Work

Reconstruction of high-resolution subsurface density and velocity in the northwest pacific via iSQG method using multi-source satellite observations

Year: 2026
Type: article
Abstract: This study employs the interior plus Surface Quasi-Geostrophic (iSQG) method to reconstruct high-resolution three-dimensional subsurface density and velocity fields using multi-source satellite data in the East China Sea Kuroshio Zone (ECS-KZ) and an Eddy-Rich Zone (ERZ) of the Northwest Pacific from 2014 to 2023. Validated against in-situ hydrographic and current data, the reconstruction exhibits distinct spatiotemporal patterns modulated by seasons, eddy polarities, and dynamical regimes. Density reconstruction achieves higher accuracy in cold seasons with depth‑averaged density anomaly root mean square errors (RMSE) of 0.20 kg m⁻³ in the ECS‑KZ and 0.23 kg m⁻³ in the ERZ, versus 0.24 kg m⁻³ and 0.26 kg m⁻³ in warm seasons. Conversely, velocity reconstruction performs better in warm seasons, yielding eddy kinetic energy (EKE) RMSEs of 0.05 m² s⁻² in the ECS‑KZ and 0.03 m² s⁻² in the ERZ compared to 0.08 and 0.05 m² s⁻² in cold seasons—a contrast driven by seasonal trade-offs between stratification and ageostrophic motions. Furthermore, cyclonic eddies (CE) are reconstructed with higher fidelity than anticyclonic eddies (AE). Specifically, the CE density anomaly RMSE is 0.20 kg m⁻³ in both regions, whereas AE values reach 0.27 kg m⁻³ in the ECS‑KZ and 0.28 kg m⁻³ in the ERZ. Similarly, CE EKE RMSE is 0.04 and 0.03 m² s⁻² compared to 0.09 and 0.04 m² s⁻² for AEs, because CEs better align with the low-order vertical modes assumed by the iSQG method. Regionally, while density errors are comparable, velocity errors are markedly larger in the ECS-KZ due to stronger nonlinear processes inherent to this western boundary current regime. These findings reveal the practical boundaries of iSQG applicability and inform satellite-based ocean structure reconstruction.
Language: English
Subfield: Oceanography
Open Access status: gold
Estimated APC paid: $2,950
OpenAlex ID: W7166201443
FWCI:
Cites: 28
Cited by:
Related to: