欢迎访问林草资源研究

林业资源管理 ›› 2017›› Issue (4): 131-134.doi: 10.13466/j.cnki.lyzygl.2017.04.020

• 科学研究 • 上一篇    下一篇

基于Landsat TM遥感数据的山核桃产量预测——以浙江临安市为例

栗晓禹1(), 黄兴召2(), 王雪军1, 高作锋1   

  1. 1.国家林业局调查规划院设计院,北京 100714
    2.安徽农业大学 林学与园林学院,合肥 230036
  • 收稿日期:2017-04-10 修回日期:2017-05-15 出版日期:2017-08-28 发布日期:2020-09-24
  • 通讯作者: 黄兴召
  • 作者简介:栗晓禹(1982-),女,辽宁本溪人,工程师,硕士,研究方向:森林经理。Email:lixiaoyu@afip.com.cn
  • 基金资助:
    国家高技术发展计划(863计划)(2013AA102605);国家自然科学基金(31170637)

Estimation of Hickory Yield Based on Landsat TM Remote Sensing Data

LI Xiaoyu1(), HUANG Xingzhao2(), WANG Xuejun1, GAO Zuofeng1   

  1. 1. Academy of Forest Inventory and Planning,SFA,Beijing 100714,China
    2. School of Forestry and Landscape Architecture,Anhui Agricultural University,Hefei 230036,China
  • Received:2017-04-10 Revised:2017-05-15 Online:2017-08-28 Published:2020-09-24
  • Contact: HUANG Xingzhao

摘要:

以浙江省临安市的山核桃为研究对象,基于2008—2011年连续4年的样地实测产量为基础,利用每年4个生长时期的Landsat TM遥感数据,系统地分析比较每个生长时期的植被指数与产量的关系。研究结果表明:NDVI 在各个生长期均与产量的相关性最高;SAVI与产量的相关性居中,DVI最低。以每个时期的NDVI为因子,建立不同时期山核桃产量的预估模型。各时期模型的预估效果为果实膨大期>花芽分化及授粉期>采摘至落叶期>休眠期。以不同时期的NDVI为因子,利用逐步回归,建立多因子的山核桃产量的预估模型。最优预估模型为y=126.51x2+26.61x1+12.56x3-67.42(R2=0.642,SEE=12.17),为山核桃产量的预测提供可行,快速,有效的方法。

关键词: 山核桃, 生长期, 遥感, 逐步回归, 模型

Abstract:

To establish the hickory yield model and vegetation index,the actual yield from 2008 to 2011 and Landsat TM remote sensing data of four growth stages in every year were used to systematically compare their relationship in hickory source region in Lin’an of Zhejiang Province.The results show that the NDVI of each growth stage has higher correlation with yield than SAVI,DVI.The model of each growth stage was built to predict hickory yield which used NDVI.The accuracy of four models was as follows: fruit expanding stage>flower bud differentiation and pollination stage>picking to defoliation stage>dormancy stage.Using the stepwise regression that the NDVI of every growth stage were factors,the model of hickory yield was established.The optimal model was y=126.51x2+26.61x1+12.56x3-67.42(R2=0.642,SEE=12.17) which provided a feasible,rapid and effective method to predict the hickory production.

Key words: hickory, growth stage, remote sensing, stepwise regression, model

中图分类号: