地理研究 ›› 2012, Vol. 31 ›› Issue (12): 2310-2320.doi: 10.11821/yj2012120018

• 地球信息科学 • 上一篇    

基于MODIS-NDVI的河南省冬小麦产量遥感估测

赵文亮1,2, 贺振2, 贺俊平2, 朱连奇1   

  1. 1. 河南大学环境与规划学院, 河南开封475004;
    2. 商丘师范学院环境与规划学院, 河南商丘476000
  • 收稿日期:2012-05-17 修回日期:2012-08-04 出版日期:2012-12-10 发布日期:2012-12-10
  • 通讯作者: 朱连奇(1963-),河南郸城县人,博导,教授,主要从事资源开发利用研究。E-mail:lqzhu@henu.edu.cn E-mail:lqzhu@henu.edu.cn
  • 作者简介:赵文亮(1963-),河南长葛人,博士生,副教授,主要从事区域资源开发研究。E-mail:zhaowenliang1169@163.com
  • 基金资助:

    国家自然科学基金资助项目(41140019);河南省科技攻关重点项目(112102210209);河南省教育厅 自然科学研究计划项目(2011B170007);河南省教育厅科学技术研究重点项目(12B420003);商丘师 范学院青年科研基金项目(2010QN18);商丘师范学院青年骨干教师科研项目

Remote sensing estimation for winter wheat yield in Henan based on the MODIS-NDVI data

ZHAO Wen-liang1,2, HE Zhen2, HE Jun-ping2, ZHU Lian-qi1   

  1. 1. College of Environment and Planning, Henan University, Kaifeng 475004, Henan, China;
    2. College of Environment and Planning, Shangqiu Normal University, Shangqiu 476000, Henan, China
  • Received:2012-05-17 Revised:2012-08-04 Online:2012-12-10 Published:2012-12-10

摘要: 粮食产量估测对于国家粮食生产宏观调控具有重要意义。以河南省为例, 利用区域 NDVI数据进行冬小麦产量监测研究。基于2000~2010年冬小麦生长关键期3~5月的MODIS~ NDVI数据集, 结合河南省18地市冬小麦生产数据, 分析了研究区小麦产量和播种面 积的时间动态变化特征, 建立了基于区域NDVI的冬小麦产量估算模型。结果表明:自2000 至2010年, 研究区冬小麦产量呈上升趋势, 播种面积基本保持稳定;利用单月NDVI建立的 冬小麦产量线性模型, 平均相对误差分别为12.02%、10.70%和9.27%;利用不同月份组合 的NDVI累积和建立的冬小麦产量模型, 平均相对误差分别为11.13%、10.38%、8.37%和 9.41%;利用多个月份组合的NDVI建立的多元线性回归模型, 平均相对误差分别为 11.00%、9.32%、9.04%和9.58%;将小麦播种面积作为限制因素引入多元线性模型后, 估 算精度得到了很大提升, 平均相对误差分别为5.65%、5.34%、6.76%和5.47%.通过误差 对比后发现, 在模型中引入播种面积后, 利用区域NDVI可以有效、快速、准确地对冬小麦 进行估产。

关键词: MODIS-NDVI, 冬小麦, 遥感估产, 河南省

Abstract: Grain production estimates has an important significance in national macro-control of food production.In this paper,winter wheat yield estimation models were devised based on regional NDVI and planting area of winter wheat in Henan province.The annual variation of NDVI and wheat planting status were investigated based on MODIS-NDVI data in March,April and May from 2000 to 2010 and winter wheat yield data in this province, and estimation models was established based on the above-mentioned data.The results showed that wheat production had a significant growth from 2000 to 2010 in the study region when wheat planting area kept stable.Through the statistical calculation and error comparison in models,it is showed that average relative errors were 12.02%,10.70% and 9.27% respectively based on linear models built with NDVI of March,April,May and winter wheat production.Average relative errors were 11.13%,10.38%,8.37% and 9. 41% respectively based on linear models built with NDVI accumulation and yield of winter wheat.Average relative errors were 11.00%,9.32%,9.04% and 9.58% respectively based on multiple linear prediction models built with NDVI of March,April,May and winter wheat production.When we took the planting area of winter wheat as an independent variable,the overall prediction accuracy of winter wheat production were 5.65%,5. 34%,6.76% and 5.47% respectively.Through comparing,the yield of winter wheat can be quickly and efficiently estimated with regional NDVI,but the accuracy and stability needs to be further improved.It is very difficult to predict winter wheat production accurately due to many influencing factors,but the accuracy can be gradually improved with various methods.

Key words: MODIS-NDVI, winter wheat, remote sensing estimation, Henan Province