Head of Atmospheric Modeling

About Convective

Convective is an early-stage startup building weather modification systems. Our mission is to unlock abundance: ending droughts, suppressing wildfires, greening deserts, and cooling cities.

About the role

We are seeking a Head of Atmospheric Modeling to lead the development our NWP and AI modeling.

You will build the bridge between theoretical validation and operational reality, implementing state of the art NWP and AI weather modeling into a cloud-native environment for real-time data assimilation from monitoring and deployment hardware to enable predictive decision-making for operational deployments.

Though not required, you could be a senior scientist-engineer coming out of NCAR, NOAA, DOE, a national lab, a leading research university, or an AI weather startup/company with deep experience in numerical weather prediction (NWP) and/or next-generation AI weather modeling emulators. You function as both a lead architect and a hands-on builder.

As we are a deployment-focused company, you will need to be excited not just about research and modeling for its own sake, but for the purposes of enabling the next generation of climate adaptation through weather modification deployed within ethical and legal guidelines.

This role includes a significant equity opportunity.

This is an early-stage startup building something that has never been built before. This is a full-time role. You should expect upwards of 50-60 hours per week, sometimes more, with frequent travel and regular weekend work. Priorities shift, roadmaps evolve, and the work products change frequently. We are looking for someone who thrives in that kind of environment. If that does not excite you, this role is probably not the right fit.

What you'll do

Operational Forecasting & Data Assimilation

• System Architecture: Architect and lead the implementation of a proprietary data assimilation system (using frameworks like WRFDA or DART) on cloud infrastructure (AWS HPC).
• Predictive Tooling: Develop the operational modeling chain that ingests historic and observational datasets for high-resolution NWP simulations (WRF, MPAS) for evaluating deployment use cases.
• AI/ML Emulation: Translate complex atmospheric dynamics into fast, actionable inference models (e.g., using FGNs, Diffusion, or Transformer-based architectures) that guide live deployment decisions.

Strategic Modeling Leadership

• Multi-Scale Hybrid Strategy: Lead the architecture of a tiered modeling stack that fuses physics simulations with AI inference across spatial scales. This ranges from Large Eddy Simulation (LES) for modeling plume dynamics and cloud physics; to WRF and AI mesoscale forecasting for predictive modeling; up to Global Climate Models (CESM) to assess long-term deployment opportunities and macro-scale system impacts.
• Team Building: Recruit, mentor, and lead a small, world-class team of meteorologists, data scientists, and atmospheric physicists.

Preferred Qualifications

• BS/MS/PhD in Atmospheric Science, Meteorology, Physics, Computer Science, Data Science, or a related field.
• NWP Mastery: Extensive experience setting up, running, and analyzing outputs from numerical weather prediction models (WRF, MPAS, GFS, CESM, etc.). You may have even contributed to the development of these models in prior roles.
• AI Exploration: Some experience building or fine-tuning AI models for meteorological emulation (e.g., Physics-Informed Neural Networks, Functional Generative Networks, or Diffusion Models).
• Data Assimilation Expertise: Experience in implementing and developing data assimilation systems (WRFDA, DART, etc,). You understand how various observations improve initial conditions for specific use cases.
• HPC & Cloud fluency: Experience implementing NWP models on high-performance computing clusters, such as Linux environments hosted on cloud platforms (AWS, GCP, Azure, etc.).
• Data Engineering: Deep experience working with large meteorological datasets and formats (e.g., NetCDF, GRIB, etc.)
• Programming Stack: Fluency in Python (PyTorch, JAX, TensorFlow) for AI development, and Fortran/C++ for core NWP.
• Scrappiness: Independently manage competing priorities under pressure to meet deadlines, while clearly communicating complex concepts.

PIc1ef79e6c9b1-30511-39692301