Parcourir les annonces Parcourir les étudiants Publier une annonce Ajouter un CV

Mapping crop phenology from high-resolution sentinel 1 satellite data

Date de début
Date de fin

Mapping crop phenology from high-resolution sentinel 1 satellite data
PhD proposal
University of Copenhagen, Kayrros (a Paris based EO startup), the Laboratoire des Sciences du Climat et
de l’Environnement (LSCE) and INRAE Bordeaux are looking for a PhD candidate on a joint project.
Accurately estimating large-scale crop productivity and yield is critical in many fields: food
security monitoring and agriculturally-related humanitarian crises, commodity markets and
international trades, design and assessment of crop insurance, etc. New optical remote sensing
satellites, such as Sentinel 2 (S1) and Planet constellation, are key tools to monitor crop
phenology, which is critical for crop management and yield estimation. However, optical systems
are limited by cloudy conditions, failing to be operational in many regions/periods (for instance,
during winter in northern Europe). The advent of high-resolution radar data from Sentinel 1 (S1)
at daily time scale opens the opportunity to monitor crops at field scale for all weather
conditions. The challenge is to develop simple statistical models linking the backscatter signal
from S1 to key crop phenological stages. The proposed PhD subject will focus on this topic in
both southern (where both S1 and S2 observations can be used) and northern (where only S1
observations can be used generally due to cloudy conditions) regions of Europe. Focus can be
made initially on crops in France. Model parameters will be optimized through machine learning
algorithms. Direct training and validation of the machine learning tools will be done based on
ancillary data from in situ inventory or other remote sensing observations (S2 and Planet when
they are available).
Overall aim
Apply high spatial and temporal resolution Sentinel 1 satellite data to monitor key crop
phenological stages for a selected region in Europe (proposed: France; if successful, other
regions will be covered)
Specific aims and working steps
• Select a region in Europe with available plot data.
• Develop a semi empirical model to relate S1 backscatter observations and other auxiliary
data, to crop phenology
• Evaluate results against very high resolution datasets
• Programming skills, preferably in Python
• Basic understanding of satellite images and spatial analyses
• Good knowledge in statistics and machine learning
Martin Brandt,
Philippe Ciais,


3 mars 2020