Analysis of crop residue cover in the North China Plain based on multispectral remote sensing and machine learning

遥感 多光谱图像 多光谱模式识别 环境科学 残留物(化学) 土地覆盖 作物残渣 封面(代数) 专题制图器 遥感应用 作物 多光谱扫描仪 反射率 随机森林 卫星图像 精准农业 作物管理 覆盖作物
作者
Jibo Yue,Ranran Yang,Yinghao Lin,Nianxu Xu,Qingjiu Tian,Jia Tian
出处
期刊:International Journal of Remote Sensing [Taylor & Francis]
卷期号:: 1-28
标识
DOI:10.1080/01431161.2026.2727176
摘要

Crop residue cover (CRC) is a critical parameter for evaluating conservation tillage practices, mitigating soil erosion, and developing models for global carbon cycle monitoring. Traditional optical remote sensing index-based methods are significantly affected by variations in soil and crop residue moisture content as well as interference from emerging subsequent crops, resulting in low estimation accuracy and poor stability of CRC. Due to the presence of subsequent crops, the non-photosynthetic vegetation fraction (fNPV) derived from linear spectral mixture analysis is typically lower than the actual field CRC. This study proposes a novel CRC estimation method based on linear spectral mixture analysis. The workflow consists of three main components: (1) development of a Crop Residue Random Forest (CRRF) index using a random forest regression model to simulate the narrowband SINDRI index – which is less sensitive to moisture – from broadband multispectral reflectance; (2) construction of a two-dimensional triangular feature space combining the CRRF and NDVI indices, followed by linear spectral mixture analysis to simultaneously estimate field fNPV, photosynthetic vegetation fraction (fPV), and bare soil fraction (fBS); and (3) retrieval of the true CRC fraction at the sowing stage using the formula CRC = fNPV / (1 − fPV). The proposed method was validated and applied regionally using multi-temporal Sentinel-2 MSI and MODIS imagery. Results demonstrate that: (1) the method effectively mitigates moisture interference and achieves high-accuracy fNPV estimation (R2 = 0.84, RMSE = 0.10); (2) the integration of Sentinel-2 and MODIS sensors enables stable estimation and spatio-temporal dynamic analysis of CRC across the North China Plain, with high temporal consistency between images acquired approximately 16 days apart (the majority of agricultural pixels showing absolute differences within ±0.15). This approach provides an efficient and robust technical solution for remote sensing monitoring of CRC fraction in complex agricultural environments and offers a valuable methodological reference for the assessment of conservation tillage practices.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
孙七喜完成签到,获得积分10
1秒前
曦贝贝完成签到,获得积分10
1秒前
小灰灰完成签到,获得积分0
1秒前
嘟嘟完成签到,获得积分10
1秒前
panda完成签到,获得积分0
2秒前
3秒前
蜀勤完成签到,获得积分10
3秒前
研友_LJGmvn发布了新的文献求助10
3秒前
甜甜元蝶的应助被suiyi采纳,获得10
3秒前
孤独丹秋完成签到,获得积分10
4秒前
5秒前
5秒前
雪碧和果冻完成签到,获得积分10
5秒前
星河完成签到 ,获得积分10
5秒前
听汐完成签到 ,获得积分10
6秒前
Zhao完成签到,获得积分10
6秒前
6秒前
ace完成签到,获得积分10
6秒前
lay完成签到,获得积分10
6秒前
欢呼以柳完成签到,获得积分10
7秒前
Nobody完成签到,获得积分10
7秒前
听话的八宝粥完成签到 ,获得积分10
8秒前
8秒前
睡觉多会瞌睡完成签到,获得积分10
8秒前
8秒前
vivi完成签到,获得积分10
9秒前
leileilei完成签到,获得积分10
9秒前
suiyi完成签到,获得积分10
9秒前
星期六睡觉完成签到,获得积分10
9秒前
七色光完成签到,获得积分10
10秒前
jokerli完成签到,获得积分10
10秒前
研友_LJGmvn完成签到,获得积分10
10秒前
10秒前
活泼冬菱完成签到,获得积分10
10秒前
慕昊强完成签到,获得积分10
10秒前
月光盏完成签到,获得积分10
11秒前
尊敬背包完成签到,获得积分10
11秒前
11秒前
zxp完成签到,获得积分10
11秒前
1Arendelle1完成签到,获得积分10
12秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
The World Atlas of Wine 9th edition 1000
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
plantTFDB 508
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8173474
求助须知:如何正确求助?哪些是违规求助? 9565719
关于积分的说明 21781727
捐赠科研通 7767700
什么是DOI,文献DOI怎么找? 3431597
关于科研通互助平台的介绍 2555855
邀请新用户注册赠送积分活动 2476832