One month in advance prediction of air temperature from Reanalysis data with Explainable Artificial Intelligence techniques

Áreas de investigación:

• Applications - Weather forecasting

Año:

2023

Tipo de publicación:

Artículo

Palabras clave:

Air temperature, Long-term air temperature prediction, Climatology, XAI, Neural networks

Autores:

• Gómez-Orellana, Antonio Manuel
• Guijo-Rubio, David
• Pérez-Aracil, Jorge
• Gutiérrez, Pedro Antonio
• Salcedo-Sanz, Sancho
• Hervás-Martínez, César

Journal:

Atmospheric Research

Volumen:

284

Número:

106608

ISSN:

0169-8095

BibTex:

@article{XAITemperaturasAt2023, author = "Antonio Manuel G{\'o}mez-Orellana and David Guijo-Rubio and Jorge P{\'e}rez-Aracil and Pedro Antonio Guti{\'e}rrez and Sancho Salcedo-Sanz and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "In this paper we have tackled the problem of long-term air temperature prediction with eXplainable Artificial Intelligence (XAI) models. Specifically, we have evaluated the performance of an Artificial Neural Network (ANN) architecture with sigmoidal neurons in the hidden layer, trained by means of an evolutionary algorithm (Evolutionary ANNs, EANNs). This XAI model architecture (XAI-EANN) has been applied to the long-term air temperature prediction at different sub-regions of the South of the Iberian Peninsula. In this case, the average August air temperature has been predicted from ERA5 Reanalysis data variables, obtaining good predictions skills and explainable models in terms of the input climatological variables considered. A cluster analysis has been first carried out in terms of the average air temperature in the zone, in such a way that a number of sub-regions with different air temperature behaviour have been defined. The proposed XAI-EANN model architecture has been applied to each of the defined sub-regions, in order to find significant differences among them, which can be explained with the XAI-EANN models obtained. Finally, a comprehensive comparison against some state-of-the-art techniques has also been carried out, concluding that there are statistically significant differences in terms of accuracy in favour of the proposed XAI-EANN model, which also benefits from being an XAI model.", awards = "JCR(2022): 5.5, position: 18/94 (Q1), category: METEOROLOGY {\&}amp; ATMOSPHERIC SCIENCES.", comments = "JCR(2022): 5.5, position: 18/94 (Q1), category: METEOROLOGY {\&}amp; ATMOSPHERIC SCIENCES.", doi = "10.1016/j.atmosres.2023.106608", issn = "0169-8095", journal = "Atmospheric Research", keywords = "Air temperature, Long-term air temperature prediction, Climatology, XAI, Neural networks", note = "JCR(2022): 5.5, position: 18/94 (Q1), category: METEOROLOGY {\&}amp; ATMOSPHERIC SCIENCES.", number = "106608", title = "{O}ne month in advance prediction of air temperature from {R}eanalysis data with {E}xplainable {A}rtificial {I}ntelligence techniques", url = "doi.org/10.1016/j.atmosres.2023.106608", volume = "284", year = "2023", }

Nota:

JCR(2022): 5.5, position: 18/94 (Q1), category: METEOROLOGY & ATMOSPHERIC SCIENCES.

Abstract:

In this paper we have tackled the problem of long-term air temperature prediction with eXplainable Artificial Intelligence (XAI) models. Specifically, we have evaluated the performance of an Artificial Neural Network (ANN) architecture with sigmoidal neurons in the hidden layer, trained by means of an evolutionary algorithm (Evolutionary ANNs, EANNs). This XAI model architecture (XAI-EANN) has been applied to the long-term air temperature prediction at different sub-regions of the South of the Iberian Peninsula. In this case, the average August air temperature has been predicted from ERA5 Reanalysis data variables, obtaining good predictions skills and explainable models in terms of the input climatological variables considered. A cluster analysis has been first carried out in terms of the average air temperature in the zone, in such a way that a number of sub-regions with different air temperature behaviour have been defined. The proposed XAI-EANN model architecture has been applied to each of the defined sub-regions, in order to find significant differences among them, which can be explained with the XAI-EANN models obtained. Finally, a comprehensive comparison against some state-of-the-art techniques has also been carried out, concluding that there are statistically significant differences in terms of accuracy in favour of the proposed XAI-EANN model, which also benefits from being an XAI model.

Comentarios:

JCR(2022): 5.5, position: 18/94 (Q1), category: METEOROLOGY & ATMOSPHERIC SCIENCES.