Global ionospheric sporadic <i>E</i> intensity prediction from GNSS RO using a novel stacking machine learning method incorporated with physical observations

Hu, Tianyang; Xu, Xiaohua; Luo, Jia; Hou, Jialiang; Liu, Haifeng

doi:https://doi.org/10.5194/acp-25-11517-2025

Articles | Volume 25, issue 18

https://doi.org/10.5194/acp-25-11517-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-25-11517-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 18

Research article

|

29 Sep 2025

Research article |

| 29 Sep 2025

Global ionospheric sporadic E intensity prediction from GNSS RO using a novel stacking machine learning method incorporated with physical observations

Tianyang Hu, Xiaohua Xu, Jia Luo, Jialiang Hou, and Haifeng Liu

Data sets

COSMIC-1 Data Products, UCAR/NCAR – COSMIC UCAR COSMIC Program https://doi.org/10.5065/ZD80-KD74

Ionosonde data, NESSDC National Earth System Science Data Centre http://wdc.geophys.ac.cn/

Ionospheric data, UKSSDC UK Solar System Data Centre https://www.ukssdc.ac.uk/wdcc1/ionosondes/secure/iono_data.shtml

OMNI Hourly Data Set, NASA Space Physics Data Facility Natalia E. Papitashvili and Joseph H. King https://omniweb.gsfc.nasa.gov/form/dx1.html

Model code and software

Paper submission support: code and data for global ionospheric sporadic E intensity prediction from GNSS RO using a novel stacking machine learning method incorporated with physical observations Tianhang Hu and Xiaohua Xu https://doi.org/10.5281/zenodo.15092794

Short summary

Sporadic E (Es) layers are an irregularity in ionospheric E region. Their formation is related to multiple atmospheric physical and chemical processes. Accurate Es intensity prediction is significant for understanding atmospheric coupling. We proposed a novel stacking machine learning method incorporating physical observations to achieve higher-precision global Es intensity prediction than previous methods. Our results indicate the importance of considering related physical factors for Es prediction.