Zhou et al. (2025) A novel Hankel spectrum analysis filtering for reducing North-South stripes in GRACE gravity solutions

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-11-14
• Authors: Hao Zhou, Liexing Li, Hanwen Hu, Haoran Xue, Hong Li, Yuelong Liu, Zhicai Luo
• DOI: 10.1016/j.jhydrol.2025.134602

Research Groups

National Gravitation Laboratory and School of Physics, Institute of Geophysics, Huazhong University of Science and Technology, Wuhan 430074, China

Short Summary

This study introduces the Hankel Spectrum Analysis Filtering (HSAF) framework to effectively reduce North-South striping noise in GRACE-derived terrestrial water storage anomalies. HSAF significantly improves the accuracy of water storage estimates by preserving hydrological signals and outperforming conventional filters across global and basin scales.

Objective

• To propose and validate a novel filtering framework, Hankel Spectrum Analysis Filtering (HSAF), to decompose GRACE-derived terrestrial water storage anomalies into orthogonal modal components, explicitly separating high-frequency North-South striping noise without extensive empirical parameter tuning, thereby improving the accuracy of derived water storage estimates.

Study Configuration

• Spatial Scale: Global and basin scales, including 112 global river basins and the Caspian Sea.
• Temporal Scale: Monitoring of time-variable gravity field and terrestrial water storage variations, including annual amplitudes and long-term trends.

Methodology and Data

• Models used: Hankel Spectrum Analysis Filtering (HSAF) for noise reduction and signal reconstruction. Compared against conventional filters such as Gaussian smoothing and DDK4.
• Data sources: GRACE and GRACE Follow-On (GRACE-FO) mission spherical harmonic coefficient (SHC) solutions, synthetic datasets, and satellite-altimetry observations for validation.

Main Results

• HSAF effectively reduces striping artifacts while accurately preserving hydrological signals.
• HSAF yields higher correlation, lower residual errors, and improved signal fidelity compared to conventional filters across a range of global and basin scale statistical metrics.
• For 112 global river basins, HSAF lowers basin-scale Root Mean Square Error (RMSE) to an average of 1.65 cm, representing a mean reduction of 22 % relative to Gaussian smoothing and outperforming conventional filters in 50 % of the basins examined.
• For the Caspian Sea, the leakage-corrected HSAF series achieves an RMSE of 6.09 cm (34 % and 44 % lower than Gaussian and DDK4, respectively) and reproduces an annual amplitude of 18.34 cm and a trend of −9.01 cm per year, both closest to satellite-altimetry observations.

Contributions

• Introduction of Hankel Spectrum Analysis Filtering (HSAF), a novel and versatile filtering framework for GRACE-type spherical harmonic coefficient solutions.
• HSAF enables localized signal reconstruction through dual singular value decomposition and adaptive windowing, explicitly separating high-frequency North-South striping noise with reduced need for empirical parameter tuning.
• Demonstrated significant improvement in the accuracy of terrestrial water storage anomaly estimates by effectively reducing noise while preserving hydrological signals, outperforming conventional filtering methods.

Funding

Not specified in the provided text.

Citation

@article{Zhou2025novel,
  author = {Zhou, Hao and Li, Liexing and Hu, Hanwen and Xue, Haoran and Li, Hong and Liu, Yuelong and Luo, Zhicai},
  title = {A novel Hankel spectrum analysis filtering for reducing North-South stripes in GRACE gravity solutions},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134602},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134602}
}

Generated by BiblioAssistant using gemini-2.5-flash (Google API)