Zhang et al. (2025) Retrieving atmospheric water vapor profiles over Europe combining NOAA-20/CrIS and ground-based GNSS-PWV data

Identification

• Journal: Atmospheric Research
• Year: 2025
• Date: 2025-11-29
• Authors: Jingyuan Zhang, Qinglan Zhang, Shirong Ye, Hong Hu, Yanlan Wu, Peng Jiang
• DOI: 10.1016/j.atmosres.2025.108666

Research Groups

• School of Resources and Environmental Engineering, Anhui University, Hefei, China
• National Geomatics Center of China, Beijing, China
• GNSS Research Center, Wuhan University, Wuhan, China
• National Engineering Research Center for Agro-Ecological Big Data Analysis and Application, Hefei, China
• School of Artificial Intelligence, Anhui University/Engineering Center for Geographic Information of Anhui Province, Hefei, China
• State Key Laboratory of Opto-Electronic Information Acquisition and Protection Technology, Hefei, China
• Anhui Engineering Research Center for Geographical Information Intelligent Technology, Hefei, China
• Information Materials and Intelligent Sensing Laboratory of Anhui Province, Hefei, China

Short Summary

This study proposes a novel method to improve the accuracy of atmospheric water vapor vertical profile retrievals over Europe by combining NOAA-20/CrIS infrared hyperspectral data with ground-based GNSS-PWV observations using Transformer and Random Forest models, demonstrating significant accuracy enhancements, especially under cloudy conditions, compared to official CrIS products.

Objective

• To improve the accuracy of atmospheric water vapor vertical profile retrievals by combining infrared hyperspectral data with ground-based GNSS observations using machine learning techniques, addressing limitations of traditional sounding methods such as low spatiotemporal resolution and strong susceptibility to cloud interference.

Study Configuration

• Spatial Scale: Europe
• Temporal Scale: Not explicitly stated in the provided text, but implies continuous monitoring capabilities due to satellite and GNSS data.

Methodology and Data

• Models used: Transformer, Random Forest (RF)
• Data sources: NOAA-20/CrIS infrared hyperspectral data, ground-based GNSS-derived Precipitable Water Vapor (GNSS-PWV) observations, Radiosonde data from IGRA (used as reference truth), official CrIS product (CrIS-EDR) for comparison.

Main Results

• Under clear-sky conditions, the incorporation of GNSS-PWV data reduced the average Root Mean Square Error (RMSE) of the Transformer and RF models by approximately 23.7% and 44.9%, respectively, achieving overall accuracy comparable to the CrIS-EDR product.
• Under cloudy conditions, the inclusion of GNSS-PWV data led to reductions in RMSE of approximately 28.4% and 37.8% for the Transformer and RF models, respectively, relative to retrievals based solely on CrIS data.
• The Bias (BIAS) values under cloudy conditions were further improved compared with the CrIS-EDR product.
• The joint retrieval using infrared hyperspectral and ground-based GNSS-PWV data significantly enhances the accuracy of specific humidity profile retrievals, showing clear advantages over the CrIS-EDR product, particularly under cloudy conditions.

Contributions

• Proposes a novel approach for retrieving atmospheric water vapor vertical profiles by synergistically combining infrared hyperspectral satellite data (CrIS) with ground-based GNSS-PWV observations.
• Applies advanced machine learning models (Transformer and Random Forest) to enhance retrieval accuracy, addressing limitations of traditional sounding methods.
• Demonstrates significant improvements in water vapor profile accuracy, particularly under challenging cloudy conditions, surpassing the performance of official CrIS products (CrIS-EDR).
• Provides a method to overcome the issues of low spatiotemporal resolution and cloud interference in traditional water vapor sounding.

Funding

• Anhui Provincial Natural Science Foundation, China (grant number 2308085MD126)
• Anhui Provincial Ecological Environment Science and Technology Program, China (grant number 2024hb001)
• Hefei Municipal Natural Science Foundation, China (grant number 202323)

Citation

@article{Zhang2025Retrieving,
  author = {Zhang, Jingyuan and Zhang, Qinglan and Ye, Shirong and Hu, Hong and Wu, Yanlan and Jiang, Peng},
  title = {Retrieving atmospheric water vapor profiles over Europe combining NOAA-20/CrIS and ground-based GNSS-PWV data},
  journal = {Atmospheric Research},
  year = {2025},
  doi = {10.1016/j.atmosres.2025.108666},
  url = {https://doi.org/10.1016/j.atmosres.2025.108666}
}