PhD title : Ecological monitoring of semi-natural grasslands: statistical analysis of dense satellite image time series with high spatial resolution

Supervisors: Mathieu Fauvel (MCF INP-ENSAT, UMR Dynafor) and Stéphane Girard (DR INRIA Grenoble Rhône-Alpes, Team Mistis)

Funding: CJS INRA-INRIA (Young Scientist Contract between INRA and INRIA) (2014-2017)

Abstract:

Grasslands are a significant source of biodiversity in farmed landscapes that is important to monitor. New generation satellites such as Sentinel-2 offer new opportunities for grassland’s monitoring thanks to their combined high spatial and temporal resolutions. Conversely, the new type of data provided by these sensors involves big data and high dimensional issues because of the increasing number of pixels to process and the large number of spectro-temporal variables. This thesis explores the potential of the new generation satellites to monitor biodiversity and factors that influence biodiversity in semi-natural grasslands. Tools suitable for the statistical analysis of grasslands using dense satellite image time series (SITS) with high spatial resolution are provided. First, we show that the spectro-temporal response of grasslands is characterized by its variability within and among the grasslands. Then, for the statistical analysis, grasslands are modeled at the object level to be consistent with ecological models that represent grasslands at the field scale. We propose to model the distribution of pixels in a grassland by a Gaussian distribution. Following this modeling, similarity measures between two Gaussian distributions robust to the high dimension are developed for the classification of grasslands using dense SITS: the High-Dimensional Kullback-Leibler Divergence and the α-Gaussian Mean Kernel. The latter outperforms conventional methods used with Support Vector Machines for the classification of grasslands according to their management practices and to their age. Finally, indicators of grassland biodiversity issued from dense SITS are proposed through spectro-temporal heterogeneity measures derived from the unsupervised clustering of grasslands. Their correlation with the Shannon index is significant but low. The results suggest that the spectro-temporal variations measured from SITS at a spatial resolution of 10 meters covering the period when the practices occur are more related to the intensity of management practices than to the species diversity. Therefore, although the spatial and spectral properties of Sentinel-2 seem limited to assess the species diversity in grasslands directly, this satellite should make possible the continuous monitoring of factors influencing biodiversity in grasslands. In this thesis, we provided methods that account for the heterogeneity within grasslands and enable the use of all the spectral and temporal information provided by new generation satellites.

Key words: Remote sensing, satellite image time series, high dimension, grassland, landscape ecology, biodiversity.

Past research experience:

2013 – 2014 (12 months): Study engineer at the Center for the Study of the Biosphere from Space (CESBIO).

2013 (6 months): Final engineering studies internship at CESBIO.

2012 (3 months): Research internship at the Center for Geospatial Research, University of Georgia, USA.

Teaching experience

2013 and 2015: Short-time teacher, French National School of Agricultural Science and Engineering (ENSAT). GIS introductory course (2 hours) and practical session (3 hours) to undergraduate students from the University of Georgia in the frame of their Maymester at ENSAT.

Supervising

2016: Donatien Dallery, Master’s degree internship (6 months) co-supervised with Mathieu Fauvel.  “Grasslands monitoring using hyperspectral data and satellite image time series: effect of age and management practice on the spectral profile.”

2015: Marc Lang, engineer internship (6 month) co-supervised with Mathieu Fauvel and David Sheeren. “Classification of grasslands types and estimation of the taxonomic diversity from satellite image time series.”

Education

11/2014 – Present: PhD in Remote Sensing (expected in 11/2017), Dynafor, INRA, Toulouse and Mistis Team, INRIA, Grenoble, France.

2010 to 2013: Diploma in Agronomy Engineering, equivalent to a Master’s degree, National School of Agricultural Science and Engineering (ENSAT), Toulouse, France.

2012: Exchange semester, major courses in Geographic Information Science, Geography Department, University of Georgia, Athens, GA, USA.

Publications

Peer-reviewed publications

M. Lopes, M. Fauvel, A. Ouin, and S. Girard. Spectro-Temporal Heterogeneity Measures from Dense High Spatial Resolution Satellite Image Time Series: Application to Grassland Species Diversity Estimation. Remote Sensing, 9(10):993, 2017. Special Issue "Dense Image Time Series Analysis for Ecosystem Monitoring".

M. Lopes, M. Fauvel, S. Girard, and D. Sheeren. Object-Based Classification of Grasslands from High Resolution Satellite Image Time Series Using Gaussian Mean Map Kernels. Remote Sensing, 9(7):688, 2017.

D. Sheeren, M. Fauvel, V. Josipović, M. Lopes, C. Planque, J. Willm, and J.-F. Dejoux. Tree species

classification in temperate forests using Formosat-2 satellite image time series. Remote Sensing, 8(9):734, 2016.

J.-P. Gastellu-Etchegorry, T. Yin, N. Lauret, T. Cajgfinger, T. Gregoire, E. Grau, J.-B. Feret, M. Lopes,

J. Guilleux, G. Dedieu, Z. Malenovský, B. D. Cook, D. Morton, J. Rubio, S. Durrieu, G. Cazanave, E.

Martin, and T. Ristorcelli. Discrete anisotropic radiative transfer (DART 5) for modeling airborne and

satellite spectroradiometer and Lidar acquisitions of natural and urban landscapes. Remote Sensing,

7(2):1667–1701, 2015.

Proceedings of international conferences

M. Lopes, M. Fauvel, A. Ouin, and S. Girard. Potential of Sentinel-2 and SPOT5 (Take5) time series for the estimation of grasslands biodiversity indices, 2017 9th International Workshop on the Analysis of Multitemporal Remote Sensing Images (MultiTemp), pp. 1-4, Brugge, Belgium, 2017.

M. Lopes, M. Fauvel, S. Girard, D. Sheeren. High dimensional Kullback-Leibler divergence for grassland management practices classification from high resolution satellite image time series, 2016 IEEE International Geoscience And Remote Sensing Symposium (IGARSS), pp. 3342-3345, Beijing, China, 2016.

Communications

Oral communications

M. Lopes, M. Fauvel, A. Ouin, and S. Girard. Estimation de la diversité en espèces des prairies à partir de leur hétérogénéité spectrale en utilisant des séries temporelles d'images satellite à haute résolution spatiale, Rencontres Ecologie des Paysages 2017, Toulouse, France, 2017.

M. Fauvel, M. Lopes, and S. Girard.  Object-based classification of grassland from high resolution satellite image time series with Gaussian mean map kernels, Journée GDR-ISIS / CCT-TSI Séries d’images multi-temporelles à haute revisite, Toulouse, France, 2017.

S. Girard, M. Lopes, M. Fauvel, and D. Sheeren. Object-based classification of grassland from high resolution satellite image time series with Gaussian mean map kernels, 27th Annual Conference of the International Environmetrics Society, Bergamo, Italy, 2017.

M. Fauvel, M. Lopes, A. Ouin, and S. Girard. Evaluation de la biodiversité des prairies semi-naturelles par télédétection hyperspectrale. 5ème Colloque scientifique du groupe thématique hyperspectral de la Société Française de Photogrammétrie et Télédétection, Brest, France, 2017.

M. Lopes, S. Girard, and M. Fauvel. Divergence de Kullback-Leibler en grande dimension pour la classification des prairies à partir de séries temporelles d’images satellite à haute résolution. 48èmes Journées de Statistique de la Société Française de Stastique, Montpellier, France, 2016.

Posters

M. Lopes, M. Fauvel, S. Girard, D. Sheeren, M. Lang. High Dimensional Kullback-Leibler Divergence for grassland object-oriented classification from high resolution satellite image time series. 2016 European Space Agency Living Planet Symposium, Prague, Czech Republic, 2016.

M. Lopes, M. Fauvel, S. Girard, D. Sheeren, and M. Lang. High Dimensional Kullback-Leibler Divergence for grassland object-oriented classification from high resolution satellite image time series. 4ème journée thématique du Programme National de Télédétection Spatiale, CNES, Paris, France, 2016.