Share Email Print

Proceedings Paper

Modelling the ground-LAI to satellite-NDVI (Sentinel-2) relationship considering variability sources due to crop type (Triticum durum L., Zea mays L., and Medicago sativa L.) and farm management
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Plant Leaf Area Index (LAI) is a key variable for several land surface processes related to vegetation dynamics. Satellite systems offer the opportunity to estimate LAI by remote, allowing to reduce costs for crop monitoring over large areas. Indeed, LAI, is a process variable of several models used for the prediction of crop biomass accumulation, yield and canopy coverage over space and time. Uncertainty of LAI estimation by means of Vegetation Indices extrapolated from multispectral satellite images is related to several factors: (i) crop type (due to the variability of plant architectures, the light extinction coefficient highly variable among crop species), (ii) the effect of environmental covariates, which affects the crop development together with the farm management. This study aimed at evaluating how the ground-LAI (Leaf Area Index) measurements and satellite vegetation indexes from Sentinel 2 reflectance were affected by the variety of crop type and the uncertainties related to different farm managements. We used Sentinel-2 imagery spanning across the growth seasons, over three agricultural sites located in the Mediterranean agroecosystem (Central Italy) and we produced satellite NDVI index to estimate LAI by remote over the sites. In the same period, with a monthly stratified random ground sampling design, the three agricultural farms with prevalent herbaceous crop cover were monitored in terms of ground- LAI variations. The novelty and the added value extracted from the results was that the uncertainty induced by farm management (FARM) variability can be relevant as much as the one caused by crop type (CROP) and seasonality (TIME) variations when ground-LAI measurements and satellite NDVI were analyzed by a set of non-liner regression models.

Paper Details

Date Published: 21 October 2019
PDF: 15 pages
Proc. SPIE 11149, Remote Sensing for Agriculture, Ecosystems, and Hydrology XXI, 111490I (21 October 2019); doi: 10.1117/12.2533446
Show Author Affiliations
M. De Peppo, Scuola Superiore Sant'Anna (Italy)
F. Dragoni, Scuola Superiore Sant'Anna (Italy)
I. Volpi, Scuola Superiore Sant'Anna (Italy)
A. Mantino, Scuola Superiore Sant'Anna (Italy)
V. Giannini, Scuola Superiore Sant'Anna (Italy)
Univ. degli Studi di Sassari (Italy)
F. Filipponi, Scuola Univ. Superiore IUSS (Italy)
A. Tornato, Scuola Univ. Superiore IUSS (Italy)
E. Valentini, Scuola Univ. Superiore IUSS (Italy)
A. Nguyen Xuan, Scuola Univ. Superiore IUSS (Italy)
A. Taramelli, Scuola Univ. Superiore IUSS (Italy)
Istituto Superiore per la Protezione e la Ricerca Ambientale (Italy)
G. Ragaglini, Scuola Superiore Sant'Anna (Italy)

Published in SPIE Proceedings Vol. 11149:
Remote Sensing for Agriculture, Ecosystems, and Hydrology XXI
Christopher M. U. Neale; Antonino Maltese, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?