Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data

Timely and accurate crop yield estimation is crucial for managing crops, trade, and food security. The combination of remote sensing technology with machine learning methods is increasingly popular for global yield prediction. However, traditional machine learning methods rely on data correlation ra...

Full description

Saved in:
Bibliographic Details
Main Authors: Fumin Wang, Jiale Li, Dailiang Peng, Qiuxiang Yi, Xiaoyang Zhang, Jueyi Zheng, Siting Chen
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10614767/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846118349224804352
author Fumin Wang
Jiale Li
Dailiang Peng
Qiuxiang Yi
Xiaoyang Zhang
Jueyi Zheng
Siting Chen
author_facet Fumin Wang
Jiale Li
Dailiang Peng
Qiuxiang Yi
Xiaoyang Zhang
Jueyi Zheng
Siting Chen
author_sort Fumin Wang
collection DOAJ
description Timely and accurate crop yield estimation is crucial for managing crops, trade, and food security. The combination of remote sensing technology with machine learning methods is increasingly popular for global yield prediction. However, traditional machine learning methods rely on data correlation rather than causality, leading to poor interpretability. To address this issue, we propose a novel approach that combines a structural causal model (SCM) with deep learning to develop a causal graph attention network model (SCM-GAT) based on causal relationships for soybean yield prediction at the county level of 10 major soybean-producing states in the United States. The SCM-GAT model considers not only conventional vegetation indices and weather variables but also the causal relationships between variables as inputs. Using independent validation and five-fold cross-validation strategies, our results show that the predictive performance of the SCM-GAT model is not only superior to traditional prediction models, such as selection operator regression (LASSO), random forest, but also superior to the common deep learning models based on correlation relationships, such as long short-term memory network, and transformer. It has also proved to be more robust than other models under extreme weather events. In addition, we identify branching to blooming and blooming to setting pods as the key growth phases for soybean yield, and adding these phases as inputs to the model further improves prediction accuracy. Our findings suggest that incorporating causal relationships in deep learning models can improve prediction accuracy. This study provides a novel approach for remote sensing prediction of crop yield.
format Article
id doaj-art-74875ffdded24fa882e4d4078dbe363f
institution Kabale University
issn 1939-1404
2151-1535
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-74875ffdded24fa882e4d4078dbe363f2024-12-18T00:00:18ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352024-01-0117141611417810.1109/JSTARS.2024.343569910614767Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing DataFumin Wang0https://orcid.org/0000-0002-5078-358XJiale Li1https://orcid.org/0000-0003-0754-1793Dailiang Peng2https://orcid.org/0009-0006-5101-2935Qiuxiang Yi3Xiaoyang Zhang4Jueyi Zheng5Siting Chen6Institute of Applied Remote Sensing and Information Technology, Zhejiang University, Hangzhou, ChinaInstitute of Applied Remote Sensing and Information Technology, Zhejiang University, Hangzhou, ChinaKey Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, ChinaSchool of Geomatics and Municipal Engineering, Zhejiang University of Water Resources and Electric Power, Hangzhou, ChinaDepartment of Geography & Geospatial Sciences, South Dakota State University, Brookings, SD, USAInstitute of Applied Remote Sensing and Information Technology, Zhejiang University, Hangzhou, ChinaInstitute of Applied Remote Sensing and Information Technology, Zhejiang University, Hangzhou, ChinaTimely and accurate crop yield estimation is crucial for managing crops, trade, and food security. The combination of remote sensing technology with machine learning methods is increasingly popular for global yield prediction. However, traditional machine learning methods rely on data correlation rather than causality, leading to poor interpretability. To address this issue, we propose a novel approach that combines a structural causal model (SCM) with deep learning to develop a causal graph attention network model (SCM-GAT) based on causal relationships for soybean yield prediction at the county level of 10 major soybean-producing states in the United States. The SCM-GAT model considers not only conventional vegetation indices and weather variables but also the causal relationships between variables as inputs. Using independent validation and five-fold cross-validation strategies, our results show that the predictive performance of the SCM-GAT model is not only superior to traditional prediction models, such as selection operator regression (LASSO), random forest, but also superior to the common deep learning models based on correlation relationships, such as long short-term memory network, and transformer. It has also proved to be more robust than other models under extreme weather events. In addition, we identify branching to blooming and blooming to setting pods as the key growth phases for soybean yield, and adding these phases as inputs to the model further improves prediction accuracy. Our findings suggest that incorporating causal relationships in deep learning models can improve prediction accuracy. This study provides a novel approach for remote sensing prediction of crop yield.https://ieeexplore.ieee.org/document/10614767/Causal inferencedeep learninggraph attention network (GAT)soybean yield predictionstructural causal model (SCM)
spellingShingle Fumin Wang
Jiale Li
Dailiang Peng
Qiuxiang Yi
Xiaoyang Zhang
Jueyi Zheng
Siting Chen
Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Causal inference
deep learning
graph attention network (GAT)
soybean yield prediction
structural causal model (SCM)
title Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
title_full Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
title_fullStr Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
title_full_unstemmed Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
title_short Estimating Soybean Yields Using Causal Inference and Deep Learning Approaches With Satellite Remote Sensing Data
title_sort estimating soybean yields using causal inference and deep learning approaches with satellite remote sensing data
topic Causal inference
deep learning
graph attention network (GAT)
soybean yield prediction
structural causal model (SCM)
url https://ieeexplore.ieee.org/document/10614767/
work_keys_str_mv AT fuminwang estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT jialeli estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT dailiangpeng estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT qiuxiangyi estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT xiaoyangzhang estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT jueyizheng estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata
AT sitingchen estimatingsoybeanyieldsusingcausalinferenceanddeeplearningapproacheswithsatelliteremotesensingdata