-
Global Estimates of L-band Vegetation Optical Depth and Soil Permittivity over Snow-covered Boreal Forests and Permafrost using SMAP Satellite Data
Authors:
Divya Kumawat,
Ardeshir Ebtehaj,
Mike Schwank,
Jean-Pierre Wigneron,
Xiaojun Li
Abstract:
This article expands the tau-omega model to properly simulate L-band microwave emission of the soil-snow-vegetation continuum through a closed-form solution of Maxwell's equations, considering the intervening dry snow layer as a loss-less medium. The feasibility and uncertainty of retrieving vegetation optical depth (VOD) and ground permittivity, given the noisy L-band brightness temperatures with…
▽ More
This article expands the tau-omega model to properly simulate L-band microwave emission of the soil-snow-vegetation continuum through a closed-form solution of Maxwell's equations, considering the intervening dry snow layer as a loss-less medium. The feasibility and uncertainty of retrieving vegetation optical depth (VOD) and ground permittivity, given the noisy L-band brightness temperatures with 1 K (1-sigma), are demonstrated through controlled numerical experiments. For moderately dense vegetation canopy and a range of 100--400 $kg.m^{-3}$ snow density, the standard deviation of the retrieval errors is 0.1 and 3.5 for VOD and ground permittivity respectively. Using L-band observations from the Soil Moisture Active Passive (SMAP) satellite, a new data set of global estimates of VOD and ground permittivity are presented over the Arctic boreal forests and permafrost areas during winter months. In the absence of dense ground-based observations of ground permittivity and VOD, the retrievals are causally validated using dependent variables including above-ground biomass, tree height, and net ecosystem exchange. Time-series analyses promise that the new data set can expand our understanding of the land-atmosphere interactions and exchange of carbon fluxes over the Arctic landscape.
△ Less
Submitted 15 July, 2023;
originally announced July 2023.
-
Vision Transformers, a new approach for high-resolution and large-scale map** of canopy heights
Authors:
Ibrahim Fayad,
Philippe Ciais,
Martin Schwartz,
Jean-Pierre Wigneron,
Nicolas Baghdadi,
Aurélien de Truchis,
Alexandre d'Aspremont,
Frederic Frappart,
Sassan Saatchi,
Agnes Pellissier-Tanon,
Hassan Bazzi
Abstract:
Accurate and timely monitoring of forest canopy heights is critical for assessing forest dynamics, biodiversity, carbon sequestration as well as forest degradation and deforestation. Recent advances in deep learning techniques, coupled with the vast amount of spaceborne remote sensing data offer an unprecedented opportunity to map canopy height at high spatial and temporal resolutions. Current tec…
▽ More
Accurate and timely monitoring of forest canopy heights is critical for assessing forest dynamics, biodiversity, carbon sequestration as well as forest degradation and deforestation. Recent advances in deep learning techniques, coupled with the vast amount of spaceborne remote sensing data offer an unprecedented opportunity to map canopy height at high spatial and temporal resolutions. Current techniques for wall-to-wall canopy height map** correlate remotely sensed 2D information from optical and radar sensors to the vertical structure of trees using LiDAR measurements. While studies using deep learning algorithms have shown promising performances for the accurate map** of canopy heights, they have limitations due to the type of architectures and loss functions employed. Moreover, map** canopy heights over tropical forests remains poorly studied, and the accurate height estimation of tall canopies is a challenge due to signal saturation from optical and radar sensors, persistent cloud covers and sometimes the limited penetration capabilities of LiDARs. Here, we map heights at 10 m resolution across the diverse landscape of Ghana with a new vision transformer (ViT) model optimized concurrently with a classification (discrete) and a regression (continuous) loss function. This model achieves better accuracy than previously used convolutional based approaches (ConvNets) optimized with only a continuous loss function. The ViT model results show that our proposed discrete/continuous loss significantly increases the sensitivity for very tall trees (i.e., > 35m), for which other approaches show saturation effects. The height maps generated by the ViT also have better ground sampling distance and better sensitivity to sparse vegetation in comparison to a convolutional model. Our ViT model has a RMSE of 3.12m in comparison to a reference dataset while the ConvNet model has a RMSE of 4.3m.
△ Less
Submitted 22 April, 2023;
originally announced April 2023.
-
High-resolution canopy height map in the Landes forest (France) based on GEDI, Sentinel-1, and Sentinel-2 data with a deep learning approach
Authors:
Martin Schwartz,
Philippe Ciais,
Catherine Ottlé,
Aurelien De Truchis,
Cedric Vega,
Ibrahim Fayad,
Martin Brandt,
Rasmus Fensholt,
Nicolas Baghdadi,
François Morneau,
David Morin,
Dominique Guyon,
Sylvia Dayau,
Jean-Pierre Wigneron
Abstract:
In intensively managed forests in Europe, where forests are divided into stands of small size and may show heterogeneity within stands, a high spatial resolution (10 - 20 meters) is arguably needed to capture the differences in canopy height. In this work, we developed a deep learning model based on multi-stream remote sensing measurements to create a high-resolution canopy height map over the "La…
▽ More
In intensively managed forests in Europe, where forests are divided into stands of small size and may show heterogeneity within stands, a high spatial resolution (10 - 20 meters) is arguably needed to capture the differences in canopy height. In this work, we developed a deep learning model based on multi-stream remote sensing measurements to create a high-resolution canopy height map over the "Landes de Gascogne" forest in France, a large maritime pine plantation of 13,000 km$^2$ with flat terrain and intensive management. This area is characterized by even-aged and mono-specific stands, of a typical length of a few hundred meters, harvested every 35 to 50 years. Our deep learning U-Net model uses multi-band images from Sentinel-1 and Sentinel-2 with composite time averages as input to predict tree height derived from GEDI waveforms. The evaluation is performed with external validation data from forest inventory plots and a stereo 3D reconstruction model based on Skysat imagery available at specific locations. We trained seven different U-net models based on a combination of Sentinel-1 and Sentinel-2 bands to evaluate the importance of each instrument in the dominant height retrieval. The model outputs allow us to generate a 10 m resolution canopy height map of the whole "Landes de Gascogne" forest area for 2020 with a mean absolute error of 2.02 m on the Test dataset. The best predictions were obtained using all available satellite layers from Sentinel-1 and Sentinel-2 but using only one satellite source also provided good predictions. For all validation datasets in coniferous forests, our model showed better metrics than previous canopy height models available in the same region.
△ Less
Submitted 20 December, 2022;
originally announced December 2022.
-
Carbon loss from forest degradation exceeds that from deforestation in the Brazilian Amazon
Authors:
Yuanwei Qin,
Xiangming Xiao,
Jean-Pierre Wigneron,
Philippe Ciais,
Martin Brandt,
Lei Fan,
Xiaojun Li,
Sean Crowell,
Xiaocui Wu,
Russell Doughty,
Yao Zhang,
Fang Liu,
Stephen Sitch,
Berrien Moore III
Abstract:
Spatial-temporal dynamics of aboveground biomass (AGB) and forest area affect the carbon cycle, climate, and biodiversity in the Brazilian Amazon. Here we investigate inter-annual changes of AGB and forest area by analyzing satellite-based annual AGB and forest area datasets. We found the gross forest area loss was larger in 2019 than in 2015, possibly due to recent loosening of forest protection…
▽ More
Spatial-temporal dynamics of aboveground biomass (AGB) and forest area affect the carbon cycle, climate, and biodiversity in the Brazilian Amazon. Here we investigate inter-annual changes of AGB and forest area by analyzing satellite-based annual AGB and forest area datasets. We found the gross forest area loss was larger in 2019 than in 2015, possibly due to recent loosening of forest protection policies. However, net AGB loss was three times smaller in 2019 than in 2015. During 2010-2019, the Brazilian Amazon had a cumulative gross loss of 4.45 Pg C against a gross gain of 3.78 Pg C, resulting in net AGB loss of 0.67 Pg C. Forest degradation (73%) contributed three times more to the gross AGB loss than deforestation (27%), given that the areal extent of degradation exceeds deforestation. This indicates that forest degradation has become the largest process driving carbon loss and should become a higher policy priority.
△ Less
Submitted 15 June, 2022;
originally announced June 2022.
-
Modelisation Des Effets De la Rugosite Sur L'Etude De L'Humidite Des Sols Par Radiometrie Micro-Ondes. Application a la Mission Spatiale Smos
Authors:
François Demontoux,
Heather Lawrence,
Jean Pierre Wigneron
Abstract:
As part of the SMOS mission (Soil Moisture and Ocean Salinity) IMS and EPHYSE laboratories are involved in the study and validation of the inversion algorithm LMEB. In the framework of this study, we developed a model using the numerical modeling software-HFSS ANSOFT. It allows the computation of the emissivity of multilayer systems (soil-litter for example). We have improved our approach to inc…
▽ More
As part of the SMOS mission (Soil Moisture and Ocean Salinity) IMS and EPHYSE laboratories are involved in the study and validation of the inversion algorithm LMEB. In the framework of this study, we developed a model using the numerical modeling software-HFSS ANSOFT. It allows the computation of the emissivity of multilayer systems (soil-litter for example). We have improved our approach to incorporate new parameters that can have a significant effect on the measurements. Indeed, so far the effect of soil roughness associated with the litter has not been studied. Moreover, the thickness of litter is never constant and we must therefore introduce a realistic profile of litter thickness. This article presents our work to integrate these profiles in our numerical model. These profiles may be derived from measurements (litter thickness profiles or soil roughness profiles) or calculations (roughness profile). In both cases the input data of our model are files of XYZ points representing our profile.
△ Less
Submitted 25 June, 2009;
originally announced June 2009.
-
Amélioration de l'étude de l'étude de l'humidité de sols par radiométrie. Caractérisation et modélisation diélectriques de profils géologiques
Authors:
François Demontoux,
Gilles Ruffié,
Jean Pierre Wigneron,
Maria-José Escorihuela
Abstract:
The surface soil moisture is a key variable to describe the exchange of water and energy between the land and the atmosphere. In hydrology, and meteorology, the amount of water in the upper soil layers permits (0-5 cm from the surface) the evaluation of the relationship between the real evaporation and the potential evaporation of the bare soil. It is also possible to determine the distribution…
▽ More
The surface soil moisture is a key variable to describe the exchange of water and energy between the land and the atmosphere. In hydrology, and meteorology, the amount of water in the upper soil layers permits (0-5 cm from the surface) the evaluation of the relationship between the real evaporation and the potential evaporation of the bare soil. It is also possible to determine the distribution of rainfall or other variables such as hydraulic conductivity. Studies have shown that microwave sensors could be used to scan the surface soil moisture. The solution choose by the team associated with the SMOS mission (Soil Moisture and Ocean Salinity) is to use a radiometer (1.4 GHz) to identify soil microwave emissions . The measurements are made on the site of CESBIO in Toulouse where a 1.4 GHz radiometer is installed . The effect of vegetation cover, soil temperatures, snow cover, topography or moisture variations have an important role in the broadcast microwave surface. Other parameters such as the presence of inclusions in the ground (holes, stones), or soil texture can also disrupt measurements. The aim of the work that we present in this paper is to develop a numerical model to simulate complex geological structures. This model must take into account all parameters that can affect the equivalent permittivity measured by the radiometer (surface, moisture variation ,... inclusions). The purpose of this model is to calculate the equivalent permittivity of these geological structures in order to be able to associate with each radiometer measurement an equivalent moisture.
△ Less
Submitted 18 January, 2008;
originally announced January 2008.
-
Lband radiometric behaviour of pine forests for a variety of surface moisture conditions
Authors:
Jennifer Grant,
Jean Pierre Wigneron,
A. A. Van De Griend,
François Demontoux,
Gilles Ruffié,
A. Della Vecchia,
N. Skou,
Bénédicte Le Crom
Abstract:
From July-December 2004 the experimental campaign 'Bray 2004' was conducted in the coniferous forest of Les Landes near Bordeaux, France, using a multi-angle L-band (1.4 GHz) radiometer to measure upwelling radiation above the forest. At the same time, ground measurements were taken of soil and litter moisture content. This experiment was done in the context of the upcoming SMOS mission in order…
▽ More
From July-December 2004 the experimental campaign 'Bray 2004' was conducted in the coniferous forest of Les Landes near Bordeaux, France, using a multi-angle L-band (1.4 GHz) radiometer to measure upwelling radiation above the forest. At the same time, ground measurements were taken of soil and litter moisture content. This experiment was done in the context of the upcoming SMOS mission in order to improve our understanding of the behaviour of the L-band signal above forested areas. Very little information exists on this subject at the moment, especially for varying hydrological conditions. Furthermore, additional measurements were done at the University of Bordeaux (PIOM laboratory) to determine the dielectric behaviour of a litter layer such as that found at the Bray site. There is some evidence that this layer may have a different influence on the L-band signal than either the soil or the vegetation, however the exact behaviour of the litter layer and the extent of its influence on the L-band signal are as yet unknown. This paper presents 1) results of the Bray experiment describing the behaviour of the above-canopy L-band emissivity for different conditions of ground moisture and 2) the relationship between soil and litter moisture content and results of the laboratory experiments on litter dielectric properties. Together this will give a first insight into the L-band radiometric properties of the different forest layers for varying hydrological conditions.
△ Less
Submitted 12 July, 2007;
originally announced July 2007.
-
Etude D'Un Modele D'Inversion Liant L'Emissivite a L'Humidite Des Sols. Contribution a la Mise Au Point De L'Algorithme De la Mission Smos
Authors:
François Demontoux,
Bénédicte Le Crom,
Gilles Ruffié,
Jean Pierre Wigneron,
Jennifer Grant,
Heather Lawrence
Abstract:
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structure soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still ignored. Its effect can be corre…
▽ More
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structure soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still ignored. Its effect can be corrected via a simple radiative model. It is thus necessary to set up an analytical model which would make it possible to correct the effect of this additional layer. The objective of this article is to present the analytical models which we retained to correct the effect of the vegetation and the litter in order to know the emissivity of the bare soil. We developed a numerical model (with software HFSS) of calculation of the emissivity of multi-layer systems in order to validate the results of the forward inversion models. Ranges of permittivity are introduced with the model in order to take account of the disturbances related to measurement and the variation of the water content of the layers.
△ Less
Submitted 12 July, 2007;
originally announced July 2007.
-
Caracterisation Electromagnetique De Materiaux Geologiques En Vue Du Suivi De L'Humidite Des Sols Par Radiometrie Micro-Ondes
Authors:
Bénédicte Le Crom,
François Demontoux,
Gilles Ruffié,
Jean Pierre Wigneron,
Jennifer Grant
Abstract:
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structures composed of soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still ignored. The object…
▽ More
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structures composed of soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still ignored. The objective of this work was to study the dielectric properties of a type of litter and ground in order to integrate these values into an analytical multi layers model of soil. The objective is to characterize the effects of this layer on the total multi layer system. This will make possible to lead to a simple analytical formulation of a model of litter which can be integrated into the calculation algorithm of SMOS in order to collect information on moisture starting from measurements of emissivity
△ Less
Submitted 12 July, 2007;
originally announced July 2007.
-
Caractérisation électromagnétique de milieux hétérogènes naturels - Application à la mesure de l'humidité du sol par radiométrie micro-onde
Authors:
François Demontoux,
Gilles Ruffié,
Jean Pierre Wigneron,
Maria-José Escorihuela,
Jennifer Grant
Abstract:
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structure soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still are very ignored. So we develope…
▽ More
The work which we present takes place within the framework of mission SMOS of the ESA which will consist to send a radiometer (1.4 GHz) in space. The goal of the research which we propose is the improvement of the comprehension of the effects of structure soil and litter. The effects of the litter and heterogeneities of the ground are probably important but still are very ignored. So we developed an experimental approach in laboratory and in situ. That makes it possible to take measurements for various configurations (frequency, temporal, polarization, incidence, Bi-statics, Brewster effect...) and in term of surface conditions(homogeneous or heterogeneous ground, more or less wet, presence of litter...). Measurements at the laboratory with waveguide enabled us to characterize the various components of the geological structure (ground, rocks) and to check the model of Dobson usually used.
△ Less
Submitted 12 July, 2007;
originally announced July 2007.
-
Inversion model validation of ground emissivity. Contribution to the development of SMOS algorithm
Authors:
François Demontoux,
Bénédicte Le Crom,
Gilles Ruffié,
Jean Pierre Wigneron,
Jennifer Grant,
Daniel Medina Hernandez
Abstract:
SMOS (Soil Moisture and Ocean Salinity), is the second mission of 'Earth Explorer' to be developed within the program 'Living Planet' of the European Space Agency (ESA). This satellite, containing the very first 1.4GHz interferometric radiometer 2D, will carry out the first cartography on a planetary scale of the moisture of the grounds and the salinity of the oceans. The forests are relatively…
▽ More
SMOS (Soil Moisture and Ocean Salinity), is the second mission of 'Earth Explorer' to be developed within the program 'Living Planet' of the European Space Agency (ESA). This satellite, containing the very first 1.4GHz interferometric radiometer 2D, will carry out the first cartography on a planetary scale of the moisture of the grounds and the salinity of the oceans. The forests are relatively opaque, and the knowledge of moisture remains problematic. The effect of the vegetation can be corrected thanks a simple radiative model. Nevertheless simulations show that the effect of the litter on the emissivity of a system litter + ground is not negligible. Our objective is to highlight the effects of this layer on the total multi layer system. This will make it possible to lead to a simple analytical formulation of a model of litter which can be integrated into the calculation algorithm of SMOS. Radiometer measurements, coupled to dielectric characterizations of samples in laboratory can enable us to characterize the geological structure. The goal of this article is to present the step which we chose to validate this analytical model.
△ Less
Submitted 6 July, 2007;
originally announced July 2007.
