimating hourly PM1 concentrations from Himawari-8 aerosol optical depth in China

Highlights: • PCA-GRNN for hourly, daily, monthly PM₁ estimation with R² of 0.65, 0.70, 0.81 • PM1-estimation performance is better at noon and in the afternoon. • PM₁ can be significantly affected by anthropogenic emissions. • PM₁ is the main component of PM_2.5 and PM₁₀, accounting for more in PM2.5 Particulate matter with diameter less than 1 μm (PM₁) has been found to be closely associated with air quality, climate changes, and even adverse human health. However, a large gap in our knowledge concerning the large-scale distribution and variability of PM₁ remains, which is expected to be bridged with advanced remote-sensing techniques. In this study, a hybrid model called principal component analysis-general regression neural network (PCA-GRNN) is developed to estimate hourly PM₁ concentrations from Himawari-8 aerosol optical depth in combination with coincident ground-based PM₁ measurements in China. Results indicate that the hourly estimated PM₁ concentrations from satellite agree well with the measured values at national scale, with R² of 0.65, root-mean-square error (RMSE) of 22.0 μg/m³ and mean absolute error (MAE) of 13.8 μg/m³. On daily and monthly time scales, R² increases to 0.70 and 0.81, respectively. Spatially, highly polluted regions of PM₁ are largely located in the North China Plain and Northeast China, in accordance with the distribution of industrialisation and urbanisation. In terms of diurnal variability, PM₁ concentration tends to peak in rush hours during the daytime. PM₁ exhibits distinct seasonality with winter having the largest concentration ($31.5\pm 3.5$ μg/m³), largely due to peak combustion emissions. We further attempt to estimate PM_2.5 and PM₁₀ with the proposed method and find that the accuracies of the proposed model for PM₁ and PM_2.5 estimation are significantly higher than that of PM₁₀. Our findings suggest that geostationary data is one of the promising data to estimate fine particle concentration on large spatial scale.