DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation
Large-scale crop yield estimation is critical for understanding the dynamics of global food security. Understanding and quantifying the temporal cumulative effect of crop growth and spatial variances across different regions remains challenging for large-scale crop yield estimation. In this study, a...
| 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/ab66cb |
| _version_ | 1797747622296944640 |
|---|---|
| author | Tao Lin Renhai Zhong Yudi Wang Jinfan Xu Hao Jiang Jialu Xu Yibin Ying Luis Rodriguez K C Ting Haifeng Li |
| author_facet | Tao Lin Renhai Zhong Yudi Wang Jinfan Xu Hao Jiang Jialu Xu Yibin Ying Luis Rodriguez K C Ting Haifeng Li |
| author_sort | Tao Lin |
| collection | DOAJ |
| description | Large-scale crop yield estimation is critical for understanding the dynamics of global food security. Understanding and quantifying the temporal cumulative effect of crop growth and spatial variances across different regions remains challenging for large-scale crop yield estimation. In this study, a deep spatial-temporal learning framework, named DeepCropNet (DCN), has been developed to hierarchically capture the features for county-level corn yield estimation. The temporal features are learned by an attention-based long short-term memory network and the spatial features are learned by the multi-task learning (MTL) output layers. The DCN model has been applied to quantify the relationship between meteorological factors and the county-level corn yield in the US Corn Belt from 1981 to 2016. Three meteorological factors, including growing degree days, killing degree days, and precipitation, are used as time-series inputs. The results show that DCN provides an improved estimation accuracy (RMSE = 0.82 Mg ha ^−1 ) as compared to that of conventional methods such as LASSO (RMSE = 1.14 Mg ha ^−1 ) and Random Forest (RMSE = 1.05 Mg ha ^−1 ). Temporally, the attention values computed from the temporal learning module indicate that DCN captures the temporal cumulative effect and this temporal pattern is consistent across all states. Spatially, the spatial learning module improves the estimation accuracy based on the regional specific features captured by the MTL mechanism. The study highlights that the DCN model provides a promising spatial-temporal learning framework for corn yield estimation under changing meteorological conditions across large spatial regions. |
| first_indexed | 2024-03-12T15:53:14Z |
| format | Article |
| id | doaj.art-d3f13d8f6a94422e950a1cf999aa5b46 |
| institution | Directory Open Access Journal |
| issn | 1748-9326 |
| language | English |
| last_indexed | 2024-03-12T15:53:14Z |
| publishDate | 2020-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Environmental Research Letters |
| spelling | doaj.art-d3f13d8f6a94422e950a1cf999aa5b462023-08-09T15:03:25ZengIOP PublishingEnvironmental Research Letters1748-93262020-01-0115303401610.1088/1748-9326/ab66cbDeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimationTao Lin0Renhai Zhong1Yudi Wang2Jinfan Xu3Hao Jiang4Jialu Xu5Yibin Ying6Luis Rodriguez7K C Ting8Haifeng Li9https://orcid.org/0000-0003-1173-6593College of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of ChinaCollege of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of China; International Campus, Zhejiang University , Haining, Zhejiang, 314400, People’s Republic of ChinaChina Academy of Electronic and Information Technology , Beijing, 100041, People’s Republic of ChinaCollege of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of ChinaCollege of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of ChinaCollege of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of ChinaCollege of Biosystems Engineering and Food Science, Zhejiang University , Hangzhou, Zhejiang, 310058, People’s Republic of China; Faculty of Agricultural and Food Science, Zhejiang A&F University , Hangzhou, Zhejiang, 311300, People’s Republic of ChinaDepartment of Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign , Urbana, IL, United States of AmericaInternational Campus, Zhejiang University , Haining, Zhejiang, 314400, People’s Republic of China; Department of Agricultural and Biological Engineering, University of Illinois at Urbana-Champaign , Urbana, IL, United States of AmericaSchool of Geosciences and Info-Physics, Central South University , Changsha, 410083, People’s Republic of China; Henan Laboratory of Spatial Information Application on Ecological Environment Protection, Zhengzhou, 450000, People’s Republic of ChinaLarge-scale crop yield estimation is critical for understanding the dynamics of global food security. Understanding and quantifying the temporal cumulative effect of crop growth and spatial variances across different regions remains challenging for large-scale crop yield estimation. In this study, a deep spatial-temporal learning framework, named DeepCropNet (DCN), has been developed to hierarchically capture the features for county-level corn yield estimation. The temporal features are learned by an attention-based long short-term memory network and the spatial features are learned by the multi-task learning (MTL) output layers. The DCN model has been applied to quantify the relationship between meteorological factors and the county-level corn yield in the US Corn Belt from 1981 to 2016. Three meteorological factors, including growing degree days, killing degree days, and precipitation, are used as time-series inputs. The results show that DCN provides an improved estimation accuracy (RMSE = 0.82 Mg ha ^−1 ) as compared to that of conventional methods such as LASSO (RMSE = 1.14 Mg ha ^−1 ) and Random Forest (RMSE = 1.05 Mg ha ^−1 ). Temporally, the attention values computed from the temporal learning module indicate that DCN captures the temporal cumulative effect and this temporal pattern is consistent across all states. Spatially, the spatial learning module improves the estimation accuracy based on the regional specific features captured by the MTL mechanism. The study highlights that the DCN model provides a promising spatial-temporal learning framework for corn yield estimation under changing meteorological conditions across large spatial regions.https://doi.org/10.1088/1748-9326/ab66cbyield estimationcornLSTMattention mechanismmulti-task learningdeep learning |
| spellingShingle | Tao Lin Renhai Zhong Yudi Wang Jinfan Xu Hao Jiang Jialu Xu Yibin Ying Luis Rodriguez K C Ting Haifeng Li DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation Environmental Research Letters yield estimation corn LSTM attention mechanism multi-task learning deep learning |
| title | DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation |
| title_full | DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation |
| title_fullStr | DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation |
| title_full_unstemmed | DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation |
| title_short | DeepCropNet: a deep spatial-temporal learning framework for county-level corn yield estimation |
| title_sort | deepcropnet a deep spatial temporal learning framework for county level corn yield estimation |
| topic | yield estimation corn LSTM attention mechanism multi-task learning deep learning |
| url | https://doi.org/10.1088/1748-9326/ab66cb |
| work_keys_str_mv | AT taolin deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT renhaizhong deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT yudiwang deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT jinfanxu deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT haojiang deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT jialuxu deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT yibinying deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT luisrodriguez deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT kcting deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation AT haifengli deepcropnetadeepspatialtemporallearningframeworkforcountylevelcornyieldestimation |