基于NOAA-NDVI的河南省冬小麦遥感估产

以河南省为研究区域,利用1982-2006年冬小麦生长关键期3-5月的NOAA-NDVI数据集,结合河南省18地市冬小麦产量数据,分析了研究区NDVI与冬小麦产量的时间动态变化特征,建立了基于NDVI的冬小麦产量估算模型,并对模型进行了验证与比较。结果表明:自1982-2006年,研究区冬小麦生长关键期的NDVI和小麦产量均表现出显著增长态势;利用3月、4月、5月NDVI分别建立的冬小麦产量线性预测模型相对误差均较小,分别为-7.56%、-6.10%、-2.63%;利用不同月份组合的NDVI累积和分别建立的冬小麦产量预测模型相对误差较小,平均误差为-5.08%;利用多个月份组合的NDVI分别建立的多元线性回归模型,虽然综合考虑了冬小麦生长期,但估测精度却低于基于单月NDVI和NDVI累积所建立的模型精度。通过误差对比分析后发现,利用5月NDVI建立的冬小麦估产模型误差最小。

Estimation of winter wheat yield based on the NOAA-NDVI data

Grain production estimates has an important significance in national macro-control of food production,the normalized difference vegetation index can accurately reflects large scale crop growth information.The changes of NDVI and winter wheat yield were studied based on NOAA-NDVI data in March,April and May from 1982 to 2006 and winter wheat yield data in 18 cities as one of the main data sources,and estimation models was established based above-mentioned data,and accuracies of the models were validated and compared.The results show that there is an increasing tendency at NDVI of winter wheat critical period,and wheat production also shows a significant growth from 1982 to 2006 in the study region.Linear models have small errors based on the NDVI of March,April,May and winter wheat production,forecast error is-7.56%、-6.10%、-2.63% respectively.Prediction model has a smaller error established with different months of combination of NDVI accumulation and yield of winter wheat,average error was-5.08%.The estimation accuracy of multiple linear prediction models based on many months NDVI is lower than the accuracy of the model based on the cumulative NDVI and single-month NDVI,although considering more growth stages of winter wheat.Error of estimation model established with NDVI in May is smallest in all of the models by comparison analysis.