Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning

Northeast China is one of China’s most important rice production bases, contributing about one-fifth of the country’s rice production. In recent years, several agricultural policies have been implemented in Northeast China to adjust crop structures, driven by economic and ecological benefits. Timely...

Full description

Saved in:
Bibliographic Details
Main Authors: Shihao Wang, Chong Huang, Lingxiao Huang, Xinliang Xu, Huading Shi, Qingbao Gu, Qiang Xue, Shiai Liu, Chenchen Zhang
Format: Article
Language:English
Published: Elsevier 2025-08-01
Series:International Journal of Applied Earth Observations and Geoinformation
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1569843225003863
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850035108570988544
author Shihao Wang
Chong Huang
Lingxiao Huang
Xinliang Xu
Huading Shi
Qingbao Gu
Qiang Xue
Shiai Liu
Chenchen Zhang
author_facet Shihao Wang
Chong Huang
Lingxiao Huang
Xinliang Xu
Huading Shi
Qingbao Gu
Qiang Xue
Shiai Liu
Chenchen Zhang
author_sort Shihao Wang
collection DOAJ
description Northeast China is one of China’s most important rice production bases, contributing about one-fifth of the country’s rice production. In recent years, several agricultural policies have been implemented in Northeast China to adjust crop structures, driven by economic and ecological benefits. Timely monitoring of the changed pattern of rice cultivation is a prerequisite for policy assessment. Current paddy rice mapping methods are experiencing uncertainties due to confusion with wetlands and are highly parameter-dependent. To these, in this study, we developed a paddy rice mapping framework that integrates automatically generated training samples, time series features from key cultivation stages, and a deep learning model to improve paddy rice identification accuracy in Northeast China, which is a typical rice-wetland coexisting area. We produced 10 m paddy rice maps for 2019–2023 in Northeast China and examined their changes under agricultural policy implementation. The resultant paddy rice maps have high accuracies, with overall accuracies >0.97, producer’s accuracies >0.94, user’s accuracies ≥0.93, and F1 scores of ≥0.95, respectively. Our proposed mapping method effectively identified small patches of paddy rice and reduced confusion with wetlands. Paddy rice areas in Northeast China estimated in this study continued to decrease annually from 71.1 × 103 km2 in 2019 to 56.4 × 103 km2 in 2023. Hot spots of rice conversion to other crops were found in the Sanjiang Plain, mainly due to the Soybean Revitalization Plan. The rice cultivation expansion was mainly found in the Songnen Plain, resulting from policies on Rehabilitation and Utilization of Saline Soils. Observed changes in paddy rice plantation emphasize the importance and necessity of timely and continuous crop cultivation monitoring under the influence of agricultural policy adjustment.
format Article
id doaj-art-aca7997a423e49348a809f64de42ed07
institution DOAJ
issn 1569-8432
language English
publishDate 2025-08-01
publisher Elsevier
record_format Article
series International Journal of Applied Earth Observations and Geoinformation
spelling doaj-art-aca7997a423e49348a809f64de42ed072025-08-20T02:57:35ZengElsevierInternational Journal of Applied Earth Observations and Geoinformation1569-84322025-08-0114210473910.1016/j.jag.2025.104739Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learningShihao Wang0Chong Huang1Lingxiao Huang2Xinliang Xu3Huading Shi4Qingbao Gu5Qiang Xue6Shiai Liu7Chenchen Zhang8State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing 100012, China; Technical Centre for Soil, Agriculture and Rural Ecology and Environment, Ministry of Ecology and Environment, Beijing 100012, ChinaState Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China; Corresponding authors.State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, ChinaState Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, ChinaTechnical Centre for Soil, Agriculture and Rural Ecology and Environment, Ministry of Ecology and Environment, Beijing 100012, ChinaState Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Sciences, Beijing 100012, ChinaGansu Province Geological Disaster Control Center, Lanzhou 730000, ChinaDongying Agriculture and Rural Bureau, Dongying 257091, ChinaSchool of Biological Sciences, Center for Earth Observation and Modeling, University of Oklahoma, Norman, OK 73019, USA; Corresponding authors.Northeast China is one of China’s most important rice production bases, contributing about one-fifth of the country’s rice production. In recent years, several agricultural policies have been implemented in Northeast China to adjust crop structures, driven by economic and ecological benefits. Timely monitoring of the changed pattern of rice cultivation is a prerequisite for policy assessment. Current paddy rice mapping methods are experiencing uncertainties due to confusion with wetlands and are highly parameter-dependent. To these, in this study, we developed a paddy rice mapping framework that integrates automatically generated training samples, time series features from key cultivation stages, and a deep learning model to improve paddy rice identification accuracy in Northeast China, which is a typical rice-wetland coexisting area. We produced 10 m paddy rice maps for 2019–2023 in Northeast China and examined their changes under agricultural policy implementation. The resultant paddy rice maps have high accuracies, with overall accuracies >0.97, producer’s accuracies >0.94, user’s accuracies ≥0.93, and F1 scores of ≥0.95, respectively. Our proposed mapping method effectively identified small patches of paddy rice and reduced confusion with wetlands. Paddy rice areas in Northeast China estimated in this study continued to decrease annually from 71.1 × 103 km2 in 2019 to 56.4 × 103 km2 in 2023. Hot spots of rice conversion to other crops were found in the Sanjiang Plain, mainly due to the Soybean Revitalization Plan. The rice cultivation expansion was mainly found in the Songnen Plain, resulting from policies on Rehabilitation and Utilization of Saline Soils. Observed changes in paddy rice plantation emphasize the importance and necessity of timely and continuous crop cultivation monitoring under the influence of agricultural policy adjustment.http://www.sciencedirect.com/science/article/pii/S1569843225003863Paddy riceAgricultural policyDeep learningDeep neural networks
spellingShingle Shihao Wang
Chong Huang
Lingxiao Huang
Xinliang Xu
Huading Shi
Qingbao Gu
Qiang Xue
Shiai Liu
Chenchen Zhang
Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
International Journal of Applied Earth Observations and Geoinformation
Paddy rice
Agricultural policy
Deep learning
Deep neural networks
title Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
title_full Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
title_fullStr Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
title_full_unstemmed Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
title_short Monitoring paddy rice cultivation adjustments in Northeast China through time series remote sensing and deep learning
title_sort monitoring paddy rice cultivation adjustments in northeast china through time series remote sensing and deep learning
topic Paddy rice
Agricultural policy
Deep learning
Deep neural networks
url http://www.sciencedirect.com/science/article/pii/S1569843225003863
work_keys_str_mv AT shihaowang monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT chonghuang monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT lingxiaohuang monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT xinliangxu monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT huadingshi monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT qingbaogu monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT qiangxue monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT shiailiu monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning
AT chenchenzhang monitoringpaddyricecultivationadjustmentsinnortheastchinathroughtimeseriesremotesensinganddeeplearning