EPS 10-days Leaf Area Index (ETLAI, LSA-406)

NRT Products available since Jan 2015


Leaf Area Index (LAI) is a dimensionless variable [m2/m2], which defines an important structural property of a plant canopy. It is defined as half the total area of green elements per unit horizontal ground area accounting for the amount of green vegetation that absorbs or scatters solar radiation.

It provides complementary information to the FVC, accounting for the surface of leaves contained in a vertical column normalized by its cross-sectional area. It defines thus the area of green vegetation that interacts with solar radiation determining the remote sensing signal, and represents the size of the interface between the vegetation canopy and the atmosphere for energy and mass exchanges.

LAI is thus a necessary input for Numerical Weather Prediction (NWP), regional and global climate modeling, weather forecasting and global change monitoring. Besides, the LAI is relevant for Land Biosphere Applications such as agriculture and forestry, environmental management and land use, hydrology, natural hazards monitoring and management, vegetation-soil dynamics monitoring and drought conditions.


Product Documentation

This pre-operational product is documented in the Algorithm Theoretical Basis Document (ATBD), Product User Manual document (PUM) and the Product Output Format document (POF) The validation results for this product are available in the (VR) document.

Please see Product Peer-Review publications in References.

The use of LSA SAF products in publications is kindly requested to be duly acknowledged:
LAI and MTLAI were provided by the EUMETSAT Satellite Application Facility on Land Surface Analysis (LSA SAF; Trigo et al., 2011)

Trigo, I. F., C. C. DaCamara, P. Viterbo, J.-L. Roujean, F. Olesen, C. Barroso, F. Camacho-de Coca, D. Carrer, S. C. Freitas, J. García-Haro, B. Geiger, F. Gellens-Meulenberghs, N. Ghilain, J. Meliá, L. Pessanha, N. Siljamo, and A. Arboleda, 2011: The Satellite Application Facility on Land Surface Analysis. Int. J. Remote Sens., 32, 2725-2744, doi: 10.1080/01431161003743199

The global LAI product is currently generated on a 10-days basis at the spatial resolution from the Advanced Very High Resolution Radiometer (AVHRR) sensor onboard the MetOp (Meteorological–Operational) satellite constellation also known as EUMETSAT Polar System (EPS).

The product is based on the three short-wave channels (VIS 0.6 µm, NIR 0.8 µm, SWIR 1.6 µm) using as input the atmospherically corrected cloud-cleared k0 BRDF (Bi-directional Reflectance Distribution Function) product derived from a parametric BRDF model (Roujean et al. 1992; Geiger et al., 2016) and atmospherically corrected cloud-cleared k0 BRDF product.

The AVHRR based ten-day vegetation products are generated pixel-by-pixel at a global scale, inheriting the temporal and spatial characteristics of the EPS ten-day albedo (ETAL) product, which is obtained through composite periods of 20 days (Geiger et al., 2016).

The LAI product is expressed in the range from 0  to 7 [m2/m2]. 

Algorithm Description

The algorithm relies on a Gaussian Processes regression multi-output algorithm (GPRmulti) within a hybrid retrieval approach and is fully described in García-Haro et al. (2018). The approach consists in running first the PROSAIL radiative transfer model to build a database of reflectance and associated biophysical parameters representing a broad set of canopy parameterizations.

The generated simulations are then used to train a non-linear non-parametric regression model through multi-output machine learning approaches for the joint retrieval LAI, FVC and FAPAR maps globally from corresponding EPS surface reflectance data.  

The GPRmulti method outperformed other multi-output kernel-based and neural network methods in terms of stability, accuracy, and robustness. The optimal addition a moderate amount of noise in simulations has increased the accuracy of the estimates, avoiding overfitting and producing more stable solutions.


Data Characteristics

The LAI ten-day vegetation products are generated pixel-by-pixel at a global scale, inheriting the temporal and spatial characteristics of the EPS ten-day albedo (ETAL) product, which is obtained through composite periods of 20 days (Geiger et al., 2016). The products are level 3 full globe rectified images in sinusoidal projection, centered at (0oN, 0oW), with a resolution of 1.1km×1.1km.

The timeslot in the filename of this product corresponds to the last day of the 20-day time-compositing period. For example, the filename HDF5_LSASAF_M01-AVHR_ETLAI_GLOBE_201611250000 with day of production 25th November corresponds to the period November 6th-25th, 2016.

The FVC product, its error estimate and the processing flag are disseminated in HDF5 format. The relevant information concerning the data fields is included in the HDF5 attributes.


Product Uncertainties

Automatic Quality Control (QC) is performed on EPS based vegetation (LAI, FVC and FAPAR) products and the quality information is provided on a pixel basis. QC contains general information about input data quality, and specific information related with the limits of application.

The error estimate define the confidence level of the different products on the basis of the theoretical model assumption and considering the statistical uncertainties of the observations and model parameters, and is a quantitative uncertainty estimate which is especially useful for data assimilation applications.

Geiger, B., Carrer, D., Hautecoeur, O., Franchistéguy, L., Roujean, J.-L., Catherine Meurey, X. C., Jacob, G. (2016). Algorithm Theoretical Basis Document (ATBD). Land Surface Albedo PRODUCTS: LSA-103 (ETAL).

García-Haro, F.J., M. Campos-Taberner, J. Muñoz-Marí, Valero Laparra, F. Camacho, J. Sanchez-Zapero, G Camps-Valls, (2018), Derivation of global vegetation biophysical parameters from EUMETSAT Polar System,  ISPRS Journal of Photogrammetry and Remote Sensing, 139: 57-74. https://doi.org/10.1016/j.isprsjprs.2018.03.005

Roujean, J.L., M. Leroy and P.Y. Dechamps, (1992). A bidirectional reflectance model of the earth's surface for the correction of remote sensing data. Journal of Geophysical Research, 97 (D18), pp. 20455-20468.