摘要
传统的估产方法难以对水稻种植面积和产量进行及时准确地预测。利用TM影像资料,根据地物光谱特征的差异,提取稻田种植面积信息,在此基础上分析早稻像元的植被指数,与作物单产建立回归方程,通过检验标准筛选出方程拟合率高的估产模型,从而预报水稻总产。结果显示:2009年奉新县早稻实际总产和平均单产与遥感估产的结果相比,早稻总产误差率为5.78%,单产误差率为1.79%。由此可以看出,结合GIS技术和资料进行水稻估产,既提高了估产精度,又降低了费用,具有一定的可靠性和实用性。
It is difficult to forecast the rice planting area and production accurately and in time by means of traditional estimating methods. By using TM image,the information of paddy area was extracted in accordance with the difference of spectral feature. Based on this,analysis on the vegetation index of early rice pix was carried out,combined with crop yields to get high fitting rate of yield estimation model which can pass the touchstone. Compared with the actual condition in Fengxin County in 2009,the results of yield estimation made the error of 5.78%in the total rice production,and 1.79% in the average yield. It can be concluded that application of GIS technology and data in rice yield estimation,not only improves the estimation accuracy,but also reduces the costs with certain reliability and practicality.
出处
《江西农业大学学报》
CAS
CSCD
北大核心
2010年第6期1103-1108,共6页
Acta Agriculturae Universitatis Jiangxiensis
基金
中国科学院知识创新工程重大项目(KSCX1-YW-09)
关键词
水稻
遥感估产
TM影像
植被指数
rice
yield assessment by remote sensing
TM image
vgetation index
