Mehr et al. (2026) Improving Meteorological Drought Forecasting Through a CPO ‐Tuned VMD ‐Liquid Neural Network

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

Identification

• Journal: International Journal of Climatology
• Year: 2026
• Date: 2026-09-13
• Authors: Ali Danandeh Mehr, Abdelkader T. Ahmed, Mir Jafar Sadegh Safari, Enrico Creaco, S. Adarsh
• DOI: 10.1002/joc.70590

Research Groups

Not specified in the provided text.

Short Summary

The study develops a hybrid CPO-VMD-LNN model for one-month-ahead meteorological drought forecasting, which significantly outperforms SARIMA, LSTM, and standalone LNN models in the Urmia Lake Basin.

Objective

• To develop a novel hybrid forecasting framework that optimizes the decomposition of the Standardized Precipitation-Evapotranspiration Index (SPEI) to improve the accuracy of one-month-ahead meteorological drought predictions.

Study Configuration

• Spatial Scale: Eastern and western regions of the Urmia Lake Basin, Iran.
• Temporal Scale: One-month-ahead forecasting.

Methodology and Data

• Models used: Crested Porcupine Optimizer (CPO), Variational Mode Decomposition (VMD), Liquid Neural Network (LNN), SARIMA, and Long Short-Term Memory (LSTM).
• Data sources: Standardized Precipitation-Evapotranspiration Index (SPEI) time series.
• Process: CPO is used to tune VMD parameters for denoising SPEI; mutual information is applied to select the most informative lagged decomposed vectors as inputs for the LNN regression model.

Main Results

• The proposed CPO-VMD-LNN hybrid model consistently outperformed the benchmark models (SARIMA, LSTM, and LNN) in both study regions.
• The model achieved high precision with Root Mean Square Error (RMSE) values below 0.25.

Contributions

• Introduction of a hybrid architecture combining a nature-inspired optimizer (CPO) for signal decomposition (VMD) and a liquid neural network (LNN) for time-series regression.
• Implementation of mutual information for feature selection to create more parsimonious and efficient drought forecasting models.

Funding

Not specified in the provided text.

Citation

@article{Mehr2026Improving,
  author = {Mehr, Ali Danandeh and Ahmed, Abdelkader T. and Safari, Mir Jafar Sadegh and Creaco, Enrico and Adarsh, S.},
  title = {Improving Meteorological Drought Forecasting Through a CPO ‐Tuned VMD ‐Liquid Neural Network},
  journal = {International Journal of Climatology},
  year = {2026},
  doi = {10.1002/joc.70590},
  url = {https://doi.org/10.1002/joc.70590}
}