PhD Position: Uncertainty Quantification in Climate Precipitation Trends using Generative Models
Develop probabilistic generative models to improve climate precipitation estimates from heterogeneous satellite data.
Best for:Best for candidates with a strong background in atmospheric science, remote sensing, and probabilistic machine learning.
Overview
The Laboratoire Atmosphères et Observations Spatiales (LATMOS) is seeking a PhD candidate to join the ANR-funded AICAtS project. The research focuses on addressing the challenges of quantifying uncertainty and confidence in climate precipitation trends derived from satellite observations. The successful candidate will develop probabilistic generative models to estimate precipitation fields, extend these methods across GPM constellation radiometers, and create robust evaluation frameworks. This position offers the opportunity to work on fundamental AI challenges applied to atmospheric sciences within an international collaborative environment.
What you'll do
- Develop a probabilistic generative model for precipitation field estimation using GMI radiometer data.
- Extend the probabilistic framework to integrate data from the full GPM satellite constellation.
- Address challenges related to heterogeneous observation geometries, spatial resolutions, and instrument sensitivities.
- Design and implement a novel evaluation framework for probabilistic precipitation products.
- Collaborate with international partners at IIT Hyderabad and the University of Oklahoma.
- Contribute to the scientific objectives of the ANR AICAtS project regarding stochasticity and uncertainty quantification.
Requirements
Degree required: Master's degree in Atmospheric Science, Physics, Computer Science, or a related field.
Degree offered: PhD
Language: English proficiency required for research and international collaboration.
Instruction: English/French
- Strong foundation in atmospheric physics or remote sensing.
- Proficiency in machine learning, specifically generative models.
- Experience with probabilistic modeling and uncertainty quantification.
- Programming skills relevant to data analysis and modeling (e.g., Python, MATLAB).
- Must hold a Master's degree or equivalent by the start date.
Funding & terms
How to apply
Deadline: 2026-10-28
Applications must be submitted through the official CNRS portal. Candidates should follow the instructions provided on the job posting page to upload their application materials.
- CV
- Cover letter
- Academic transcripts
Create a free profile to unlock your personal match score, the full requirements, funding details and how to apply — in English or Persian.