论文部分内容阅读
多源数据与其技术方法逐渐被应用于植被物候的研究当中,但基于多源数据物候识别方法间的差异性比较及定量化评估工作还有待加强。以山东禹城农田生态系统为例,探讨了基于多源数据,NDVI、EVI、数字相机图片、碳通量数据(NEE)以及人工实测数据获取的冬小麦主要生育日期的结果进行差异比较及定量化评估。结果表明:(1)通过碳通量数据获取的主要生育日期的计算结果与人工实测结果最接近,各阶段差异均<3d;通过数字相机图片获取的结果仅次于通过碳通量数据获取的结果,而通过遥感数据NDVI、EVI获取的结果与人工实测结果差距最大;(2)通过NDVI、EVI两种数据获取的冬小麦主要生育期结果具有极显著的相关性,最高达到R2=0.857(P<0.001);(3)基于多源数据获取的冬小麦主要生育期的计算结果,均显示出禹城站冬小麦返青期提前,蜡熟期推迟,生长季长度变长的年际变化特征。
Multi-source data and its technical methods are gradually being applied to the study of vegetation phenology, but the comparison and quantitative assessment based on the multi-source data phenological identification methods need to be strengthened. Taking Yucheng farmland ecosystem in Shandong Province as an example, the differences and quantification of the main growth dates of winter wheat based on multi-source data, NDVI, EVI, digital camera pictures, carbon flux data (NEE) and artificial measured data were discussed Evaluation. The results showed that: (1) The calculated results of the main fertility dates obtained from the carbon flux data were the closest to the results of artificial measurements, with differences of <3d in each stage; the results obtained with digital camera pictures were second only to those obtained from carbon flux data (2) The main growth period of winter wheat obtained by NDVI and EVI data has a very significant correlation, up to R2 = 0.857 (P <0.001). (3) The calculation results of winter wheat main growth period based on multi-source data all showed the interannual variability of winter wheat returning stage, waxing stage and long growing season in Yucheng Station.