Sun et al. (2025) Reducing Evapotranspiration Simulation and Forecast Uncertainties Due To Initial and Model Errors Over the Tibetan Plateau

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Journal of Geophysical Research Atmospheres
• Year: 2025
• Date: 2025-12-24
• Authors: Guodong Sun, Mu Mu, Q. M. Zhang, Qinglong You
• DOI: 10.1029/2025jd043736

Research Groups

Not specified in the abstract.

Short Summary

This study introduces a combined approach using minimization, data assimilation, and the conditional nonlinear optimal parameter perturbation ensemble prediction (CNOP-PEP) method to reduce uncertainties in evapotranspiration (ET) simulations and predictions over the Tibetan Plateau. The proposed method significantly improves ET ensemble prediction accuracy compared to traditional methods by optimizing initial conditions and perturbing model parameters.

Objective

• To reduce the impact of initial errors on evapotranspiration (ET) simulation uncertainties over the Tibetan Plateau using a minimization technique and a data assimilation method.
• To evaluate the simulation and forecast performances of ET with imperfect model physical parameters using the CNOP-PEP method, building upon the reduction of initial errors.

Study Configuration

• Spatial Scale: Tibetan Plateau
• Temporal Scale: Annual or long-term averages, inferred from error units (mm/year).

Methodology and Data

• Models used: Minimization technique, data assimilation method, conditional nonlinear optimal parameter perturbation ensemble prediction (CNOP-PEP) method, one-at-a-time (OAT) method, stochastically perturbed parametrization (SPP) method.
• Data sources: Not specified in the abstract.

Main Results

• The minimization technique and data assimilation method effectively reduce uncertainties in ET over the Tibetan Plateau caused by initial value errors.
• Ensemble prediction of ET using the CNOP-PEP method with assimilated initial conditions performs better than without assimilated initial conditions.
• The ensemble prediction of ET using the CNOP-PEP method (17.3% improvement, likely relative to a baseline not explicitly stated but implied by comparison) is superior to traditional methods like the one-at-a-time (OAT) method (2.0%) and the stochastically perturbed parametrization (SPP) method (10.6%).
• Absolute errors using anomaly-based metrics are 71.80 mm/year for CNOP-PEP, 101.06 mm/year for OAT, and 105.52 mm/year for SPP.

Contributions

• First application of a combined minimization technique and data assimilation method to reduce initial error impact on ET simulation uncertainties over the Tibetan Plateau.
• Introduction and demonstration of the CNOP-PEP method for ensemble prediction of ET, integrating perturbed parameters with optimized and assimilated initial conditions.
• Quantitative evidence showing the superior performance of the CNOP-PEP method over traditional ensemble prediction methods (OAT, SPP) for ET forecasting on the Tibetan Plateau.

Funding

Not specified in the abstract.

Citation

@article{Sun2025Reducing,
  author = {Sun, Guodong and Mu, Mu and Zhang, Q. M. and You, Qinglong},
  title = {Reducing Evapotranspiration Simulation and Forecast Uncertainties Due To Initial and Model Errors Over the Tibetan Plateau},
  journal = {Journal of Geophysical Research Atmospheres},
  year = {2025},
  doi = {10.1029/2025jd043736},
  url = {https://doi.org/10.1029/2025jd043736}
}