Work

Retrieval of Leaf Area Index Using Sentinel-2 Imagery in a Mixed Mediterranean Forest Area

Year: 2020

Type: article

Abstract: Leaf area index (LAI) is a crucial biophysical indicator for assessing and monitoring the structure and functions of forest ecosystems. Improvements in remote sensing instrumental characteristics and the availability of more efficient statistical algorithms, elevate the potential for more accurate models of vegetation biophysical properties including LAI. The aim of this study was to assess the spectral information of Sentinel-2 MSI satellite imagery for the retrieval of LAI over a mixed forest ecosystem located in northwest Greece. Forty-eight field plots were visited for the collection of ground LAI measurements using an ACCUPAR LP-80: PAR & LAI Ceptometer. Spectral bands and spectral indices were used for LAI model development using the Gaussian processes regression (GPR) algorithm. A variable selection procedure was applied to improve the model’s prediction accuracy, and variable importance was investigated for identifying the most informative variables. The model resulting from spectral indices’ variables selection produced the most precise predictions of LAI with a coefficient of determination of 0.854. Shortwave infrared bands and the normalized canopy index (NCI) were identified as the most important features for LAI prediction.

Source: ISPRS International Journal of Geo-Information

DOI: 10.3390/ijgi9110622

Authors: Irene Chrysafis, Georgios Korakis, Apostolos P. Kyriazopoulos, Giorgos Mallinis

Institutions: Democritus University of Thrace, Aristotle University of Thessaloniki

Language: English

Topic: Remote Sensing in Agriculture

Subfield: Ecology

Field: Environmental Science

Domain: Physical Sciences

SDG: Life on land

Open Access status: gold

Estimated APC paid: $1,056