期刊
REMOTE SENSING
卷 14, 期 19, 页码 -出版社
MDPI
DOI: 10.3390/rs14195045
关键词
crop yield prediction; remote sensing; deep learning; feature importance; attention
类别
资金
- National Natural Science Foundation of China [31861143015]
- Strategic Priority Research Program of the Chinese Academy of Sciences [XDA23100202]
- Science and Technology Planning Project of Hebei Academy of Sciences of China [22A03]
In this study, the researchers proposed a transformer-based model, Informer, to predict rice yield in the Indian Indo-Gangetic Plains. By integrating time-series satellite data, environmental variables, and rice yield records, Informer achieved better performance than other models for end-of-season prediction. The model was also able to achieve stable performances for within-season prediction after late September.
As the second largest rice producer, India contributes about 20% of the world's rice production. Timely, accurate, and reliable rice yield prediction in India is crucial for global food security and health issues. Deep learning models have achieved excellent performances in predicting crop yield. However, the interpretation of deep learning models is still rare. In this study, we proposed a transformer-based model, Informer, to predict rice yield across the Indian Indo-Gangetic Plains by integrating time-series satellite data, environmental variables, and rice yield records from 2001 to 2016. The results showed that Informer had better performance (R-2 = 0.81, RMSE = 0.41 t/ha) than four other machine learning and deep learning models for end-of-season prediction. For within-season prediction, the Informer model could achieve stable performances (R-2 approximate to 0.78) after late September, which indicated that the optimal prediction could be achieved 2 months before rice maturity. In addition, we interpreted the prediction models by evaluating the input feature importance and analyzing hidden features. The evaluation of feature importance indicated that NIRV was the most critical factor, while intervals 6 (mid-August) and 12 (mid-November) were the key periods for rice yield prediction. The hidden feature analysis demonstrated that the attention-based long short-term memory (AtLSTM) model accumulated the information of each growth period, while the Informer model focused on the information around intervals 5 to 6 (August) and 11 to 12 (November). Our findings provided a reliable and simple framework for crop yield prediction and a new perspective for explaining the internal mechanism of deep learning models.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据