Retrieval of biophysical vegetation products from RapidEye imagery

For operational applications in the agricultural sector canopy biophysical products are required at a high spatial and temporal resolution. In this study, recently available multispectral RapidEye data were tested for their operational suitability in the Italian Campania region. For this purpose, two model inversion methods and two commonly used vegetation indices were applied to estimate leaf area index (LAI), canopy (CCC) and leaf chlorophyll content (LCC) from a range of crops. The physically based approaches outperformed the empirical methods, with a slightly higher retrieval accuracy of the look-up table (LUT) than of the neural network (NN) approach. However, the NN method performs much faster, rendering it potentially more suitable for application in large areas. The empirical models showed dependencies of sensor and crops, but still performed reasonable in the estimation of LAI and CCC. Results demonstrated the suitability of RapidEye sensor data to retrieve canopy characteristics. To foster the use of physically based approaches, providers of image processing software should add modules for direct and inverse modelling.

VUOLO Francesco;  ATZBERGER Clement;  RICHTER Katja;  D'URSO Guido;  DASH Jadunandan; 

2010-10-06

International Society for Photogrammetry and Remote Sensing (ISPRS)

JRC55768

http://www.isprs.org/proceedings/XXXVIII/part7/a/comm7a.html,    https://publications.jrc.ec.europa.eu/repository/handle/JRC55768,