Monitoring Nitrogen Status And Growth With Hyper Spectral In Early Rice

Rice is the main food crop in our country, through real-time monitoring of nutrition and growth conditions of rice plants by spectroscopic techniques; it can provide certain theoretical basis and technical for rice nondestructive monitoring and precise management.Rice plants under different nitrogen management will vary in the growth vigor,consequently forming certain spectral characteristics. Based on the inversion of the spectral characteristics of rice nitrogen nutrition status, provide the basis for the diagnosis of nitrogen nutrition. This study focuses on early rice, uses a combination of leaf and canopy reflectance spectroscopy and physiological and biochemical indexes of testing technology,and then analyzes the relationship between Spectral reflection characteristic of early rice and nitrogen content in leaves, biomass and leaf-area index under different levels of nitrogen, finally chooses spectral parameters which are sensitive to agronomic parameters and builds the corresponding model for monitoring, the main research results are as follows:(1) Through analyzing spectral reflectance characteristics of leaves and canopy of the early rice, it can be found that, with the growth of rice, the canopy spectral reflectance in near infrared region reaches the maximum at the heading stage, indicating a certain relationship between the canopy spectral reflectance the growth parameters of rice, like leaf –area index and biomass. With the increase of the nitrogen, leaf spectral reflectance reduces in the visible region, this is because with the increase of nitrogen, photosynthesis increases and absorption of red and blue light enhances, leading to the reflectivity decreases gradually. These shows that leaf spectrum has certain effect on estimating nitrogen content, and there is close relationship between spectral characteristics of rice canopy and leaf-area index and biomass, providing certain theoretical basis for the feasibility of rice growth monitoring.(2) Based on the analysis of the variation of leaf nitrogen content in early rice under different nitrogen levels, this study researches on the relationship between leaf nitrogen content and leaf spectral reflectance with determines the sensitive wave band. The result shows that, the correlation coefficient for largest derivative value(Dr) within red edge and the nitrogen content reaches the maximum. Through correlation coefficient between leaf nitrogen content and the first derivative spectral reflectance,it is determine sensitive band of about 740 nm. Finally, with the first derivative of spectral reflectance at wavelength of740 nm for the results of the independent variables to build a power equation model is better than other bands fit. By checking that the RMSE is 0.52, the RPD is 1.54.(3) By analyzing the variation of rice growth parameters at different growth stagesnitrogen levels with advancing, this study researches on the relationship between rice leaf area index and biomass with canopy spectral reflectance and spectral parameters. The result shows that, LAI variation in rice under different nitrogen levels roughly curve showing the variation of parabola. And rice LAI sensitive in the near infrared region of the main band. Among three indices NDVI(760.560)、NDVI(810.560)、RVI(810.560)the model accuracy higher than others. The coefficient of determination is greater than 0.7,the RPD in the midst of 1.4 to 1.8. Rice with high biomass(SDr-SDb)/(SDr+SDb) and SDr/ SDy established model accuracy. The coefficient of determination is greater than 0.8, the RPD greater than 1.5. This suggests that use the normalized difference vegetation index and ratio vegetation index can better estimate the biomass of rice.