Showing papers by "Nadine Gobron published in 2004"

Radiation Transfer Model Intercomparison (RAMI) exercise: Results from the second phase

Abstract: [1] The Radiation Transfer Model Intercomparison (RAMI) initiative is a community-driven exercise to benchmark the models of radiation transfer (RT) used to represent the reflectance of terrestrial surfaces. Systematic model intercomparisons started in 1999 as a self-organized, open-access, voluntary activity of the RT modeling community. The results of the first phase were published by Pinty et al. [2001]. The present paper describes the benchmarking protocol and the results achieved during the second phase, which took place during 2002. This second phase included two major components: The first one included a rerun of all direct-mode tests proposed during the first phase, to accommodate the evaluation of models that have been upgraded since, and the participation of new models into the entire exercise. The second component was designed to probe the performance of three-dimensional models in complex heterogeneous environments, which closely mimic the observations of actual space instruments operating at various spatial resolutions over forest canopy systems. Phases 1 and 2 of RAMI both confirm not only that a majority of the radiation transfer models participating in RAMI are in good agreement between themselves for relatively simple radiation transfer problems but also that these models exhibit significant discrepancies when considering more complex but nevertheless realistic geophysical scenarios. Specific recommendations are provided to guide the future of this benchmarking program (Phase 3 and beyond).

Canopy structure parameters derived from multi-angular remote sensing data for terrestrial carbon studies

Abstract: Recent studies have highlighted the importance of vegetation structure, both in the context of landscape dynamics and with regard to ecosystem productivity. This paper addresses the feasibility to retrieve information on canopy structure on the basis of quasi-simultaneous multi-spectral and multi-directional remote sensing measurements from space. After a brief summary of both active and passive remote sensing approaches that are commonly used to address vegetation structure retrievals, this contribution focuses on the state-of-the-art in physically based interpretations relating the anisotropy of multi-directional reflectance measurements to the structure and heterogeneity of the underlying surface. New sets of ecology-oriented parameters are identified that permit a geophysical interpretation of the directional signature of the surface leaving radiation field. The availability of such terrestrial surface structure information, at the within-pixel scale and for the entire globe, will undoubtedly lead to better estimates of ecosystem productivity, carbon stocks and fluxes, as well as changes thereof.

Synergy between 1‐D and 3‐D radiation transfer models to retrieve vegetation canopy properties from remote sensing data

Abstract: [1] We devise a computer efficient and flexible inversion technique to retrieve vegetation canopy parameters, in particular the Leaf Area Index, from the radiance field emerging at the top of a structurally heterogeneous systems overlying an anisotropic spatially uniform surface background. The proposed inversion strategy focuses on a reanalysis of multiangle and multispectral measurements unhindered by the many specific constraints imposed by the operational application of the current algorithms and their associated limitations on data staging. This technique capitalizes on the decoupling between contributions due to the canopy only and those invoking the background reflectance properties. These contributions are decomposed into the wavelength dependent and independent contributions. A quasi-linear relationship is thus obtained between the radiance/reflectance emerging from the top of the canopy layer and the background reflectance. Although all individual contributions can be estimated from accurate three-dimensional radiation transfer models, we propose appropriate approximations in order to estimate the minor terms. These approximations exploit the relatively limited dependency exhibited by these relatively smaller contributions with respect to the azimuthal coordinate. Moreover, additional mathematical developments are proposed to further approximate these terms by their corresponding solutions obtained in the limit case of a plane-parallel turbid medium scenario. They require defining effective values of the state variables entering the plane-parallel turbid medium model. The resulting reflectance of a three-dimensional spatially heterogeneous vegetation layer is driven by a sum of contributions that can be precomputed offline on the basis of the three-dimensional and plane-parallel homogeneous turbid medium model capabilities. The decoupling of the intrinsic vegetation and the background contributions allows many of the contributions to be precomputed and stored in look-up tables. This development yields a simple and computer efficient inversion scheme that allows us to jointly retrieve the values of the main vegetation layer attributes and the underlying background radiative properties. Demonstration tests based on actual multiangular and multispectral data set are currently being investigated.

Meris global vegetation index: evaluation and performance

Abstract: This paper overviews the evaluation and the performance of the Medium Resolution Imaging Spectrometer (MERIS) Global Vegetation Index (MGVI) algorithm which is implemented in the MERIS ground segment as the main land surface product, also called the Top Of Atmosphere Vegetation Index (TOAVI). MGVI values represent the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) which act as an integrated indicator of the state of the plant canopy and can be retrieved by remote sensing techniques with acceptable accuracy. FAPAR can be used to quantify the presence of vegetation with good reliability on a global scale. The retrieval algorithm is optimized to be insensitive to atmospheric and angular effects as well as sensor spectral properties. The evaluation of the algorithm consists of comparisons of daily FAPAR values time series derived from MERIS against values derived from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) data acquired over local sites in Europe during the year 2003. Scatter-plots of contemporaneous product values are analyzed, taking illumination and viewing geometries into account. We also investigate time series of FAPAR to demonstrate the feasibility of monitoring vegetation in the context of land application.

Monitoring FAPAR over land surfaces with remote sensing data

Abstract: Temporal changes of terrestrial vegetation have traditionally been monitored using empirical remote sensing tools, which are sensitive to perturbations as well as to the spectral properties of the sensor. Advances in the understanding of radiation transfer theory, and the availability of higher performance modern instruments, have led to the development of physically-based inverse methods to derive biogeophysical products. Jointly, these developments allow the retrieval of reliable, accurate information on the state and evolution of terrestrial environments. A series of optimized algorithms has been developed to document biogeophysical variables, and in particular to estimate the Fraction of Absorbed Photosynthetically Active Radiation (FAPAR) from a variety of optical instruments. As a result, monitoring managed (e.g., agriculture) or natural ecosystems will benefit from the availability of local, regional and global time series of remote sensing products such as FAPAR. This paper outlines the methodology and exhibits selected results in the form of temporal composites derived from the SeaWiFS sensor.

Analysis of multi-angular data to retrieve indicators of ecosystem structure

Abstract: Vegetation structure significantly impacts the degree of anisotropy of the scattererd radiation field. The proper analysis of multiangular data, such as those provided by the Multi-angle Imaging SpectroRadiometer (MISR) instrument on board Terra, could thus in principle yield statistical information on the structure of the observed environment. Preliminary investigation in this direction suggests that useful information on the heterogeneity of the vegetation can be retrieved at the subpixel scale.