Monitoring Nitrogen Status And Physicochemical Parameters With Hyperspectral Sensing In Winter Wheat

This study discusses the change law of hyperspectral characteristics of wheat canopy bycomprehensive application of hyperspectral remote sensing and lab analysis in differentnitrogen treatments and growth stages, and establishes the hyperspectral estimation models ofnitrogen content, chlorophyll content, and LAI were established. The purpose of the researchis to monitor winter wheat canopy chlorophyll content and leaf area index and nitrogenNutrition Diagnosis, and guide the farming measures. The main conclusions are as follows:(1) The winter wheat canopy leaf nitrogen content, chlorophyll content and leaf areaindex significantly increased with enhancing nitrogen levels and had significant differences inall stages. The spectral reflectance decreased with the increase of nitrogen levels in the visiblespectrum, while on the contrary it increased in near infrared spectrum; The red edge positioncaused red shift, red edge amplitude and red edge area increased gradually.(2) There were higher correlationship between original spectrum, derivative spectrumand canopy leaf nitrogen content, chlorophyll content and LAI at different growth stages. Theoriginal spectral reflectance of different growth stages was negatively correlated with leafnitrogen content in visible spectrum. In the “red stage”, the negative correlation turned to apositive one and reached maximum at red edge shoulder; The correlation coefficients ofderivative spectrum and leaf nitrogen content was negatively correlated in green andpositively in red. The correlation coefficients of original spectral reflectance and chlorophyllcontent、 LAI was separated by700nm, the correlation coefficients of original spectralreflectance and chlorophyll content was negatively correlated in green-turning and jointingstages and positively correlated in heading and filling stages while less than700nm; thecorrelation coefficients was positively correlated in green-turning and filling stages whilejointing and heading stages was negative of original spectral reflectance and LAI. thecorrelation coefficients more than700nm were basically similar and changed steadily, thecorrelation turned to a positive one and reached maximum at red edge shoulder. Thecorrelation coefficients of derivative spectrum and chlorophyll content、 LAI were steadilyand significantly negative in red while steadily and significantly positive in red.(3) Through the regressive analysis of winter wheat canopy leaf nitrogen content,chlorophyll content and LAI and different sensitive bands, spectrum feature parameters and vegetation index and choose the best to establish the estimation models. The simulative effectand precision of established models was tested by self specie and cross specie to obtained thebest models with reproducibility and universality.