Design a parametric crop insurance index tailored for smallholder farmers using satellite-derived indicators and rainfall data.
Use remote sensing time series to build robust, scalable indices that can be automated for large areas.
Objectives & main tasks
Extract and preprocess satellite and rainfall datasets (NDVI from MODIS/Sentinel-2 via Google Earth Engine, CHIRPS rainfall).
Define candidate parametric index formulations, calibrate thresholds and loss triggers suitable for smallholder contexts and validate against ground truth or proxy yield data.
Technical environment & tools
Languages & libraries: Python (pandas, numpy), R (optional), Scikit-learn for basic modeling.
Platforms & tools: Google Earth Engine for satellite data extraction and processing, Jupyter Notebook for analysis, QGIS for spatial mapping.
Storage & visualization: PostgreSQL/PostGIS for spatial data storage and Metabase or Dash/Plotly for interactive visualization.
Expected deliverables
Cleaned datasets and reproducible extraction scripts (Google Earth Engine + ingestion to PostgreSQL/PostGIS).
Analysis notebooks showing index formulation, calibration, validation results and simple models (scikit-learn).
Spatial maps and interactive visualizations (QGIS exports and Dash/Plotly or Metabase dashboards) plus a final written report describing methodology and recommendations.
Skills & competencies sought
Strong data processing and Python experience (pandas, numpy); familiarity with remote sensing concepts (NDVI, time-series) and Google Earth Engine.
Experience with spatial databases (PostGIS), basic machine learning for index calibration, and dashboarding/visualization tools.
How to apply
Apply via the company site: https://www.hydatis.com
Or send your application email to stages@hydatis.fr including CV, cover letter, and a short description of relevant projects (use the subject line specified below).