Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt
Forecasting crop yields is becoming increasingly important under the current context in which food security needs to be ensured despite the challenges brought by climate change, an expanding world population accompanied by rising incomes, increasing soil erosion, and decreasing water resources. Temp...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2020-01-01
|
| Series: | Environmental Research Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/1748-9326/ab68ac |
| _version_ | 1797747605068840960 |
|---|---|
| author | Aleksandra Wolanin Gonzalo Mateo-García Gustau Camps-Valls Luis Gómez-Chova Michele Meroni Gregory Duveiller You Liangzhi Luis Guanter |
| author_facet | Aleksandra Wolanin Gonzalo Mateo-García Gustau Camps-Valls Luis Gómez-Chova Michele Meroni Gregory Duveiller You Liangzhi Luis Guanter |
| author_sort | Aleksandra Wolanin |
| collection | DOAJ |
| description | Forecasting crop yields is becoming increasingly important under the current context in which food security needs to be ensured despite the challenges brought by climate change, an expanding world population accompanied by rising incomes, increasing soil erosion, and decreasing water resources. Temperature, radiation, water availability and other environmental conditions influence crop growth, development, and final grain yield in a complex nonlinear manner. Machine learning (ML) techniques, and deep learning (DL) methods in particular, can account for such nonlinear relations between yield and its covariates. However, they typically lack transparency and interpretability, since the way the predictions are derived is not directly evident. Yet, in the context of yield forecasting, understanding which are the underlying factors behind both a predicted loss or gain is of great relevance. Here, we explore how to benefit from the increased predictive performance of DL methods while maintaining the ability to interpret how the models achieve their results. To do so, we applied a deep neural network to multivariate time series of vegetation and meteorological data to estimate the wheat yield in the Indian Wheat Belt. Then, we visualized and analyzed the features and yield drivers learned by the model with the use of regression activation maps. The DL model outperformed other tested models (ridge regression and random forest) and facilitated the interpretation of variables and processes that lead to yield variability. The learned features were mostly related to the length of the growing season, and temperature and light conditions during this time. For example, our results showed that high yields in 2012 were associated with low temperatures accompanied by sunny conditions during the growing period. The proposed methodology can be used for other crops and regions in order to facilitate application of DL models in agriculture. |
| first_indexed | 2024-03-12T15:53:00Z |
| format | Article |
| id | doaj.art-8f4d738702d1491ea93abbb2a5619807 |
| institution | Directory Open Access Journal |
| issn | 1748-9326 |
| language | English |
| last_indexed | 2024-03-12T15:53:00Z |
| publishDate | 2020-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Environmental Research Letters |
| spelling | doaj.art-8f4d738702d1491ea93abbb2a56198072023-08-09T15:04:21ZengIOP PublishingEnvironmental Research Letters1748-93262020-01-0115202401910.1088/1748-9326/ab68acEstimating and understanding crop yields with explainable deep learning in the Indian Wheat BeltAleksandra Wolanin0https://orcid.org/0000-0002-9029-6911Gonzalo Mateo-García1Gustau Camps-Valls2Luis Gómez-Chova3Michele Meroni4Gregory Duveiller5https://orcid.org/0000-0002-6471-8404You Liangzhi6Luis Guanter7Remote Sensing and Geoinformatics Section, GFZ German Research Centre for Geosciences, Helmholtz-Centre, Potsdam, GermanyImage Processing Laboratory, Universitat de València , València, SpainImage Processing Laboratory, Universitat de València , València, SpainImage Processing Laboratory, Universitat de València , València, SpainEuropean Commission, Joint Research Centre (JRC), Ispra, ItalyEuropean Commission, Joint Research Centre (JRC), Ispra, ItalyEnvironment and Production Technology Division, The International Food Policy Research Institute (IFPRI), Washington, D.C., United States of AmericaCentro de Tecnologías Físicas, Universitat Politècnica de València , València, SpainForecasting crop yields is becoming increasingly important under the current context in which food security needs to be ensured despite the challenges brought by climate change, an expanding world population accompanied by rising incomes, increasing soil erosion, and decreasing water resources. Temperature, radiation, water availability and other environmental conditions influence crop growth, development, and final grain yield in a complex nonlinear manner. Machine learning (ML) techniques, and deep learning (DL) methods in particular, can account for such nonlinear relations between yield and its covariates. However, they typically lack transparency and interpretability, since the way the predictions are derived is not directly evident. Yet, in the context of yield forecasting, understanding which are the underlying factors behind both a predicted loss or gain is of great relevance. Here, we explore how to benefit from the increased predictive performance of DL methods while maintaining the ability to interpret how the models achieve their results. To do so, we applied a deep neural network to multivariate time series of vegetation and meteorological data to estimate the wheat yield in the Indian Wheat Belt. Then, we visualized and analyzed the features and yield drivers learned by the model with the use of regression activation maps. The DL model outperformed other tested models (ridge regression and random forest) and facilitated the interpretation of variables and processes that lead to yield variability. The learned features were mostly related to the length of the growing season, and temperature and light conditions during this time. For example, our results showed that high yields in 2012 were associated with low temperatures accompanied by sunny conditions during the growing period. The proposed methodology can be used for other crops and regions in order to facilitate application of DL models in agriculture.https://doi.org/10.1088/1748-9326/ab68acwheat yieldIndian Wheat Beltfood securityremote sensingexplainable artificial intelligence (XAI)deep learning (DL) |
| spellingShingle | Aleksandra Wolanin Gonzalo Mateo-García Gustau Camps-Valls Luis Gómez-Chova Michele Meroni Gregory Duveiller You Liangzhi Luis Guanter Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt Environmental Research Letters wheat yield Indian Wheat Belt food security remote sensing explainable artificial intelligence (XAI) deep learning (DL) |
| title | Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt |
| title_full | Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt |
| title_fullStr | Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt |
| title_full_unstemmed | Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt |
| title_short | Estimating and understanding crop yields with explainable deep learning in the Indian Wheat Belt |
| title_sort | estimating and understanding crop yields with explainable deep learning in the indian wheat belt |
| topic | wheat yield Indian Wheat Belt food security remote sensing explainable artificial intelligence (XAI) deep learning (DL) |
| url | https://doi.org/10.1088/1748-9326/ab68ac |
| work_keys_str_mv | AT aleksandrawolanin estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT gonzalomateogarcia estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT gustaucampsvalls estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT luisgomezchova estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT michelemeroni estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT gregoryduveiller estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT youliangzhi estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt AT luisguanter estimatingandunderstandingcropyieldswithexplainabledeeplearningintheindianwheatbelt |