Monitoring Water Status With Hyperspectral Sensing In Wheat

Fast and non-destructive monitoring of crop water status is an important research area in agricultural sensing. It is extremely important to the precise irrigation and crop production. The primary objective of this study was to explore the optimum wavebands, spectral indices and quantitative models for estimating leaf water content (LWC), plant water content (PWC) and canopy leaf water content (CLWC) through systematic extraction of hyperspectral information at canopy and leaf levels, on the basis of multiple field experiments under varied water levels management and cultivars in wheat. The anticipated results would provide new wavebands choice for manufacturing portable water monitoring instrument and utilizing space-borne remote sensing data, and thus assist in real-time estimation and precise diagnosis of plant water in wheat.Obtained the major growth stages of wheat canopy reflectance spectral and plant water content (PWC), canopy leaf water content (CLWC) of the data layer under the different soil water conditions. Then a systematic analysis was undertaken on quantitative relationships of PWC and CLWC to major hyperspectral indices composed of any two wavebands with original, first derivative, reciprocal and antilog reflectance within the full spectral range of350～2500nm. The results showed that sensitive to moisture content of wheat reflectance spectra mainly in the visible and near infrared region. The optimum PWC monitoring models were NDVI(R836,R793) and RVI(Rg50,R780) by original spectral, NDVI(RC446,RC485) and RVI(RCg37,RC793) by reciprocal spectral; the optimum CLWC monitoring model was RVI(Rg93,Rgo5) by original spectral. Analysis the different types of canopy spectral data by the systematic precise methods and construct the monitoring models of PWC and CLWC, the origin spectral models and reciprocal spectral models performed better than the first derivative models, and antilog models performed the worst of all. The effective of PWC monitor models was better than CLWC monitor models.Relationship between different position leaf water content (LWC) and single leaf reflectance spectral was studied. A systematic analysis was undertaken on quantitative relationships of LWC to major hyperspectral indices composed of any two wavebands with original, first derivative, reciprocal and antilog reflectance within the full spectral range of350-2500nm.The results showed that the optimum LWC monitoring models were NDVI(R1532,R1621) and RVI(R1532,R1623) by original spectral, NDVI(RC1621,RC1532) and RVI(RC1622,RC1532) by reciprocal spectral. Analysis the different types of canopy spectral data by the systematic precise methods and construct the monitoring models of LWC, the origin spectral models and reciprocal spectral models performed better than the first derivative models, and antilog models performed the worst of all. LWC in the top position leaves were significantly related to leaf spectral reflectance. The leaf combination of1st leaf and2nd leaf construct water monitor models is the best of all leave positions.