Long-term temperature prediction with hybrid autoencoder algorithms

Research areas:

• Applications - Weather forecasting

Year:

2024

Type of Publication:

Article

Keywords:

Autoencoder, Temperature prediction, Hybrid models, Heatwave

Authors:

• Pérez-Aracil, Jorge
• Fister, D.
• Marina, C. M.
• Peláez-Rodríguez, C.
• Cornejo-Bueno, L.
• Gutiérrez, Pedro Antonio
• Giuliani, M.
• Castelleti, A.
• Salcedo-Sanz, Sancho

Journal:

Applied Computing and Geosciences

Volume:

23

Pages:

100185

Month:

September

ISSN:

2590-1974

BibTex:

@article{JorgeAE2024ACG, author = "Jorge P{\'e}rez-Aracil and D. Fister and C.M. Marina and C. Pel{\'a}ez-Rodr{\'i}guez and L. Cornejo-Bueno and Pedro Antonio Guti{\'e}rrez and M. Giuliani and A. Castelleti and Sancho Salcedo-Sanz", abstract = "This paper proposes two hybrid approaches based on Autoencoders (AEs) for long-term temperature prediction. The first algorithm comprises an AE trained to learn temperature patterns, which is then linked to a second AE, used to detect possible anomalies and provide a final temperature prediction. The second proposed approach involves training an AE and then using the resulting latent space as input of a neural network, which will provide the final prediction output. Both approaches are tested in long-term air temperature prediction in European cities: seven European locations where major heat waves occurred have been considered. The long-term temperature prediction for the entire year of the heatwave events has been analysed. Results show that the proposed approaches can obtain accurate long-term (up to 4 weeks) temperature prediction, improving Persistence and Climatology in the benchmark models compared. In heatwave periods, where the persistence of the temperature is extremely high, our approach beat the persistence operator in three locations and works similarly in the rest of the cases, showing the potential of this AE-based method for long-term temperature prediction.", awards = "JCR (2023): 2.6, Position: 98/253 (Q2), Category: GEOSCIENCES, MULTIDISCIPLINARY", comments = "JCR (2023): 2.6, Position: 98/253 (Q2), Category: GEOSCIENCES, MULTIDISCIPLINARY", doi = "10.1016/j.acags.2024.100185", issn = "2590-1974", journal = "Applied Computing and Geosciences", keywords = "Autoencoder, Temperature prediction, Hybrid models, Heatwave", month = "September", note = "JCR (2023): 2.6, Position: 98/253 (Q2), Category: GEOSCIENCES, MULTIDISCIPLINARY", pages = "100185", title = "{L}ong-term temperature prediction with hybrid autoencoder algorithms", url = "doi.org/10.1016/j.acags.2024.100185", volume = "23", year = "2024", }

Note:

JCR (2023): 2.6, Position: 98/253 (Q2), Category: GEOSCIENCES, MULTIDISCIPLINARY

Abstract:

This paper proposes two hybrid approaches based on Autoencoders (AEs) for long-term temperature prediction. The first algorithm comprises an AE trained to learn temperature patterns, which is then linked to a second AE, used to detect possible anomalies and provide a final temperature prediction. The second proposed approach involves training an AE and then using the resulting latent space as input of a neural network, which will provide the final prediction output. Both approaches are tested in long-term air temperature prediction in European cities: seven European locations where major heat waves occurred have been considered. The long-term temperature prediction for the entire year of the heatwave events has been analysed. Results show that the proposed approaches can obtain accurate long-term (up to 4 weeks) temperature prediction, improving Persistence and Climatology in the benchmark models compared. In heatwave periods, where the persistence of the temperature is extremely high, our approach beat the persistence operator in three locations and works similarly in the rest of the cases, showing the potential of this AE-based method for long-term temperature prediction.

Comments:

JCR (2023): 2.6, Position: 98/253 (Q2), Category: GEOSCIENCES, MULTIDISCIPLINARY