基于HJ-1A/1B数据的冬小麦成熟期遥感预测

基于遥感数据获取作物成熟度信息，制订收割顺序，是遥感在精准农业中的一个重要应用课题，目前的作物物候监测在空间分辨率和预测的时效性上尚无法满足精准农业的要求。以山东禹城市为研究区，选取观测点，分析冬小麦成熟期临近过程中水分和叶绿素的动态变化，获取基于遥感监测小麦成熟期的依据。以HJ-1A CCD数据构建的植被指数来反映作物绿度的变化，以HJ-1B IRS数据构建的归一化水指数来反映作物含水率的变化，通过回归分析建立了冬小麦成熟期的遥感预测模型，实现了禹城市冬小麦成熟期的遥感监测。小麦成熟期预测值与观测值的相关性达极显著水平，预测与观测小麦成熟期的先后顺序较为一致，在以误差小于1 d作为成功预测衡量标准的情况下，预测结果的准确度为65%。提出了使用HJ-1A/1B数据开展冬小麦成熟期预测的模型，而使用不同时相遥感数据预测成熟期的通用模型将成为下一步研究重点。

Predicting mature date of winter wheat with HJ-1A/1B data

Predicting crop mature date and producing harvesting order is an important field of applying remote sensing in precision farming. Current crop phenophase monitoring method with remote sensing cannot meet the needs of precision farming, due to its low spatial resolution and time lag in information acquiring. Taking Yucheng (Shandong) as study area, this paper analyzed the dynamic variation of moisture content and chlorophyll in the maturing period of winter wheat to provide theoretical basis for mature date prediction. Winter wheat mature date predicting model was developed through regression analysis by using VI from HJ-1A CCD to describe the change of chlorophyll and NDWI from HJ-1B IRS to describe the change of water content. The winter wheat mature date map of Yucheng was produced. The a correlation between predicted and observed mature dates has reached very significant level. A rather consistent maturing order could be concluded. Taking the predictions with errors less than 1 day as successful prediction, the accuracy was 65%. The study showed that HJ-1A/1B data can be effectively used for winter wheat mature data prediction, universal predicting model with remote sensing data of different temporal period will be the focus in the following research.