AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models
Accurately forecasting reference evapotranspiration (ET<sub>0</sub>) values is crucial to improve crop irrigation scheduling, allowing anticipated planning decisions and optimized water resource management and agricultural production. In this work, a recent state-of-the-art architecture...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Agronomy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4395/12/3/656 |
| _version_ | 1797447371924176896 |
|---|---|
| author | Juan Antonio Bellido-Jiménez Javier Estévez Joaquin Vanschoren Amanda Penélope García-Marín |
| author_facet | Juan Antonio Bellido-Jiménez Javier Estévez Joaquin Vanschoren Amanda Penélope García-Marín |
| author_sort | Juan Antonio Bellido-Jiménez |
| collection | DOAJ |
| description | Accurately forecasting reference evapotranspiration (ET<sub>0</sub>) values is crucial to improve crop irrigation scheduling, allowing anticipated planning decisions and optimized water resource management and agricultural production. In this work, a recent state-of-the-art architecture has been adapted and deployed for multivariate input time series forecasting (transformers) using past values of ET<sub>0</sub> and temperature-based parameters (28 input configurations) to forecast daily ET<sub>0</sub> up to a week (1 to 7 days). Additionally, it has been compared to standard machine learning models such as multilayer perceptron (MLP), random forest (RF), support vector machine (SVM), extreme learning machine (ELM), convolutional neural network (CNN), long short-term memory (LSTM), and two baselines (historical monthly mean value and a moving average of the previous seven days) in five locations with different geo-climatic characteristics in the Andalusian region, Southern Spain. In general, machine learning models significantly outperformed the baselines. Furthermore, the accuracy dramatically dropped when forecasting ET<sub>0</sub> for any horizon longer than three days. SVM, ELM, and RF using configurations I, III, IV, and IX outperformed, on average, the rest of the configurations in most cases. The best NSE values ranged from 0.934 in Córdoba to 0.869 in Tabernas, using SVM. The best RMSE, on average, ranged from 0.704 mm/day for Málaga to 0.883 mm/day for Conil using RF. In terms of MBE, most models and cases performed very accurately, with a total average performance of 0.011 mm/day. We found a relationship in performance regarding the aridity index and the distance to the sea. The higher the aridity index at inland locations, the better results were obtained in forecasts. On the other hand, for coastal sites, the higher the aridity index, the higher the error. Due to the good performance and the availability as an open-source repository of these models, they can be used to accurately forecast ET<sub>0</sub> in different geo-climatic conditions, helping to increase efficiency in tasks of great agronomic importance, especially in areas with low rainfall or where water resources are limiting for the development of crops. |
| first_indexed | 2024-03-09T13:55:12Z |
| format | Article |
| id | doaj.art-93c0729951b04dc489b754703f4ea4f1 |
| institution | Directory Open Access Journal |
| issn | 2073-4395 |
| language | English |
| last_indexed | 2024-03-09T13:55:12Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agronomy |
| spelling | doaj.art-93c0729951b04dc489b754703f4ea4f12023-11-30T20:44:38ZengMDPI AGAgronomy2073-43952022-03-0112365610.3390/agronomy12030656AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based ModelsJuan Antonio Bellido-Jiménez0Javier Estévez1Joaquin Vanschoren2Amanda Penélope García-Marín3Projects Engineering Area, Department of Rural Engineering, Civil Constructions and Engineering Projects, University of Córdoba, 14071 Córdoba, SpainProjects Engineering Area, Department of Rural Engineering, Civil Constructions and Engineering Projects, University of Córdoba, 14071 Córdoba, SpainData Mining Group, Department of Mathematics and Computer Science, Eindhoven University of Technology, 5612 Eindhoven, The NetherlandsProjects Engineering Area, Department of Rural Engineering, Civil Constructions and Engineering Projects, University of Córdoba, 14071 Córdoba, SpainAccurately forecasting reference evapotranspiration (ET<sub>0</sub>) values is crucial to improve crop irrigation scheduling, allowing anticipated planning decisions and optimized water resource management and agricultural production. In this work, a recent state-of-the-art architecture has been adapted and deployed for multivariate input time series forecasting (transformers) using past values of ET<sub>0</sub> and temperature-based parameters (28 input configurations) to forecast daily ET<sub>0</sub> up to a week (1 to 7 days). Additionally, it has been compared to standard machine learning models such as multilayer perceptron (MLP), random forest (RF), support vector machine (SVM), extreme learning machine (ELM), convolutional neural network (CNN), long short-term memory (LSTM), and two baselines (historical monthly mean value and a moving average of the previous seven days) in five locations with different geo-climatic characteristics in the Andalusian region, Southern Spain. In general, machine learning models significantly outperformed the baselines. Furthermore, the accuracy dramatically dropped when forecasting ET<sub>0</sub> for any horizon longer than three days. SVM, ELM, and RF using configurations I, III, IV, and IX outperformed, on average, the rest of the configurations in most cases. The best NSE values ranged from 0.934 in Córdoba to 0.869 in Tabernas, using SVM. The best RMSE, on average, ranged from 0.704 mm/day for Málaga to 0.883 mm/day for Conil using RF. In terms of MBE, most models and cases performed very accurately, with a total average performance of 0.011 mm/day. We found a relationship in performance regarding the aridity index and the distance to the sea. The higher the aridity index at inland locations, the better results were obtained in forecasts. On the other hand, for coastal sites, the higher the aridity index, the higher the error. Due to the good performance and the availability as an open-source repository of these models, they can be used to accurately forecast ET<sub>0</sub> in different geo-climatic conditions, helping to increase efficiency in tasks of great agronomic importance, especially in areas with low rainfall or where water resources are limiting for the development of crops.https://www.mdpi.com/2073-4395/12/3/656machine learningtransformersneural networkssupport vector machinereference evapotranspirationforecasting |
| spellingShingle | Juan Antonio Bellido-Jiménez Javier Estévez Joaquin Vanschoren Amanda Penélope García-Marín AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models Agronomy machine learning transformers neural networks support vector machine reference evapotranspiration forecasting |
| title | AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models |
| title_full | AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models |
| title_fullStr | AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models |
| title_full_unstemmed | AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models |
| title_short | AgroML: An Open-Source Repository to Forecast Reference Evapotranspiration in Different Geo-Climatic Conditions Using Machine Learning and Transformer-Based Models |
| title_sort | agroml an open source repository to forecast reference evapotranspiration in different geo climatic conditions using machine learning and transformer based models |
| topic | machine learning transformers neural networks support vector machine reference evapotranspiration forecasting |
| url | https://www.mdpi.com/2073-4395/12/3/656 |
| work_keys_str_mv | AT juanantoniobellidojimenez agromlanopensourcerepositorytoforecastreferenceevapotranspirationindifferentgeoclimaticconditionsusingmachinelearningandtransformerbasedmodels AT javierestevez agromlanopensourcerepositorytoforecastreferenceevapotranspirationindifferentgeoclimaticconditionsusingmachinelearningandtransformerbasedmodels AT joaquinvanschoren agromlanopensourcerepositorytoforecastreferenceevapotranspirationindifferentgeoclimaticconditionsusingmachinelearningandtransformerbasedmodels AT amandapenelopegarciamarin agromlanopensourcerepositorytoforecastreferenceevapotranspirationindifferentgeoclimaticconditionsusingmachinelearningandtransformerbasedmodels |