New Spectral Index For Estimating The Leaf Chlorophyll Content Based On Multiangular Reflectance Factor

Remote sensing technology has been widely used to estimate the biophysical and chemical parameters of vegetation because of its short time-consuming,small loss and high efficiency.As one of the main pigments of green plants,chlorophyll is an important parameter of the biophysical and chemical characteristics of vegetation,and the change of its content can be used to characterize the physiological changes and growth status of plants.Therefore,using remote sensing information to accurately estimate vegetation biochemical parameters has become an important research content of vegetation remote sensing.In previous studies,many spectral indices have been developed to can make good estimates of leaf chlorophyll content?LCC?.There are many differences in leaf surface structure?roughness,fluff,waxy layer,etc.?of different types of plants.Therefore,the reflection information obtained when making remote sensing observations from different directions will be vary,even for leaves with the same chlorophyll content.In reality,when the remote sensing platform acquires reflection information,the reflection information entering the sensor will come from different detection angles.However,most of the existing spectral indices are developed based on the zenith direction or a single detection angle,without considering the effect of angular reflection information in estimating the accuracy of LCC.In this study,three plant species?Pachira macrocarpa,Juglans and Schefflera microphylla Merr.?were selected to observe the microstructure of the leaf surface and measure the multiangle reflection information of the main plane.The results show that the differences in the surface structure of the three plant species not only affect the reflection distribution characteristics of the leaf surface,but also make the multiangle reflection information differ.Among the 32 published spectral indices,the relationship between most of them and the measured LCC is significantly different at different angles,and the sensitivity of different species?blade surface structure?to different observation angles is also different.A few spectral indices,such(R₈₅₀-R₇₁₀)/(R₈₅₀-R680),are less sensitive to the observation angle,and have stable prediction ability for each angle,but the prediction accuracy for individual species is not ideal.It is revealed that most of the spectral indexes obtained based on single angle measurement in the past have instability in the estimation of leaf chlorophyll content in multi-angle measurement results,which vary significantly with the observation angle,resulting in the inability to accurately estimate the leaf chlorophyll content at certain angles.Therefore,in this study,the MDatt index(R₇₂₀-R₇₆₁)/(R₇₂₀-R₆₇₂)was obtained by optimizing the difference ratio index based on multiangle reflectance factor,which not only conforms to the theory of band selection proposed by the predecessors,but also has a high correlation with LCC of different species at different observation angles.When using MDatt index to estimate the LCC of multiangle data,the overall performance is very good,R²=0.97,RMSE=4.06 ?g/cm².Meanwhile,the prediction accuracy for different species from different angles is also very high,R² is greater than 0.95,and RMSE is less than 5?g/cm².At the same time,when using other published data sets to evaluate each spectral index,the MDatt index performs better than other existing spectral indices.Thus,it can be proved that the MDatt index(R₇₆₁-R₆₇₂)/(R₇₂₀-R₆₇₂) is an effective index to estimate the leaf chlorophyll content of different plants more stably,which is less affected by leaf surface structure and less limited by the observation angle.