|
| 协同Sentinel-2和随机森林算法的黄河三角洲冬小麦精细化绘制与时空演化分析 |
| Refined mapping and spatiotemporal evolution analysis of winter wheat in the Yellow River Delta using Sentinel-2 and the random forest algorithm |
| 投稿时间：2024-02-01 |
| DOI：10.13254/j.jare.2024.0107 |
| 中文关键词: 冬小麦，黄河三角洲，Sentinel-2，随机森林，动态变化监测 |
| 英文关键词: winter wheat, Yellow River Delta, Sentinel-2, random forest, dynamic change monitoring |
| 基金项目:国家自然科学基金项目（42171113）；山东省自然科学基金项目（ZR2021QD113）；济南市校融合项目（JNSX2023065） |
| 作者 | 单位 | E-mail | | 李静娴 | 山东建筑大学测绘地理信息学院, 济南 250101 | | | 刘建涛 | 山东建筑大学测绘地理信息学院, 济南 250101 | liujiantao18@sdjzu.edu.cn | | 王志平 | 山东省地质测绘院, 济南 250003 | | | 冯权泷 | 中国农业大学土地科学与技术学院, 北京 100083 | | | 孟飞 | 山东建筑大学测绘地理信息学院, 济南 250101 | | | 王会蒙 | 山东建筑大学测绘地理信息学院, 济南 250101 | | | 张冬 | 齐鲁空天信息研究院, 济南 250132
中国科学院空天信息创新研究院, 北京 100086 | | | 彭玉 | 齐鲁空天信息研究院, 济南 250132
中国科学院空天信息创新研究院, 北京 100086 | |
|
| 摘要点击次数: 2483 |
| 全文下载次数: 3546 |
| 中文摘要: |
| 近年来气候变化、设施农业发展、城市化进程加速、环境污染等因素影响了黄河三角洲东营市的冬小麦种植。准确、高效、精细地获取该区域冬小麦种植面积、空间分布以及变化趋势对于提升冬小麦管理的针对性和种植结构布局的有效性至关重要。本研究以Sentinel-2多光谱数据作为数据源，基于随机森林、支持向量机和面向对象等方法开展了2017、2019、2022年山东省东营市冬小麦种植区的绘制。通过对比发现，随机森林在东营市冬小麦种植区域绘制时展现出了优越的性能，三年冬小麦提取总体精度分别达到了 92.24%、92.66%、89.25%，Kappa系数分别达到 0.911 4、0.916 1、0.898 0；通过进一步分析随机森林生成的时间序列冬小麦空间分布数据，发现东营市各县区冬小麦种植面积总体增长，并呈现向东部和北部蔓延的趋势。研究结果对于改善该区域冬小麦管理效益以及种植结构布局具有重要参考价值。 |
| 英文摘要: |
| Winter wheat is a major staple crop in northern China, crucial for national food security and social stability. In recent years, winter wheat cultivation in Dongying City, located in the Yellow River Delta, has been significantly impacted by factors such as climate change, development of facility agriculture, accelerated urbanization, and environmental pollution. It is crucial to accurately, efficiently and precisely obtain the information, such as winter wheat planting area, spatial distribution and temporal trends in order to enhance the targeted management and effective layout of winter wheat cultivation in this region. This study utilized Sentinel-2 multispectral data as the data source and employed random forest, support vector machine, and object-based classification methods to delineate the winter wheat planting areas in Dongying City for the year 2017, 2019 and 2022. Through comparative experiments, it was found that random forest exhibited superior performance in delineating the winter wheat planting areas in Dongying City. The overall accuracy of winter wheat extraction reached 92.24%, 92.66% and 89.25% for the year 2017, 2019 and 2022, respectively, with corresponding Kappa coefficients of 0.911 4, 0.916 1 and 0.898 0. Through further analysis of the time-series winter wheat spatial distribution data generated by random forest, it was observed that the overall winter wheat planting area in each county of Dongying City showed a general increase and exhibited a tendency to spread eastward. The spatiotemporal characteristics of winter wheat planting areas discovered in this study are of significant reference value for improving the management efficiency and structural layout of winter wheat cultivation in the region. |
| HTML
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
