The supply partner is not in a position to guarantee the accuracy, up-to-dateness, integrity, completeness of the data and in particular that it is free from errors or inaccuracies, in particular with regard to location, identification or qualification. No guarantee is given by the supplying partner as to the suitability of the data for a particular purpose. Users use the data under their own responsibility, without any recourse against the partner supplier, who cannot be held liable for any damage resulting directly or indirectly from the use of these data. In particular, it is up to the users to assess, under their sole responsibility: the appropriateness of using the data; the compatibility of the files with their computer systems; the adequacy of the data to their needs; that they have sufficient competence to use the data. The partner supplier is not responsible for any elements outside the data, in particular the analysis tools, hardware, software, networks, etc., used to consult and/or process the data. The user shall ensure that the information made available is up to date and compatible with the use he/she makes of it.

Free use subject to the obligatory mentions on all distribution documents: "Source: under the direction of the OSU Nantes Atlantique".

The data is provided under the Open License 2.0 ( https://www.etalab.gouv.fr/wp-content/uploads/2018/11/open-licence.pdf).

fre

Région Pays-de-la-Loire

The microphytobenthos is largely dominated by diatoms , which form a biofilm on the surface of muddy sediments during the diurnal period of low water. The microphytobenthos consists of important primary producers and is the main food source for benthic invertebrates such as oysters cultivated . The estimate of biomass microphytobenthos therefore a key parameter for studying the functioning of mudflat ecosystems. Here has been chosen as the chlorophyll a biomass estimate because microphytobenthos productivity is essentially based on the photosynthetic pigment.

The work was made ​​from hyperspectral images taken using a VNIR camera whose characteristics depend on the sensor used. The images were taken between May and September , at midday and at low tide, that is to say when the microphytobenthos biofilm develops. The flight plan was scheduled to have enough time to cover all of the intertidal zone at low tide.

Before analyzing the images underwent atmospheric correction with flaash algorithm ( Fast Line -of-sight Atmospheric Analysis of Spectral Hypercubes ) that includes the codes of radiative transfer MODTRAN4 (see Matthew et al , 2000. Status of atmospheric corrected using a MODTRAN4 based algorithm . SPIE Proceedings , Algorithms for Multispectral , Hyperspectral , and ultraspectral imagery VI . Vol. 4049 , pp . 199-207 ) to convert images of radiance images free of airborne reflectance characteristics . Measurements were carried out previously in the laboratory , these representatives a small fraction of the intertidal zone ( as it is difficult to access the mudflat ) and were extrapolated to a larger scale, with care , with the use hyperspectral images.

The setting point has coordinates used : -2.171169543 ; 47.29112657 , located in the municipality of MONTOIR-DE-BRETAGNE.

The first step of analyzing hyperspectral images has been to differentiate the biofilm microphytobenthos diatoms, and thus mask pixels containing other photosynthetic organisms. Only diatoms are taken into account here because the physical model developed to quantify biomass takes into account only the characteristics of this group of microalgae. The euglena, which are another important group microphytobenthos, are therefore also excluded from the mask. It has been shown in the laboratory that the biofilm microphytobenthos not absorb or reflect light in wavelength ranges of near infrared, in contrast to macroalgae. Microalgae contain less pigment than macroalgae, and have low and narrow absorption characteristics that give a characteristic peak at about 580 nm reflectance due to absorption bands of chlorophyll and carotenoid. This helped to define an index microphytobenthos: MPBI to restrict the area of ​​interest. In the laboratory, the euglena and diatoms, which are the two most represented groups in the microphytobenthos the Region Pays de la Loire, exhibit every 2 a high concentration of chlorophyll A, and have for this pigment the same reflectance spectrum the absorption band of 673 nm. By cons, there are differences in reflectance by their own assembly pigments:

-the euglena contain chlorophyll B, which broadens the absorption function to 673 nm,

-the diatoms contain chlorophyll c and fucoxanthin especially, the latter having a strong absorption peak between 540 and 550 nm. These properties were used to select the most representative pixels of diatoms on images, working on the presence / absence of fucoxanthin. Thus the peak of the absorption spectrum was set to 553 nm for euglena and 600 nm for diatoms, to define an index of diatoms IDiatomée and an index of euglena: IEuglène.

Here is the process used to determine the diatomaceous pixels:

1) A first mask was used from NDVI (Vegetation Index Normalized Difference) to calculate a binary mask non vegetated surface using a minimum NDVI value of 0.1 (here calculated from the sensor resolution HySpeX)

NDVI = (R800-R673) / (R800 + R673) (R = reflectance)

2) Then a second mask has been applied, calculated from the index microphytobenthos

IMPB = (2R586 / (R495 + R673)) - 1 to differentiate microphytobenthos the remainder of the growing

If IMPB exceeds NDVI, the pixel is kept as it corresponds more to microphytobenthos, otherwise hide it. It was also defined that if IMPB is less than 0.25 it corresponds to the naked sediment and is also masked.

3) A final mask was applied to differentiate diatoms euglena, calculated from the index and the index euglena diatoms:

IEuglène = (2R553 / (R600 + R495)) - 1

IDiatomée = (2R600 / (R459 + R673)) - 1

Only then are kept the pixels having a higher IEuglène than IDiatomée.

The second step was to determine the biomass microphytobenthos from images obtained following the selection of the pixels, expressed in mg of chlorophyll A per m². A radiative transfer model microphytobenthic biofilm (MPBOM) developed previously (see thesis Kazemipour Farzaneh, 2011) was used to determine the amount of diatom biomass. The MPBOM was to calculate the spectrum of the absorption coefficient of the measured reflectance spectrum. In the laboratory it was possible to determine the linear regression between the absorption coefficient of chlorophyll a to 673 nm and the biomass values ​​then determine in situ from images, the amount of biomass from the obtained absorption coefficient with MPBOM.

Microphytobenthos biomass was determined by the formula:

biomass = 96.9 x coefficient of chlorophyll A absorption at 673 nm (R2 = 96%).

To calculate the MPBOM, were taken into account here the reflectance spectrum of the substrate and environmental parameters. This helped to achieve biomass mapping microphytobenthos.the image acquisition work was funded by GEOPAL programs ( State - Region of Pays de la Loire - ERDF) and OSUNA , and directed by the ITF CONSULTING society. The sensor was funded by a State-Region Contract Plan, the Loire Atlantic, FNADT and ERDF . The work was made ​​from hyperspectral images taken using a VNIR HySpex camera with a spectral resolution of 4.5 nm to 160 channels of 400 to 1000 nm spectral range . The camera was used from a plane at 2,500 meters, to an opening angle of 17 °. The images were taken between May and September , at midday and at low tide, that is to say when the microphytobenthos biofilm develops.

The work analyzes were performed by the Laboratory of Planetology and Geodynamics of Nantes, CNRS UMR 6112 , Laboratoire Sea Molecules and Health , EA 2160 , Université de Nantes .

fre

fbfede52-b320-4986-8afc-deef4dc722aa

ISO 19115

1.0