基于GEE的苹果园地遥感信息提取研究OACSTPCD
Remote Sensing Extraction of Apple Orchard Based on GEE-A Case Study of Weibei Plateau Region in Shaanxi Province
准确掌握苹果园空间分布及动态变化对苹果产业发展及规划具有重要意义.以陕西主要种植区——渭北旱塬区14个县(区)为研究区,基于Google Earth Engine(GEE)云计算平台,以2020年Sentinel-2卫星影像的Level-2A级数据为数据源,结合实地采样获得的461个苹果园样本点,构建包含光谱特征、遥感植被指数、纹理特征和地形特征等共计21个特征作为分类特征集,利用随机森林分类方法对研究区苹果园地空间信息进行提取.结果表明,研究区苹果种植面积为23.03万hm2,连片种植的苹果园主要分布于延安市洛川县、渭南市白水县、咸阳市淳化县和咸阳市彬州市,种植面积分别为3.54、3.21、2.90万hm2和2.51万hm2;包含光谱特征、遥感植被指数、纹理特征和地形特征的分类特征集在提取苹果园空间格局方面效果最优,苹果园地提取的总体精度和Kappa系数分别为94%和0.86,高于其他分类特征组合.研究成果可为未来陕西省苹果产业合理规划及果园精细化管理提供科学依据和数据支撑.
Shaanxi Province is the best apple production region in the world with the largest contiguous ap-ple planting area.Accurate understanding of the spatial patterns and dynamic changes of apple orchards is of paramount importance for the advancement of strategic planning of the apple industry.Fourteen counties located in the Weibei Plateau of Shaanxi Province were taken as the research area.Based on the Google Earth Engine(GEE)cloud computing platform,leveraged Level-2 A data from Sentinel-2 remote sensing images in 2020 were used as data source.Combining with the 461 on-site sample points from apple or-chards,a comprehensive set of 21 classification features including spectral features,remote sensing vegeta-tion index,texture features,and topographic features were constructed.The random forest classification method was used to extract the spatial information of apple orchards in the Weibei Plateau.The results showed that the planting area of apple orchards was 200 300 hm2.Apple orchards with contiguous distribu-tion were mainly distributed in Luochuan of Yan'an,Baishui of Weinan,Chunhua of Xianyang,and Binzhou of Xianyang,with planting areas of 35 400,32 100,29 000 hm2 and 25 100 hm2,respectively.The scheme combining spectral features,remote sensing vegetation index,texture features,and topographic features performed the best among all these classification schemes with overall accuracy and the Kappa coefficient of 94%and 0.86,respectively.Our findings provide a scientific basis and data support for the rational plan-ning of the apple industry and the refined management of apple orchards in Shaanxi.
陈印;刚成诚;刘欢欢;刘悦;范蒙恩;陈宇;张曼;于子涵
西北农林科技大学草业与草原学院,陕西杨陵 712100西北农林科技大学水土保持研究所,陕西杨陵 712100||中国科学院水利部水土保持研究所,陕西杨陵 712100西北农林科技大学资源环境学院,陕西杨陵 712100
农业科学
Google Earth Engine苹果园Sentinel-2纹理特征渭北旱塬区
Google Earth Engineapple orchardSentinel-2texture featureWeibei Plateau region
《西北林学院学报》 2024 (001)
36-43 / 8
国家自然科学基金(31602004);青海省防灾减灾重点实验室开放基金(QFZ-2021-Z06);陕西省自然科学基金(2021JQ-171).
评论