The core idea of this machine learning approach is to rapidly convert the CO spectral features extracted from GIIRS measurements into columns through a trained model and simultaneously estimate the uncertainty based on the error propagation theory. The model is trained in spatially and temporally representative radiative transfer simulations. Comparisons with the retrieval results of traditional physical methods and ground-based observations reveal consistent spatial distribution and temporal variation across these different datasets.
Dr. Dasa Gu, a leading researcher on the project, stated that “our results confirm that machine learning methods have the potential to provide reliable CO products without the computationally intensive iterative process required by traditional retrieval methods. However, characterizing the instrument sensitivity of machine learning retrieval results is one issue that needs to be addressed before operational retrieval.”
###
References
DOI
Original Source URL
https://doi.org/10.34133/remotesensing.0289
Funding information
The work is supported by the Hong Kong Research Grants Council (26304921, 16306224) and Hong Kong Environment and Conservation Fund (78-2019, 13-2022). C.Z. and M.Z. are supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (grant XDB41000000) and the USTC Research Funds of the Double First-Class Initiative.
About Journal of Remote Sensing
The Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.