Translator Disclaimer
Paper
14 October 2015 Application of a regularized model inversion system (REGFLEC) to multi-temporal RapidEye imagery for retrieving vegetation characteristics
Author Affiliations +
Abstract
Accurate retrieval of canopy biophysical and leaf biochemical constituents from space observations is critical to diagnosing the functioning and condition of vegetation canopies across spatio-temporal scales. Retrieved vegetation characteristics may serve as important inputs to precision farming applications and as constraints in spatially and temporally distributed model simulations of water and carbon exchange processes. However significant challenges remain in the translation of composite remote sensing signals into useful biochemical, physiological or structural quantities and treatment of confounding factors in spectrum-trait relations. Bands in the red-edge spectrum have particular potential for improving the robustness of retrieved vegetation properties. The development of observationally based vegetation retrieval capacities, effectively constrained by the enhanced information content afforded by bands in the red-edge, is a needed investment towards optimizing the benefit of current and future satellite sensor systems. In this study, a REGularized canopy reFLECtance model (REGFLEC) for joint leaf chlorophyll (Chll) and leaf area index (LAI) retrieval is extended to sensor systems with a band in the red-edge region for the first time. Application to time-series of 5 m resolution multi-spectral RapidEye data is demonstrated over an irrigated agricultural region in central Saudi Arabia, showcasing the value of satellite-derived crop information at this fine scale for precision management. Validation against in-situ measurements in fields of alfalfa, Rhodes grass, carrot and maize indicate improved accuracy of retrieved vegetation properties when exploiting red-edge information in the model inversion process.
© (2015) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Rasmus Houborg and Matthew F. McCabe "Application of a regularized model inversion system (REGFLEC) to multi-temporal RapidEye imagery for retrieving vegetation characteristics", Proc. SPIE 9637, Remote Sensing for Agriculture, Ecosystems, and Hydrology XVII, 963717 (14 October 2015); https://doi.org/10.1117/12.2196378
PROCEEDINGS
10 PAGES


SHARE
Advertisement
Advertisement
Back to Top