Evaluation and comparison of Himawari-8 L2 V1.0, V2.1 and MODIS C6.1 aerosol products over Asia and the Oceania regions

Abstract

Himawari-8 aerosol products are useful for forecasting weather, environmental monitoring, and climate change research. In 2018, after receiving much attention from researchers, the Himawari-8 Level2 (L2) aerosol retrieval algorithm was updated from Version 1.0 (V1.0) to Version 2.1 (V2.1). Although previous studies have examined the accuracy of Himawari-8 aerosol products, only a few studies have evaluated and compared the V1.0 and V2.1 algorithms or described how the Himawari-8 aerosol products differ from those of MODerate-resolution Imaging Spectrometer (MODIS). In this study, we validated and compared the Himawari-8 L2 V1.0 and V2.1 aerosol products at 500 nm using data from all Asian Aerosol Robotic Network (AERONET) sites for the period from 2016 to 2017. Furthermore, a comprehensive comparison between the Himawari-8 L2 V2.1 and MODIS C6.1 AOD (Terra and Aqua) products was conducted. The V1.0 and V2.1 AOD agreed well with AERONET AOD (R: 0.67–0.72) with 44.65% and 49.38% of retrievals fell within the EE, respectively. Meanwhile, V1.0 and V2.1 AOD bias increased with the AOD magnitude, but V1.0 AOD tended to underestimate the AOD. The underestimation has now improved in the V2.1 AOD products. Both V1.0 and V2.1 AOD performed better over forests and grasslands than croplands, as well as urban and barren lands and had the best performance in summer, while the worst in spring. Overall, the V2.1 AOD outperformed in term of retrievals falling within the EE compared to V1.0 AOD, but it still has large estimation uncertainties in high aerosol loadings areas and sparsely vegetated areas, which indicating the aerosol algorithm needed to improve in the future. Furthermore, the data quality of V2.1 AE/FMF products significantly improved compared to that of V1.0. Compared to the MODIS C6.1 AOD products, Himawari-8 L2 V2.1 AOD accuracy is slightly lower but demonstrates a similar spatial distribution.