The fine mode ratio (FMF) is crucial for distinguishing between natural aerosols and anthropogenic aerosols. However, currently most satellite based land aerosol fine model fraction products are highly unreliable. Here, we have developed a new satellite based global land daily FMF dataset (Phy DL FMF) that covers the period from 2001 to 2020 at a spatial resolution of 1°, leveraging the advantages of collaborative physics and deep learning methods. Based on the analysis of 361089 data samples from 1170 AERONET sites worldwide, the Phy DL FMF dataset is comparable to the measurement results of the Aerosol Robot Network (AERONET). Overall, the root mean square error (RMSE) of Phy DL FMF is 0.136, the correlation coefficient is 0.68, and the proportion of results within the expected error (EE) range of ± 20% is 79.15%. In addition, off-site validation of the Surface Radiation Budget (SURFRED) observation results showed that the root mean square error of Phy DL FMF was 0.144 (72.50% of the results were within ± 20% of the expected error (EE) range). The performance of Phy DL FMF is superior to other deep learning or physical methods (such as the spectral deconvolution algorithm proposed in our previous research), especially in forest, grassland, cultivated land, as well as urban and barren land types. As a long-term dataset, Phy DL FMF is able to show a significant downward trend in the overall terrestrial regions worldwide (with a significance level of 95%). According to the trend analysis of Phy DL FMF in different countries, the upward trend of FMF is particularly evident in India and the western United States. In summary, this study provides a new aerosol full factor dataset for global land regions, which helps to enhance our understanding of the spatiotemporal variations of fine and coarse aerosol patterns.
In this study, MODIS C6.1 L1B MOD02SSH data (i.e. top of atmosphere (TOA) reflectance of bands 1 to 7), MODIS C6.1 L3 MOD09CMG data (surface reflectance of bands 1 to 7), and MODIS C6.1 L3 MOD08 daily data from 2001 to 2020 were used to retrieve FMF. AERONET provides a global network of solar photometers for ground aerosol characteristics. The five meteorological variables (i.e. 2-meter temperature, PBLH, surface pressure, 10 meter wind component, and 2-meter dew point temperature) are all derived from the fifth generation products of the European Centre for Medium Range Weather Forecasts (ERA5).
We leverage the advantages of collaborative physics methods and deep learning to retrieve aerosol FMFs over land using MODIS data on a global scale. We tested and validated this hybrid model using data from twenty years (2001-2020) and generated a new long-term frequency modulation dataset called Physics Deep Learning Frequency Modulation (Phy DL FMF). Unlike previous studies, the proposed hybrid model considers both physical characteristics and nonlinear relationships to limit FMF calculations.
The data quality is good.
| # | number | name | type |
| 1 | 42030606 | National Natural Science Foundation of China |
This work is licensed under a
CC BY 4.0.
CSTR:11738.11.NCDC.ZENODO.DB6467.2024
10.5281/zenodo.5105617
