Retrieval of spruce leaf chlorophyll content from airborne image data using continuum removal and radiative transfer

Abstract

We investigate combined continuum removal and radiative transfer (RT) modeling to retrieve leaf chlorophyll a & b content (Cab) from the AISA Eagle airborne imaging spectrometer data of sub-meter (0.4 m) spatial resolution. Based on coupled PROSPECT-DART RT simulations of a Norway spruce (Picea abies (L.) Karst.) stand, we propose a new Cab sensitive index located between 650 and 720 nm and termed ANCB650-720. The performance of ANCB650-720 was validated against ground-measured Cab of ten spruce crowns and compared with Cab estimated by a conventional artificial neural network (ANN) trained with continuum removed RT simulations and also by three previously published chlorophyll optical indices: normalized difference between reflectance at 925 and 710 nm (ND925&710), simple reflectance ratio between 750 and 710 nm (SR750/710) and the ratio of TCARI/OSAVI indices. Although all retrieval methods produced visually comparable Cab spatial patterns, the ground validation revealed that the ANCB650-720 and ANN retrievals are more accurate than the other three chlorophyll indices (R2 = 0.72 for both methods). ANCB650-720 estimated Cab with an RMSE = 2.27 μg cm− 2 (relative RRMSE = 4.35%) and ANN with an RMSE = 2.18 μg cm− 2 (RRMSE = 4.18%), while SR750/710 with an RMSE = 4.16 μg cm− 2 (RRMSE = 7.97%), ND925&710 with an RMSE = 9.07 μg cm− 2 (RRMSE = 17.38%) and TCARI/ OSAVI with an RMSE = 12.30 μg cm− 2 (RRMSE = 23.56%). Also the systematic RMSES was lower than the unsystematic one only for the ANCB650-720 and ANN retrievals. Our results indicate that the newly proposed index can provide the same accuracy as ANN except for Cab values below 30 μg cm− 2, which are slightly overestimated (RMSE = 2.42 μg cm− 2). The computationally efficient ANCB650-720 retrieval provides accurate high spatial resolution airborne Cab maps, considerable as a suitable reference data for validating satellite-based Cab products. Disciplines Medicine and Health Sciences | Social and Behavioral Sciences Publication DetailsMalenovsky, Z., Homolova, L., Zurita-Milla, R., Lukes, P., Kaplan, V., Hanus, J., Gastellu-Etchegorry, J. & Schaepman, M. E. (2013). Retrieval of spruce leaf chlorophyll content from airborne image data using continuum removal and radiative transfer. Page | 2 Abstract 26We investigate combined continuum removal and radiative transfer ( considerable as a suitable reference data for validating satellite-based C ab products. 48

Discrete Anisotropic Radiative Transfer (DART 5) for Modeling Airborne and Satellite Spectroradiometer and LIDAR Acquisitions of Natural and Urban Landscapes

“…DART forward simulations of vegetation reflectance were successfully verified by real measurements [32] and also cross-compared against a number of independently designed 3D reflectance models (e.g., FLIGHT [26], Sprint [33], Raytran [27]) in the context of the RAdiation transfer Model Intercomparison (RAMI) experiment [34][35][36][37][38]. To date, DART has been successfully employed in various scientific applications, including development of inversion techniques for airborne and satellite reflectance images [39,40], design of satellite sensors (e.g., NASA DESDynl, CNES Pleiades, CNES LIDAR mission project [41]), impact studies of canopy structure on satellite image texture [42] and reflectance [32], modeling of 3D distribution of photosynthesis and primary production rates in vegetation canopies [43], investigation of influence of Norway spruce forest structure and woody elements on canopy reflectance [44], design of a new chlorophyll estimating vegetation index for a conifer forest canopy [45], and studies of tropical forest texture [46][47][48], among others. DART creates and manages 3D landscapes independently from the RT modeling (e.g., visible and thermal infrared IS, LIDAR, radiative budget).…”

The Optimal Leaf Biochemical Selection for Mapping Species Diversity Based on Imaging Spectroscopy

“…New airborne remote sensing technologies, especially LiDAR (Light Detection and Ranging), and physical modeling approaches, such as 4-SCALE [49], ACRM (Two-Layer Canopy Reflectance Model) [50], FRT (Forest Reflectance and Transmittance) [51] and PROSPECT-DART (Discrete Anisotropic Radiative Transfer) [52], have been developed. LiDAR is an active remote sensing technology, and its scanners can emit a high-frequency pulse that can penetrate vegetation canopy gaps and, by recording the return time and intensity of backscatter from targets, provide detailed information of both the ground and multiple points within the forest canopy [53][54][55].…”

Detection of chlorophyll and leaf area index dynamics from sub-weekly hyperspectral imagery

“…In addition, the validity of the empirical regressions beyond the training data will need further investigation. A key objective of ongoing work is to fully utilize the rich information content of the hyperspectral data streams by complementing the existing list of explanatory spectral indices (Table 2) with additional hyperspectral metrics, such as those based on continuum removal 12,53,54 .…”