Hyperspectral Characteristics And Chlorophyll Content Estimation Of Rice And Wheat Under Ozone Stress

Rice and wheat are the main food crops in China and their production plays a vital role in national food security.In recent years,there have been frequent incidents of ozone pollution in China.Surface ozone not only damages human health,but also affects the growth and development of crops,thus threatening crop yields.Hyperspectral technology has been widely used in crop production as a real-time,non-destructive and convenient monitoring method to obtain subtle information about plants.The use of hyperspectral techniques to monitor rice and wheat under ozone stress is useful in providing early warning of disasters.However,research on the identification mechanisms and damage assessment of spectral features of crops under ozone stress is still incomplete.Moreover,the impact of future warming on ozone stress monitoring in the context of global warming is still unclear.To this end,ozone fumigation experiments were conducted on rice and winter wheat using the OTC platform and the O3-FACE platform,respectively.By analysing the relationship between chlorophyll content and hyperspectral characteristics of rice and winter wheat in response to ozone stress,ozone stress was identified;by analysing the correlation between spectral parameters and chlorophyll content,and using spectral parameters to construct models for estimating chlorophyll content in rice and wheat,the extent of damage caused by ozone pollution to rice and winter wheat was assessed;by analysing the spectra of winter wheat under warming,ozone and warming-ozone interactions The main results include the following The main results of this paper include the following.（1）The hyperspectral characteristics of rice and wheat under ozone stress changed significantlyThe hyperspectral characteristics of rice under different ozone concentrations varied significantly.The higher the ozone concentration,the higher the green peak in the visible region and the closer the position of the red edge to the visible region.The hyperspectral features of winter wheat under ozone stress also changed significantly at different stages of fertility.In the early stages of ozone stress,winter wheat spectra increased in the near-infrared band,with no significant differences in the visible region;in the later stages,due to the cumulative effect of ozone,the reflectance in the visible region of winter wheat increased,while the reflectance in the near-infrared region remained largely stable.The most obvious changes in the hyperspectral characteristics of winter wheat in the middle and late stages of filling are the period of greatest ozone stress,which is conducive to ozone stress monitoring.（2）New vegetation indices can improve the accuracy of rice chlorophyll content estimation modelsThe accuracy of rice chlorophyll content estimation models constructed using the traditional remote sensing vegetation indices and spectral characteristics was generally low,among which the accuracy of the chlorophyll content estimation model constructed using DD was the best,with R² of 0.54 and RMSE of 1.56/10 mg·m^-2.The model for estimating chlorophyll using the new vegetation index SR610（598 nm,610 nm）has an R² of 0.71 and RMSE of 1.24/10 mg·m^-2.Compared to the traditional remote sensing vegetation index DD,the accuracy is improved by 31.4%and the error is reduced by 20%.（3）Inconsistent models for estimating the optimal chlorophyll content of winter wheat at different scalesCompared with the leaf scale,the canopy scale model for estimating chlorophyll content of wheat was better overall.At the leaf scale,the partial least squares regression model based on spectral parameters was the optimal model with R² of 0.97 and RMSE of 2.8 SPAD,while at the canopy scale,the BP neural network model based on remotely sensed vegetation index and spectral parameters was the optimal model with R² of 0.99 and 0.98,and RMSE of 2.52SPAD and 1.43 SPAD,respectively.（4）Temperature increase has a greater effect on the identification of hyperspectral features in winter wheat under ozone stressCompared with ozone stress,the damage of winter wheat under warming treatment was greater,and the combined warming and ozone effect would aggravate the damage degree of winter wheat,and the spectral curves of winter wheat canopy under warming,ozone and warming and ozone combined effect would show increased"green peak"and blue shift of"red edge position"at around 550 nm and 720 nm.Seriously affect ozone stress monitoring.The difference of index DOD can be used to distinguish the effects of ozone stress and warming in the visible band at heading stage and the near infrared band at early flowering stage of winter wheat.The DOD also can be used to distinguish the effects of ozone stress and warming and ozone combination in the near infrared band at early flowering stage and the visible band at middle and late filling stage of winter wheat.In conclusion,hyperspectral technology can be used to identify rice and wheat damage under ozone stress,and the chlorophyll content estimation model constructed by remote sensing vegetation index and spectral characteristic parameters is highly feasible,which provides technical support for ozone stress assessment.Although temperature increase will affect ozone stress monitoring,it is of great significance for ozone stress monitoring under future global warming to distinguish the effect of temperature increase on ozone stress monitoring by using the distinguishing index.