Surface marine wind data base front ERS-1 scatterometer has been processed to define the characteristic seasonal distribution of wind-stress curl throughout the southwestern Indian Ocean between January and December 1994. A compact model to obtain satellite-derived wind-stress curl fields on smaller scales than previously available is proposed and evaluated. The results indicate the strong capability of the ERS-1 scatterometer to monitor wind-driven variations in the mesoscale ocean patterns.

Satellite remote sensing data and purse seining catch data are used to analyse the ecological behaviour of tuna in the Indian Ocean. A set of environmental parameters deduced from satellite data is extracted at the fishing points by using an interface developed at the “Institut de Recherche pour le Développement” (IRD). Statistical analysis were then carried out and some linear model and nonlinear models (neural networks) have been set and evaluated. A methodology based on genetic algorithms is used in order to avoid overfitting and to obtain suitable models.

The littoral of French Guiana is characterized by the north-westward migration of large mud banks along shore and by high suspended particulate matter concentrations (SPMC) coming from the Amazon River. A correspondence function is established between in situ optical data and SPMC measurements obtained from three coastal surveys on the Kaw River, the Mahury estuary, and the adjacent continental shelf, respectively. This function is applied to SPOT satellite images to estimate SPMC distributions near the sea surface on five dates between 1998 and 2001. Two typical situations have been observed: the mud suspensions are generally confined near the coast by a strong north-westward coastal current, while in one case (SPOT scene of 2 July 2001) the turbid plumes, probably resulted from the wave action on the mud banks, are directed seaward. SPMC varies from more than 300 mg l−1 adjacent to the intertidal zone to a few mg l−1 offshore. Turbid plumes are observed downstream, with regard to the Guyana Current, of small islands along the coast. Mud undulations are observed on the emergent surface of the mudflats, parallel to the crest waves and characterized by a wavelength of 200–1000 m. These structures are interpreted to form by wave action on the soft mudflat surface. Mangrove colonization of the mudflat follows these primary topographic structures.

Satellites products, as sea surface temperature maps, were and are still used to locate zones of high probability of concentration of pelagic fish, thus reducing search time. The fishing industry request for satellite information really became significant on the late 90s, when the offer became more complete with mainly the provision of maps of water color and maps of altimetry (anomalies of the ocean level and related currents). Therefore, the knowledge of oceanic landscape is more and more increasing and should profit to all the fishing field actors. In this context, the goals of the SeAGeRH project are (1) to combine competences and acquired experience from IRD and CLS to rapidly develop a comprehensive and competitive offer in the field of spatial oceanography products dedicated to the-fisheries market and (2) to propose an adapted version of the range of products for the specific requirements of the institutions responsible for the management of fisheries resources, a new market.

Satellites products, as sea surface temperature maps, were and are still used to locate zones of high probability of concentration of pelagic fish, thus reducing search time. The fishing industry request for satellite information really became significant on the late 1990s, when the offer became more complete with mainly the provision of maps of water color and maps of altimetry (anomalies of the ocean level and related currents). Therefore, the knowledge of oceanic landscape is more and more increasing and should profit to all the fishing field actors. In this context, the goals of the SeAGeRH project are (1) to combine competences and acquired experience from IRD and CLS to rapidly develop a comprehensive and competitive offer in the field of spatial oceanography products dedicated to the fisheries market and (2) to propose an adapted version of the range of products for the specific requirements of the institutions responsible for the management of fisheries resources, a new market.

The determination of shallow water depth in high-energy tidal inlets is essential to model and to forecast the navigation channel position or the short-term topographic evolution of beaches. Colour satellite imagery provides, at low cost, complete maps of areas that are difficult to map by traditional hydrographic means due to their size and their rough underwater morphology. In order to calibrate SPOT images to derive bathymetric maps, a simple method applied to shallow waters of a moderately turbid tidal inlet has been carried out. Its development is based on a set of field measurements, including the reflectance of the water and of the bottom sediment, the vertically averaged diffuse attenuation coefficient, and the concentrations of inorganic particles in suspension, of chlorophyll a and pheopigments and of dissolved organic carbon (DOC). From these data, it appears, first, that the water reflectance is directly linked to the depth, and second, that the water reflectance varies slowly with the turbidity in this area for total suspended matter (SM) concentrations lower than 9 mg l−1. The extinction of light with depth has been characterized in the inlet. The relationship obtained is slightly different from that observed in clear waters. The bathymetric models established for the clearest waters do not calculate the depth accurately. In the contrary, reflectance code adapted to Case II water type, calibrated with in situ measurements, allows us to retrieve depth down to 6 m and for total SM concentrations ranging from 0.2 to 9 mg l−1. This relationship has been applied to five SPOT images of the mouth of Arcachon lagoon. The accuracy of the derived map has been assessed by in situ depth measurements. The mean difference between measured depths and computed depths is about 20%. This accuracy is adequate to assess the inlet morphodynamics quantitatively, which is necessary for middle-term to short-term mathematical modeling of the 3D inlet evolution.

Remote sensing from satellite sensors is a suitable technique for synoptic monitoring of coastal waters. However quantitative use of remotely sensed data is difficult due to the complex behaviour of these waters. Lahet et al. (2001) demonstrated the suitability of associating spectral optical measurements and a water colour classification to provide reliable estimates of water quality variables in coastal waters by low turbidity. In this Letter, we investigate the robustness of the colour classification method by considering experimental reflectances sampled at the central wavelength of various satellite sensor wavebands.

This paper presents a colour classification of coastal waters from features of spectral reflectance. Experimental data were collected in coastal waters of the Ebro river mouth area during two periods of non-bloom conditions and low suspended load. Examination of the reflectance spectra reveals a few significant shapes corresponding to different apparent colours and specific values of hydrological parameters. A colour classification is proposed. An empirical method and a semi-analytical method are presented in order to retrieve suspended particulate matter concentration, namely SPM, chlorophyll a concentration, chl, sediment refractive index, mr , and absorption coefficient of yellow substance at 440 nm, a y (440). The first method is a direct application of the colour classification. It is based on correlations found for each water type between spectral features of the first derivative of reflectance and SPM, chl, mr , ay (440) values. A second method uses the inversion of a three component reflectance model. Measured reflectance spectra and typical values of mr , deduced from the colour classification, are assimilated in the model. As a result, it appears that estimates of SPM and chl are more accurate by calculating chl from the correlation technique and SPM along with ay (440) from inversion of the reflectance model.

A three-component reflectance model that takes into account the contributions of dissolved organic matter, phytoplankton, and mineral particles is described. The principle of inversion of the model, in terms of chlorophyll a plus phaeopigments concentration, chl, sediment refractive index, mr, and absorption coefficient of yellow substance at 440 nm, ay(440), is exposed. The inversion procedure is applied to experimental data collected in the Ebro river plume area. The obtained estimates of chl are accurate only in order of magnitude. However, we demonstrate that mr and ay(440) determinations are robust against chl variations within −50%, +100% of experimental values. Investigation of possible variations with wavelength of mr, using measured concentrations and ay(440) estimates, evidences a slight dependency as was previously observed. We support in conclusion that the inversion procedure could be exploited in a two-step process. First ay(440) and the mean value of mr are estimated from measured water quality parameters. Then mr is calculated at each wavelength, considering ay(440) value determined in the first step.

This article deals with the inversion of reflectance spectra measured in the Rhône River mouth area during a flood event in terms of parameters characterizing the turbid water components. The optically active constituents considered are the total suspended matter, mainly composed of sediment, of known concentration ranging from 3.7 mg L−1 to 234 mg L−1, and yellow substance. A simple reflectance model is presented. Scattering is treated by Mie theory applied to a collection of particles of real refractive index mp and obeying a Junge size distribution law. A sensitivity study performed on the model shows that both concentration and refractive index and both cutoff particle diameters of the Junge size distribution law cannot be accurately determined from a given reflectance spectrum. Two model parameters, among a total of six, are estimated from model inversion: mp and a parameter characterizing the yellow substance absorption spectrum, ay0=ay (440 nm). The derived values are realistic as compared to the values that can be found in literature. Other combinations of parameters could be determined by inversion but measurement of mp and ay0 from reflectance spectra already has interesting applications in a satellite remote sensing perspective. Increasing the number of parameters of the model is feasable, but much work is needed to envisage the inverse problem.

Spectroradiometers are important tools used in the study of the optical properties of the ocean and in the validation of satellite ocean color sensors. The authors present the technical characteristics of a field instrument and the methods for measuring spectral reflectance and beam attenuation coefficient. Then the inversion of a reflectance model, in terms of various relevant parameters, is exposed

This paper concerns the modelling of the reflectance spectra of turbid coastal waters and the inversion of two parameters: the sediment refractive index and a parameter characterizing the yellow substance absorption spectrum. The study uses experimental data of reflectance spectra and of suspended matter concentrations collected in the Rhone river mouth. The results of inversion are quite realistic and the restitution of measured reflectance spectra is satisfactory. We envisage in the future to increase the number of parameters to invert and to apply the inversion procedure to satellite data.

This paper presents a colour classié cation of coastal waters from features of spectral reè ectance.Experimental data were collected in coastal waters of the Ebro river mouth area during two periods of non-bloom conditions and low suspended load. Examination of the reè ectance spectra reveals a few signié c- ant shapes corresponding to di Verent apparent colours and specié c values of hydrological parameters.Acolour classié cation is proposed.An empirical method and a semi-analytical method are presented in order to retrieve suspended partic- ulate matter concentration, namely SPM, chlorophyll a concentration, chl, sedi- ment refractiveindex, m r ,and absorption coe Ycient ofyellowsubstance at 440nm, a y (440). The é rst method is a direct application of the colour classié cation. It is based on correlations found for each water type between spectral features of the é rst derivative of reè ectance and SPM, chl, m r , a y (440) values. A second method uses the inversion of a three component reè ectance model. Measured reè ectance spectra and typical values of m r , deduced from the colour classié cation, are assimilated in the model. As a result, it appears that estimates of SPM and chl are more accurate by calculating chl from the correlation technique and SPM along with a y (440) from inversion of the reè ectance model.