slideshow 1 slideshow 2 slideshow 3 slideshow 3 slideshow 3 slideshow 6 slideshow 7 slideshow 8 slideshow 9 slideshow 10 slideshow 11 slideshow 12 slideshow 13 slideshow 14 slideshow 15 slideshow 16

Vous êtes ici

√ Niveau requis: 
√ Début du stage: 
02-01-2018
√ Durée: 
6 mois
√ Lieu: 
2 place Viala 34060 Montpellier cedex 2 FRANCE
√ Indemnité: 
€577
√ Contact: 
Jean-Stéphane Bailly > téléphone : +33 (0)4 99 61 28 09 > email : bailly@agroparistech.fr
√ Description: 

Spatial estimation from mixed spatial data (continuous and ranked data) : application to temporary flooded areas elevation fields

6 months Master Internship 2018 - Montpellier - South of France

Supervision :
BAILLY Jean-Stéphane, bailly@agroparistech.fr, AgroParisTech, LISAH
BELAUD Gilles, gilles.belaud@supagro.fr, Supagro, G-EAU
DELENNE Carole, carole.delenne@umontpellier.fr, Univ. Montpellier, HSM/INRIA Lemon
ROUSSEAU Antoine, antoine.rousseau@inria.fr, INRIA Lemon

Grant : ~ 577 EUR /month

Skills : Applied mathematics, statistics, computational sciences or hydrologist with mathematical/computational background

Motivation
The understanding and modeling of natural or anthropogenic temporary flooded areas along rivers networks (riverbeds, wetland, deltas, small dams, etc) are of high importance for both ecosystems and human population (e.g. water resource assessment, flood risk evaluation). The modeling of these area needs a proper and precise
description of the geometry, i.e. the topography or elevation field, which may be highly dynamic and need to be monitored ([Albergel et al., 2005], [Bouteffeha et al. 2015]). Moreover, these areas are less prone to conventional topographic surveys since they are flat and need high precision topographic instruments and/or the temporary
immersed character of these areas make them unsuitable for infrared LiDAR survey.
    
Optical and radar remote sensing images available now at high spatial and temporal resolutions allows the dynamic delineation of flooded limits (see Figure), even for small waterbodies ([Ogilvie et al., 2015], [Soti et al., 2005]) and their use for hydrodynamic model calibration ([Ogilvie et al., 2015], [Hostache et al., 2010], [Schumann et al., 2007]).
In this context, the topographic information resulting from multi-temporal waterbodies delineation from remote sensing images consists in a series of ranked contourlines with unknown elevation (not labeled). This information may also be ranked with some known topographic points within the flooded area.
Figure - A flooded area in The Inner Niger Delta - West Africa (from [Wymenga et al., 2002]) at two different dates.

The internship aims to develop a generic spatial estimation method to reconstruct a 2D field (topographic field) from a set of both punctual data and ranked but not labeled contourlines.

Delenne et al. (2016) and Dartevelle (2016) proposed a geostatistical method to estimate such a 2D field (topographic field) from a set of both ponctual data and ranked unlabeled contourlines. This method relies on spatial model inference from a numerous set of data points and assume the spatial ergodicity/stationarity of the studied
spatial field (elevation field), which make it poorly generic. We aim here to find an alternative deterministic methodpassing over inference and stationarity problems. This generic and deterministic method will rely on mixed objective optimization criteria linked at least to i) usual interpolation distance to data points and ii) rank of contourlines.
Different interpolation methods and optimization criteria taking account of data uncertainty will be under study.

Methods will be tested on synoptic datasets as well as some actual topographical fields on study sites of the supervisor teams: Small dams in France and Maghreb, Mékong floodplain, Camargue delta, Niger inner delta, etc.

References :
[Albergel et al., 2005] Albergel, J., Collinet, J., Pepin, Y., Nasri, S., Boufaroua, M., Droubi, A., & Merzouk, A. (2005). Sediment budgets on hill reservoirs of small catchments in North Africa and the Middle East. Iash, 291(1), 323-331.
[Bouteffeha et al., 2015] Bouteffeha, M., Dages, C., Bouhlila, R., & Molenat, J. (2015). A water balance approach for quantifying subsurface exchange fluxes and associated errors in hill reservoirs in semiarid regions. Hydrological processes, 29(7), 1861-1872.
[Dartevelle, 2015] Dartevelle, M. (2016), 'Méthodes d'estimation de topographies à partir de données mixtes (continues et ordinales)', Master's thesis, INSA-Toulouse.
[Delenne et al., 2016] C. Delenne, J.-S. Bailly, M. Dartevelle, N. Marcy, and A. Rousseau. Combining punctual and ordinal contour data for accurate floodplain topography mapping (poster and 8p. paper). In J. S. Bailly, D. Griffith, and D. Josselin, editors, Spatial accuracy : International symposium on "Spatial Accuracy Assessment in Natural Resources and Environmental Sciences", Montpellier (France), 5-8 July 2016.
[Hostache et al., 2010] Hostache, R., Lai, X., Monnier, J. et Puech, C. (2010). Assimilation of spatially distributed water levels into a shallow-water flood model. part ii : Use of a remote sensing image of mosel river. Journal of Hydrology, 390(3-4):257 - 268.
[Ogilvie et al., 2015] Ogilvie, A., Belaud, G., Delenne, C., Bader, J.-C., Oleksiak, A., Bailly, J. S., Ferry, L. et Martin, D. (2015). Decadal monitoring of the Niger Inner Delta flood dynamics using MODIS optical data. Journal of Hydrology, 523:358-383.
[Schumann et al., 2007] Schumann, G., Matgen, P., Hoffmann, L., Hostache, R., Pappen- berger, F. et Pfister, L. (2007). Deriving distributed roughness values from satellite radar data for flood inundation modelling. Journal of Hydrology, 344(1-2):96 - 111.
[Soti et al., 2009] Soti, V., Tran, A., Bailly, J. S., Puech, C., Seen, D. L., & Bégué, A. (2009). Assessing optical earth observation systems for mapping and monitoring temporary ponds in arid areas. International Journal of Applied Earth Observation and Geoinformation, 11(5), 344-351.
[Wymenga et al., 2002] Wymenga, E., Kone, B., Kamp, J. V. D. et Zwarts, L. (2002). Delta intérieur du fleuve Niger. Mali-PIN publication.

Application :
Send your application (CV and letter) by email to : bailly@agroparistech.fr, carole.delenne@umontpellier.fr, antoine.rousseau@inria.fr or gilles.belaud@supagro.fr

Other considerations :
Internship length : 5 to 6 months. Can start in January 2018
Grant :~ 577 EUR /month + daily lunch at the INRA-Supagro restaurant at preferential rate
Internship location : LISAH laboratory, Campus INRA-Supagro, Montpellier City center, France, (see location on googlemap)

Statut stage: