Effects Of Leaf Hair On Estimation Of Chlorophyll Content For Urban Vegetations

Spectral vegetation indices can be used to estimate vegetation leaf chlorophyll contentand other biochemistries. Since spectral vegetation indices could monitor health condition ofvegetation and their environment without damage, they have been widely used. But most ofthe vegetation indices assume that the spectral characteristics of plants are only related to theirbiochemistries, without considering the influence of the hair structure of the leaf surface onleaf spectrum, or the effect of biochemistries estimation with these indices. However, manystudies have shown that the leaf surface hair has a great influence on leaf spectral reflectance.This thesis studied the effect of leaf surface hair on the estimation of leaf chlorophyllcontent by removing the leaf hair with transparent tape. Six species of urban vegetation withdifferent leaf surface hair structure were selected as samples. We found that the hair densityhas an impact on inversion of chlorophyll content using vegetation indices, the greater thehair density, the smaller the correlation coefficient between the vegetation indices and thechlorophyll content. In each plant species, we randomly selected half of the leaf samples asthe training samples to model the chlorophyll content with the ratio or reflectance differenceof spectral bands within400-2500nm. It is found that before and after hair removal, the faraxial plane of Lonicera Ruprechtiana, R²of R₄₆₃/R₄₈₁and R₁₁₈₅/R₂₀₆₅and total leafchlorophyll content were0.864,0.892and0.733,0.736; before and after hair removal ofnearly axial plane Lonicera Ruprechtiana, R²of R₁₁₀₂/R₁₁₄₇and total leaf chlorophyll contentwere0.869,0.898;before and after hair removal of Tilia mandshurica leaves, the R²ofR₁₀₈₄-R₁₁₂₉and total leaf chlorophyll content were0.594,0.600; before and after hair removalof Elaeagnus angustifolia leaves, the R²of R₈₈₆/R₉₁₃, R₁₇₁₄/R₁₇₂₃and total leaf chlorophyllcontent were0.497,0.588and0.679,0.608; before and after hair removal of the far axial planeof Populus alba, the R²of R₁₀₈₄/R₁₁₄₇and total leaf chlorophyll content were0.388,0.372.And through the test of remaining samples, we found that before and after hair removal of faraxial plane of Lonicera ruprechtiana, the RMSE of R₄₆₃/R₄₈₁and R₁₁₈₅/R₂₀₆₅were0.00268,0.00326and0.00634,0.00722; before and after hair removal of neaely axial plane ofLonicera ruprechtiana, the RMSE of R₁₁₀₂/R₁₁₄₇were0.00571、0.00581;before and after hairremoval of the far axial plane of Tilia mandshurica, the RMSE of R₁₀₈₄-R₁₁₂₉were0.00468,0.00372; before and after hair removal of Elaeagnus angustifolia leaf the RMSE ofR₈₈₆/R₉₁₃and R₁₇₁₄/R₁₇₂₃were0.00481、0.00369and0.00354、0.00421; before and after hairremoval of the far axial plane of Populus alba, the RMSE of R₁₀₈₄/R₁₁₄₇were0.00259,0.00303.In comparing correlation coefficient of estimating chlorophyll content by existing vegetationindices and found some vegetation indices of this article performed better in estimatingchlorophyll content. Because their inversion precision before and after hair removing arerobust, they are hair insensivtive vegetation indices in inversion of chlorophyll content.