Yong et al. (2025) Analysis of hydrological time-lag effects using multiple GNSS techniques: GNSS-R-retrieved soil moisture, GNSS-derived coordinates, and GNSS-based water vapor data

Identification

• Journal: GPS Solutions
• Year: 2025
• Date: 2025-09-12
• Authors: Weiao Yong, Xiaolei Wang, Jinsheng Tu, Ying Gao
• DOI: 10.1007/s10291-025-01954-1

Research Groups

School of Earth Sciences and Engineering, Hohai University, Nanjing, People’s Republic of China

Short Summary

This study proposes an integrated approach using multiple Global Navigation Satellite System (GNSS) techniques, including GNSS reflectometry (GNSS-R) soil moisture retrieval, GNSS positioning, and GNSS-based water vapor data, to analyze hydrological time-lag effects in the Southern United States and Central America. The research reveals significant and spatially heterogeneous time-lag relationships among precipitable water vapor (PWV), soil moisture (SM), GNSS vertical displacement, and vegetation water content (VWC).

Objective

• To analyze the time-lag relationships among precipitable water vapor (PWV), GNSS vertical displacement, soil moisture (SM), and vegetation water content (VWC) by comprehensively utilizing multiple GNSS techniques.

Study Configuration

• Spatial Scale: Southern United States to Northern Central America (7°N to 37°N, 60°W to 125°W), with gridded data at 36 km × 36 km (EASE-Grid 2.0) and point observations from 50 GNSS stations.
• Temporal Scale: One year (January 1 to December 31, 2023) with daily temporal resolution for most variables.

Methodology and Data

• Models used:
  • CatBoost machine learning model for soil moisture retrieval.
  • Lead-lag analysis for quantifying time-lag effects.
  • Singular Spectrum Analysis (SSA) for denoising GNSS vertical displacement time series.
  • GipsyX software for GNSS data processing (PWV, vertical displacement).
  • Vienna Mapping Function 1 (VMF1) and European Centre for Medium-Range Weather Forecasts (ECMWF) data for PWV derivation.
  • Precise Point Positioning (PPP) for GNSS vertical displacement.
  • Bilinear interpolation for gridding PWV data.
• Data sources:
  • CYGNSS: Level 1 (L1) V3.2 data (January 1 to December 31, 2023) for GNSS-R soil moisture retrieval.
  • SMAP: Level 3 (L3) Radiometer Daily Soil Moisture product (36 km EASE-Grid 2.0) as reference SM, and auxiliary parameters (Roughness Coefficient (RC), Vegetation Opacity (VO), Vegetation Water Content (VWC), Soil Surface Temperature (SST)).
  • Nevada Geodetic Laboratory (NGL): GNSS Precipitable Water Vapor (PWV) and GNSS vertical displacement data for 2023 from 50 stations.
  • Pekel water mask: 30 meter resolution for excluding water-covered areas.
  • Land cover type data for regional analysis.

Main Results

• The CatBoost model achieved high accuracy in retrieving soil moisture from CYGNSS data, with an average correlation coefficient (R) of 0.93 and a root mean square error (RMSE) of 0.04 cm³/cm³.
• Significant time-lag effects were observed among hydrological variables, with notable heterogeneity influenced by vegetation cover type:
  • GNSS vertical displacement lags PWV by approximately 2–4 months (e.g., 80 days at TNMT, 105 days at MXTX, 115 days at ROD1).
  • Soil moisture lags PWV by approximately 0–20 days in woody savannas.
  • Vegetation water content (VWC) lags PWV by approximately 0–80 days (40–80 days in forests, 0–40 days in woody savannas, 30–60 days in shrubland and grasslands).
  • VWC lags soil moisture by approximately 0–30 days.
• The time-lag analysis results derived from CYGNSS SM data showed high consistency with those obtained using SMAP SM data, with mean absolute errors (MAE) of R between SM and VWC of 0.13, and between PWV and SM of 0.10. Most absolute errors in lag days were within 0–20 days.

Contributions

• Proposes and validates a novel integrated approach for analyzing hydrological time-lag effects by combining multiple GNSS techniques (GNSS-R SM, GNSS coordinates, and GNSS-based water vapor).
• Successfully retrieves large-scale, continuous, and accurate soil moisture data for the Southern United States and Central America using CYGNSS data and the CatBoost machine learning model.
• Quantifies and demonstrates the significant and spatially heterogeneous time-lag relationships among PWV, GNSS vertical displacement, SM, and VWC, providing new insights into hydrological dynamics under climate change.
• Highlights the effectiveness of GNSS technology for high spatial and temporal resolution hydrological variable analysis.

Funding

• National Nature Science Foundation of China (Grant No: 42474048, 42004018).

Citation

@article{Yong2025Analysis,
  author = {Yong, Weiao and Wang, Xiaolei and Tu, Jinsheng and Gao, Ying},
  title = {Analysis of hydrological time-lag effects using multiple GNSS techniques: GNSS-R-retrieved soil moisture, GNSS-derived coordinates, and GNSS-based water vapor data},
  journal = {GPS Solutions},
  year = {2025},
  doi = {10.1007/s10291-025-01954-1},
  url = {https://doi.org/10.1007/s10291-025-01954-1}
}