Estimation of Winter Wheat Yield Based on NOAA-NDVI Data

The paper studied the changes of NDVI and winter wheat yild based on NOAA-NDVI data in March, April and May from 1982 to 2006 and winter wheat yield of Henan Province, estimation models was established with above data, and accuracy of the models were validated and compared. The results showed that there was an increasing tendency at NDVI of winter wheat critical period, and wheat yield also showed a significant growth during study period. Linear models have small errors based on the NDVI of March, April, May and winter wheat yield, forecast error is -7.56%, -6.10%, -2.63% respectively. Prediction model established with different months of combination of NDVI accumulation and yield of winter wheat, has a smaller error, the average was -5.08%. The accuracy of multiple linear prediction models based on NDVI is lower than accuracy of the model based on cumulative NDVI and single-month NDVI, although considering more growth stages of winter wheat. Regional NDVI can quickly and efficiently estimates winter wheat yield, but the accuracy and stability needs further enhanced.