Search

International audience

International audience; An improved version of the MARMIT model is proposed. Wet soil is represented as a dry soil layer surmounted by a layer of a mixture of water and soil particles.

International audience; Surface soil moisture content (SMC) is known to impact soil reflectance at all wavelengths of the solar spectrum. As a consequence, many semi-empirical methods aim at inferring SMC from soil reflectance, but very few rely on physically-based models. This article presents a multilayer radiative transfer model of soil reflectance called MARMIT (multilayer radiative transfer model of soil reflectance) as a function of SMC given on a mass basis and a method called MARMITforSMC to estimate it from soil reflectance spectra. This model depicts a wet soil as a dry soil covered with a thin film of water. It is used to assess SMC over seven independent laboratory datasets gathered from the literature. A learning phase is required to link the thickness of the water film with the SMC. For that purpose, a sigmoid function, the parameters of which are related to soil physical and chemical properties such as porosity, grain size and mineralogy composition, is fitted. SMC can be inferred with good accuracy (RMSE ≈ 3%) if the learning step is applied soil by soil. The link between SMC and water thickness actually depends on soil texture and chemical composition. If the soils are divided into classes and if the learning phase is applied to a class, the RMSE slightly increases up to 5%. Finally, MARMITforSMC provides lower RMSE than any other existing semi-empirical or physically-based method.

Going on toward the objectives of NAOMI (New Advanced Observation Method Integration) research project, Total and ONERA are working on hyperspectral imagery to detect, characterize and quantify spills at sea. An important part of this work consists in building a database of oil and water-in-oil emulsion reflectance. This database of spectral signatures will be used to analyze the properties of a slick thanks to hyperspectral imagery in the VNIR+SWIR domain and spectral matching techniques. The characterization of the hydrocarbons performed first in laboratory has been completed with a pool experiment. The aim of such an experiment is to measure more realistic spectral signatures in term of background and thickness than in laboratory. Starting from the sample of the emulsion released at sea during NOFO 2015 experiment, emulsion has been remixed once for laboratory measurements and second for the pool experiment. Indeed, its reflectance was measured in the laboratory but for a quite large thickness and it was difficult to predict how the thickness would be once the emulsion freely spread at the water surface. Moreover, depending on the thickness, a signature mixing emulsion and water background could be obtained. In such case, the signatures measured in the lab and in the pool may differ significantly. As a consequence, the use of spectral signature measured in laboratory may give poor spectral matching results. In order to get the answers, a pool experiment, piloted by ONERA in the frame of the NAOMI collaboration with TOTAL, was organized at CEDRE in Brest (France). CEDRE’s pool is usually used for oil recovery equipment testing or people operating such equipment training. Thus, the pool is large (1900 m²), fairly deep (3 m) and filled with ocean water. Known volumes of several products, including the NOFO 2015 emulsion, were successively poured into a delimited area within the pool. Two hyperspectral cameras put on board a cherry picker located on the pool quay were lifted at about 15 m above the delimited area in order to take images of each small spill. The obtained spectral signatures have been compared with the laboratory ones. Detection algorithms have been applied to the pool hypercubes in order to identify the pixels covered by the NOFO 2015 emulsion and a thickness assessment has been performed. A wide characterization of the NOFO2015 emulsion has been done thanks to laboratory measurements, in pool experiment and airborne images from the NOFO campaign in 2015. The paper will present the pool experiment and the corresponding image processing. A quick recall of the laboratory measurements will be done before presenting the pool and laboratory spectral signatures comparison. Then the spectral signatures will be compared with data from an airborne image of the 2015 NOFO campaign. Finally, a conclusion will be drawn concerning the information that can be extracted from hyperspectral airborne imagery for such kind of emulsion and more generally.

One of the objectives of the NAOMI (New Advanced Observation Method Integration) research project, fruit of a partnership between Total and ONERA, is to work on the detection, the quantification and the characterization of offshore hydrocarbon at the sea surface using airborne remote sensing. In this framework, work has been done to characterize the spectral signature of hydrocarbons in lab in order to build a database of oil spectral signatures. The main objective of this database is to provide spectral libraries for data processing algorithms to be applied to airborne VNIRSWIR hyperspectral images. A campaign run by the NOFO institute (Norwegian Clean Seas Association for Operating Companies) took place in 2015 to test anti-pollution equipment. During this campaign, several hydrocarbon products, including an oil emulsion, were released into the sea, off the Norwegian coast. The NOFO team allowed the NAOMI project to acquire data over the resulting oil slicks using the SETHI system, which is an airborne remote sensing imaging system developed by ONERA. SETHI integrates a new generation of optoelectronic and radar payloads and can operate over a wide range of frequency bands. SETHI is a pod-based system operating onboard a Falcon 20 Dassault aircraft, which is owned by AvDEF. For these experiments, imaging sensors were constituted by 2 synthetic aperture radar (SAR), working at X and L bands in a full polarimetric mode (HH, HV, VH, VV) and 2 HySpex hyperspectral cameras working in the VNIR (0,4 to 1 μm) and SWIR (1 to 2,5 μm) spectral ranges. A sample of the oil emulsion that was used during the campaign was sent to our laboratory for analysis. Measurements of its transmission and of its reflectance in the VNIR and SWIR spectral domains have been performed at ONERA with a Perkin Elmer spectroradiometer and a spectrogoniometer. Several samples of the oil emulsion were prepared in order to measure spectral variations according to oil thickness, illumination angle and aging. These measurements have been used to build spectral libraries. Spectral matching techniques, relying on these libraries have been applied to the airborne hyperspectral acquisitions. These data processing approaches enable to characterize the oil emulsion by estimating the properties taken into account to build the spectral library, thus going further than unsupervised spectral indices that are able to detect the presence of oil. The paper will describe the airborne hyperspectral data, the measurements performed in the laboratory, and the processing of the optical images with spectral indices for oil detection and with spectral matching techniques for oil characterization. Furthermore, the issue of mixed oil-water pixels in the hyperspectral images due to limited spatial resolution will be addressed by estimating the areal fraction of each.

International audience; A laboratory experiment is set up to study both surface and in-depth soil moisture content (SMC). For that purpose, an aquarium is filled successively with two soils, a clay loam and a sand. Reflectance spectra are acquired in the solar domain (400–2400 nm) on the soil surface using an ASD FieldSpec 3 HR spectroradiometer and in-depth through the aquarium glass wall using two hyperspectral cameras. Successive amounts of water ranging from low to heavy rainfall in a temperate region are uniformly poured into the aquarium. The MARMITforSMC method based on the MARMIT (MultilAyer Radiative Transfer Model for soIl reflecTance) model is applied to each reflectance spectrum to determine gravimetric SMC. In particular, vertical profiles of SMC are provided with unprecedented spatial accuracy (~0.287 mm). The results are compared with volumetric SMC measured by two time-domain reflectometry (TDR) sensors. The in-depth SMC image produced on the sand shows lower values within the first 2 cm (5%) than below (17%). In contrast, the SMC image produced on the clay loam shows evenly distributed values whatever the position in the aquarium, even 1 h after moistening. The difference in grain size between the soils explains this result.; Une expérience est montée en laboratoire pour étudier l’humidité (SMC) du sol à la fois en surface et en profondeur. Pour cela, un aquarium est successivement rempli avec deux sols, un limon argileux et un sable. Les spectres en réflectance sont acquis dans le domaine solaire (400-2400 nm) en utilisant un spectroradiomètre ASD FieldSpec 3 HR pour la surface et deux caméras hyperspectrales pour le profil derrière la paroi de l’aquarium. Des quantités successives d’eau, allant de la pluie faible à forte pour une région tempérée, sont versées uniformément dans l’aquarium. La méthode MARMITforSMC s’appuyant sur le modèle MARMIT (Modèle de Transfert Radiatif Multicouches pour la réflectance du sol) est appliquée à chaque spectre en réflectance pour déterminer l’humidité gravimétrique correspondante. Ceci permet en particulier d’obtenir des profils verticaux d’humidité avec une résolution spatiale sans précédent (0 ,287 mm). Les résultats sont comparés aux mesures d’humidité volumétrique obtenues à l’aide de sonde de réflectométrie dans le domaine temporel (TDR). Les cartes de profil d’humidité générées pour le sable montrent des valeurs plus faibles pour les 2 premiers centimètres (5 %) qu’au-delà (17 %). Par contre, les cartes de profil d’humidité générées pour le limon argileux montrent des valeurs homogènes quelle que soit la profondeur, même 1 heure après l’humidification. La différence de granulométrie des sols explique ces résultats.

This paper deals with the in-the-field absolute calibration of airborne or spaceborne passive optical sensors in the reflective domain. One way to proceed is to image an area whose radiance is known and to have information about atmosphere. Usually, in order to limit the sensitivity to unwell-known parameters, areas are chosen uniform, large, flat, with stability of their reflectance and their atmospheric context. This leads for spaceborne sensors to desert areas, African deserts or Antarctica for instance. For high resolution sensors such as airborne sensors, it is necessary to find closer areas. This paper presents a calibration method relying on a dark and bright target. The use of two reflectances simplifies the assessment of the radiance entering in the camera and probably reduces the sensitivity to the environment of the target. This method is tested for the ONERA airborne camera named PELICAN.