Subseasonal Rainfall Forecasting Skill of S2S-ECMWF and Its Potential for Bias Correction Using ANN over the Central Highlands of Vietnam During the Transition from the Dry to the Rainy Season
Main Article Content
Abstract
This study evaluates the subseasonal rainfall forecasting skill of the S2S-ECMWF model over the Central Highlands of Vietnam during the transition from the dry to the rainy season (March–April) for the period 2003–2022. It also examines the potential of applying an Artificial Neural Network (ANN) for correcting biases in the raw forecast products. The results show that the original S2S-ECMWF forecasts capture some broad features of rainfall occurrence, particularly for light rainfall events and short lead times. However, their quantitative forecasting skill remains limited, as indicated by large errors, low or even negative correlations, and a clear decline in skill at higher rainfall thresholds. The model’s ability to detect moderate and heavy rainfall events decreases substantially, and its discrimination between rainfall and non-rainfall events above the selected thresholds is very weak. After ANN-based bias correction, forecast performance improves markedly, especially during weeks 1–2. The corrected forecasts also exhibit much stronger discriminatory skill, particularly at short lead times. Although forecast skill still decreases with increasing lead time and heavy rainfall prediction remains challenging, the results demonstrate the potential of ANN-based post-processing for improving subseasonal rainfall forecasts over the Central Highlands of Vietnam.
References
[2] P. T. Ha, Seasonal and Sub-seasonal Prediction of the Rainy Season Onset Date in Vietnam Based on Numerical Model Products, Doctoral Dissertation, VNU University of Science, Hanoi, Vietnam, 2022 (in Vietnamese).
[3] E. Olaniyan, E. A. Adefisan, F. Oni, E. Afiesimama, A. A. Balogun, K. A. Lawal, Evaluation of the ECMWF Sub-seasonal to Seasonal Precipitation Forecasts During the Peak of West Africa Monsoon in Nigeria, Frontiers in Environmental Science, Vol. 6, No. 4, 2018, https://doi.org/10.3389/fenvs.2018.0 0004.
[4] F. M. De Andrade, M. P. Young, D. MacLeod, L. C. Hirons, S. J. Woolnough, E. Black, Subseasonal Precipitation Prediction for Africa: Forecast Evaluation and Sources of Predictability, Weather and Forecasting, Vol. 36, No. 1, 2021, pp. 265-284, https://doi.org/10.1175/WAF-D-20-0054.1.
[5] F. Vitart, Evolution of ECMWF Sub-seasonal Forecast Skill Scores Over the Past 10 Years, ECMWF Technical Memorandum, European Centre for Medium-Range Weather Forecasts, Reading, United Kingdom, 2013.
[6] E. W. Benet, M. Pyrina, B. J. Esteve, E. Fraenkel, J. Cohen, D. I. V. Domeisen, Subseasonal Prediction of Central European Summer Heatwaves with Linear and Random Forest Machine Learning Models, Artificial Intelligence for the Earth Systems, Vol. 2, No. 2, 2023,
pp. e220038,
https://doi.org/10.1175/AIES-D-22-0038.1.
[7] X. Liu, S. Yang, A. Kumar, S. Weaver, X. Jiang, Diagnostics of Subseasonal Prediction Biases of the Asian Summer Monsoon by the NCEP Climate Forecast System, Climate Dynamics, Vol. 41,
No. 5-6, 2013, pp. 1453-1474, https://doi.org/10.1007/ s00 382-012-1553-3.
[8] S. He, X. Li, T. DelSole, P. Ravikumar, A. Banerjee, Sub-Seasonal Climate Forecasting Via Machine Learning: Challenges, Analysis, and Advances, Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35, No. 1, 2021,
pp. 169-177.
[9] R. P. Worsnop, M. Scheuerer, T. M. Hamill, T. A. Smith, J. Schlör, RUFCO: A Deep Learning Framework to Postprocess Subseasonal Precipitation Accumulation Forecasts, Artificial Intelligence for the Earth Systems, 2024, https://doi.org/10.1175/AIES-D-24-0020.1.
[10] C. K. Mengist, K. H. Seo, Improving Subseasonal to Seasonal Global 2-m Temperature and Precipitation Forecasts Using a Deep Learning-Based Postprocessing Approach, Monthly Weather Review, Vol. 154, No. 3, 2026, pp. 391-405, https://doi.org/10.1175/MWR-D-25-0179.1.
[11] D. H. Yen, Assessment of Rainfall Forecasting Skill of the ECMWF Model over Vietnam. Master's Thesis, VNU University of Science, Hanoi, Vietnam, 2023 (in Vietnamese).
[12] Q. Tran-Anh, T. Ngo-Duc, E. Espagne, L. Trinh-Tuan, A 10-km CMIP6 Downscaled Dataset of Temperature and Precipitation for Historical and Future Vietnam Climate, Scientific Data, Vol. 10, 2023, pp. 257.
[13] X. T. Nguyen, D. T. Nguyen, H. Kamimera, T. L. Trinh, J. Matsumoto, T. Inoue, V. T. Phan, The Vietnam Gridded Precipitation (VnGP) Data Set: Construction and Validation, Scientific Online Letters on the Atmosphere, Vol. 12, 2016,
pp. 291-295.
[14] A. Q. Tran, D. T. Nguyen, E. Espagne, T. L. Trinh, A High-Resolution Projected Climate Dataset for Vietnam: Construction and Preliminary Application in Assessing Future Change, Journal of Water and Climate Change, Vol. 13, No. 9, 2022, pp. 3379-3396.
[15] V. T. Le, V. T. Phan, Recent Changes in Temperature Extremes During Cold Season Over Northern Vietnam and Teleconnections With ENSO, IOD, NAO and AO, International Journal of Climatology, Vol. 46, No. 2, 2025, pp. e70190.
[16] M. Kubat, An Introduction to Machine Learning (2nd ed.), Springer International Publishing, Cham, Switzerland, 2017.
[17] I. T. Jolliffe, D. B. Stephenson (Eds.), Forecast Verification: A Practitioner's Guide in Atmospheric Science (2nd ed.), Wiley-Blackwell, Chichester, United Kingdom, 2012.