An investigation into robust spectral indices for leaf chlorophyll estimation |
| |
| Authors: | Russell Main Moses Azong Cho Renaud Mathieu Martha M O’Kennedy Abel Ramoelo Susan Koch |
| |
| Institution: | aThe Council for Scientific and Industrial Research, Natural Resources and Environment Department, Ecosystems Earth Observation Group, P.O. Box 395, Pretoria 0001, South Africa;bThe Council for Scientific and Industrial Research, Biosciences Department, P.O. Box 395, Pretoria 0001, South Africa;cPlant Protection Research Institute, Agricultural Research Council, Private Bag X134, Pretoria 0001, South Africa |
| |
| Abstract: | Quantifying photosynthetic activity at the regional scale can provide important information to resource managers, planners and global ecosystem modelling efforts. With increasing availability of both hyperspectral and narrow band multispectral remote sensing data, new users are faced with a plethora of options when choosing an optical index to relate to their chosen or canopy parameter. The literature base regarding optical indices (particularly chlorophyll indices) is wide ranging and extensive, however it is without much consensus regarding robust indices. The wider spectral community could benefit from studies that apply a variety of published indices to differing sets of species data. The consistency and robustness of 73 published chlorophyll spectral indices have been assessed, using leaf level hyperspectral data collected from three crop species and a variety of savanna tree species. Linear regression between total leaf chlorophyll content and bootstrapping were used to determine the leafpredictive capabilities of the various indices. The indices were then ranked based on the prediction error (the average root mean square error (RMSE)) derived from the bootstrapping process involving 1000 iterative resampling with replacement. The results show two red-edge derivative based indices (red-edge position via linear extrapolation index and the modified red-edge inflection point index) as the most consistent and robust, and that the majority of the top performing indices (in spite of species variability) were simple ratio or normalised difference indices that are based on off-chlorophyll absorption centre wavebands (690–730 nm). |
| |
| Keywords: | Leaf level reflectance Leaf chlorophyll Red-edge Vegetation indices Photosynthetic activity |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
