Mercor is hiring experienced meteorologists to help train a large-scale physics AI model by interpreting High-Resolution Rapid Refresh (HRRR) outputs. You will analyze single-frame HRRR model visualizations (4 images per timestamp) and produce high-quality natural-language labels that clearly explain:
* What weather phenomena are present, and
* Why they are occurring, based strictly on the model data.
This is a high-judgment, expert role for meteorologists comfortable diagnosing synoptic and mesoscale features directly from model fields.
What You’ll Do
* Interpret HRRR outputs at a specific time step
* Identify key atmospheric features (e.g., troughs, jet streaks, fronts, instability, convection)
* Explain the physical mechanisms driving observed conditions (e.g., vorticity advection, lift, moisture transport, upper-level divergence)
* Produce clear, structured, technically accurate written explanations
All labels must be grounded strictly in HRRR data (no radar/satellite/AFD references).
Requirements
* Degree in Meteorology / Atmospheric Science OR 2+ years of operational forecasting experience
* Strong synoptic and mesoscale analysis skills
* Deep understanding of atmospheric dynamics and thermodynamics
* Clear, precise technical writing ability
Evaluation Process
* Behavioral interview (forecast reasoning & decision-making)
* Short technical assessment (500 mb chart interpretation)