The Hyper-spectral Inversion Of Physicochemical Parameters In Dry-hot Wind Disaster Which Has An Effect On Winter Wheat

Dry-hot wind disasters seriously affect winter wheat yields,but the monitoring of dry-hot wind disasters is still lagging behind.In this study,winter wheat was suffered from different degrees of hot-dry wind in the early filling stage.Measure the physicochemical index of wheat by chemical methods and measure its’hyperspectral reflectance accordingly.Analyze the correlations between physicochemical index and degree of dry-hot wind,then screen out the index which is sensitive to different degree of dry-hot wind.Extract the most relevant feature information between sensitive index and spectrum information which consumed by the spectral analysis method based on the original spectrum,the first derivative spectrum,optimization NDVI and RVI.And four kinds of functions are used to construct hyperspectral models upon sensitive physical.Compare models by different methods and select the most optimal model which can sensitively indicate the wheat suffered by different degree of dry-hot wind.The result can provide technical support to rapid monitor the dry-hot wind disaster.The main conclusions of this study are as follows:（1）Under the different degrees of dry-hot wind,the difference of the equivalent water thickness（EWT）,POD,CAT activity and malondialdehyde（MDA）content in winter wheat is significant（P<0.05）.Compared with the wheat with no disaster affected,physicochemical index in the wheat under heavy dry-hot wind have large changed,EWT is about 33%,POD activity is about 25%,CAT activity is about 14%,and MDA content is about 2.5 times.The difference in the moisture content（FMC）between heavy dry hot air and other conditions is about 12%.SPAD and SOD activity only showed small changes with the degree of dry-hot wind.SPAD only changed about 2%,SOD activity changed about 3%,and there was no significant correlation with different degrees of dry hot wind effects（P>0.05）.Therefore,FMC,EWT,POD,CAT activity and MDA content were selected as the physicochemical parameters to analyze winter wheat under different degrees of dry-hot wind.（2）Under the different degrees of dry-hot wind,near the infrared wavelength band,the reflectance of the original spectrum under heavy dry-hot wind is about 0.03⁰.04higher than that under normal conditions.And the reflectance at 800¹300nm can clearly distinguish wheat under different degrees of dry-hot wind,the difference between 0.01⁰.03.Under the heavy dry-hot wind,the first-derivative in yellow-band is 0.0006 higher than that under other levels of stress,thus it can distinguish whether the wheat is exposed to heavy dry-hot wind.Under the three levels of dry-hot wind,the first-derivative in red-band is separated.And under different stresses,the difference is 0.0003.The amplitude and area of the blue-edge are the most sensitive to different degrees of dry-hot wind among the“three edges”parameter.Under the different degrees of dry-hot wind,the differences of amplitude and area are0.0005 and 0.012,respectively.The three spectral analysis methods are effective to distinguish wheat under different degrees of dry-hot wind,among which the original spectrum and blue edge parameters are the most efficient way to distinguish.（3）Correlate the selected physicochemical parameters with hyper-spectral data under different treatments.By analyzing the characteristic information,select the parameters that have the highest correlation between the characteristic spectral information and the physicochemical parameters.The results are as follows:R’1917 is selected as the feature information which has the highest accuracy in the FMC estimation（R²=0.9029）,and the estimation model isC_FMC（28）24506X_R￠1917（10）86.269;SDy/SDb is selected as the feature information which has the highest precision in the EWT estimation model（R²=0.8341）,and the estimation model isC_EWT（28）-0.0229X_SDy/SDb-0.0012;R’1799 is selected as the characteristic information which has the highest precision in the POD activity estimation model（R²=0.3882）,and the estimation model isC_POD（28）9′10⁷X_R￠1799（10）26284;R’1674 is selected as the characteristic information which has the highest precision in the CAT activity estimation（R²=0.4079）,and the estimation model is C_CAT（28）-10⁶X_R￠1674（10）153.08;R’1925 is selected as the characteristic information which has the highest precision in the MDA content estimation（R²=0.7322）,the estimation model isC_MDA（28）-388769X_R￠1925（10）0.9553.In this study,the sensitive physical and chemical parameters of winter wheat leaves underwent dry-hot wind stress are summarized and the hyperspectral model was established.So as to provide a theoretical basis for fast and nondestructive monitoring of whether winter wheat suffers dry-hot wind disaster.