| Nitrogen fertilizer is the largest and most widely used fertilizer in wheat.Due to the widespread abuse nitrogen fertilization phenomenon,the nitrogen fertilizer utilization efficiency is low,and the yield increase is slow or even reduced.Therefore,it is imperative to apply rapid and convenient nitrogen fertilizer application monitoring technology.Numerous studies at home and abroad have shown that under the same growth environment,canopy temperature changes can reflect the growth and physiological metabolism of wheat,and there is a significant relationship with wheat yield and composition factors.Therefore,it is useful to use canopy temperature to monitor the physiological metabolism of wheat,evaluate the advantages and disadvantages of nitrogen application,and guide wheat high-yield cultivation.It will have great application potential in digital agriculture and smart agriculture.In this study,a hand-held thermal infrared instrument and a thermal infrared camera equipped with UAV were used to obtain the canopy temperature distribution map during wheat flowering,and the photosynthesis ability of wheat in the same period was obtained.The effects of 0 kg·hm-2(N1)、180kg·hm-2(N2)、240 kg·hm-2(N3)and 300 kg·hm-2(N4)four N treatment on yangmai 23 canopy temperature,plant nitrogen content,photosynthetic physiological traits,biomass accumulation of different organs,wheat yield and its components were studied at 7 days after flowering,14 days after flowering and 21 days after flowering.The relationship between nitrogen content during flowering and wheat canopy temperature was known.On this basis,according to the research progress of the predecessors and the results of this study,the parameters are summed up as R,G,B,NRI,NGI,NBI,GMR,GR,ExG,ExR,ExGMR,NDIg,NDIb,NDI,canopy leaf temperature CT,wheat canopy temperature ST,and canopy coverage CC to construct the best estimation model of nitrogen content after wheat flowering.Then,we found the image characteristic parameters suitable for constructing the nitrogen content estimation model of wheat flower and modeled and tested by analyzing the correlation between the above-mentioned characteristic parameters and nitrogen content at 7d after flowering,14d after flowering and 21d after flowering.The results of the model construction are as follows:(1)CT showed significant correlations with the nitrogen content of post-anthesis leaves during the flowering stage.The best estimation model of leaf nitrogen content at 7d after flowering fitted by multiple stepwise regression analysis was y=-6.0934*CT+195.41 and its R2,RMSE were 0.7101,4.964 mg·g-1,respectively.The optimal estimation model of nitrogen content in leaves of 14d after flowering was y=-2.5351*CT+69.708,the R2 of the model was 0.6481,the RMSE was 1.807 mg·g-1.In addition,the best estimation model for leaf nitrogen content after 21 days of flowering was y=-1.037*CT+36.221 and its R2,RMSE were 0.6224,0.834 mg·g-1,respectively.Therefore,it is feasible to estimate leaf nitrogen content after flowering by using image characteristic parameter CT.(2)The correlation between PT and the nitrogen content in panicles was the best during the flowering stage.The best estimation model of panicles nitrogen content at 7d after flowering fitted by multiple stepwise regression analysis was y=-2.6926*PT+89.4,the R2 of the model was 0.6886,the RMSE was 2.962 mg·g-1.The optimal estimation model of nitrogen content in panicles of 14d after flowering was y=-3.0443*PT+75.442,the R2 of the model was 0.6698,the RMSE was 3.224 mg·g-1.In addition,the best estimation model for panicle nitrogen content after 21 days of flowering was y=-1.8058*PT+64.086 and its R2,RMSE were 0.683,2.735mg·g-1,respectively.This showed that the image characteristic parameter PT had significant effect on the estimation of wheat panicle nitrogen content after flowering.(3)The correlation between CT and biomass was the best during the flowering stage.The best estimation model of biomass at 7d after flowering fitted by multiple stepwise regression analysis was y=1051.5*CT2-60418*CT+876864,the R2 of the model was 0.6192,the RMSE was1280.207 kg·hm-2.The optimal estimation model of biomass of 14d after flowering was y=-1408.7*CT2+53704*CT-496833,the R2 of the model was 0.6407,the RMSE was1159.665 kg·hm-2.In addition,the best estimation model for biomass after 21 days of flowering was y=-87.291*LT2+2348.9*LT+33.92 and its R2,RMSE were 0.6165,1457.815 kg·hm-2,respectively.This showed that the image characteristic parameter CT had significant effect on the estimation of wheat biomass after flowering.So,it is feasible to forecast the wheat yield by using canopy temperature. |
PDF Full Text Request