Research on Yield Estimate Dynamic Model of Rice in Chongqing

In this paper, MODIS-250m vegetation index from first 10-day of March to last 10-day of November in 2010and 2011, land use data and the yield data from 9 counties such as Jiangjin, Wanzhou, Fengdu, Liangping, Qijiang, Liangping, Fuling, Yunyang, Kaixian, Wixi in Chongqing were used. Based on GIS, RS technique and NDVI analysis results, the rice planting area was extracted. Four key growth period of rice such tillering, jointing, heading and milk were selected to establish dynamic assessment model. In nine counties, the average relative error of the remaining 8 counties was between 5.36%-17.09%, and the mean was 12.31% except Qijiang that estimation results and actual yield deviation was too large. The results show that building rice yield estimation model using normalized difference vegetation index was feasible that can realize the dynamic yield estimation of rice.