Joseph Mouallem

Joseph Mouallem

Computational Scientist & Research Software Engineer

Princeton University

GFDL / NOAA

About

High-performance computing and scientific software for large-scale simulation - building the numerical methods and coupled Earth-system models behind next-generation weather and climate prediction.

I am a Computational Scientist and Research Software Engineer at the Cooperative Institute for Modeling the Earth System at Princeton University and in the Weather and Climate Division at NOAA’s Geophysical Fluid Dynamics Laboratory (GFDL). I work at the intersection of high-performance computing and scientific software engineering, designing scalable parallel algorithms and building, optimizing, and maintaining production simulation codes that run on large HPC systems.

I apply that work to numerical methods and coupled Earth-system models for high-resolution weather and climate simulation. My work centers on GFDL’s FV3 dynamical core and the SHiELD modeling framework, where I develop scalable capabilities for coupled atmosphere-ocean-land simulation on modern HPC architectures.

Before joining Princeton and GFDL in 2020, I began working in computational fluid dynamics during my M.Sc. at INSA Lyon with the turbomachinery team, earned my Ph.D. in Mechanical Engineering at the University of São Paulo on multiscale multiphase flows, and held a postdoctoral fellowship at the University of Waterloo studying heated particle-laden turbulence. That trajectory, from classical CFD and multiphase flow, to numerical methods and dynamical cores, to coupled Earth-system modeling, shapes how I approach model development today.

Education
  • Ph.D. in Mechanical Engineering, 2018

    University of São Paulo, Brazil

  • M.Sc. in Fluid Mechanics, 2014

    INSA Lyon, France

  • B.Sc. in Mechanical Engineering, 2014

    ULFG2, Lebanon

Research Areas

High-Performance Computing

Scalable algorithms, parallel numerical methods, performance optimization, and portability across modern HPC architectures.

Scientific Software

Architecture, implementation, testing, and maintenance of production-scale computational physics and Earth-system model infrastructure.

Numerical Methods & Algorithms

Dynamical cores, grid design, nesting, regridding, and transport schemes for multiscale geophysical simulation.

Earth-System Modeling

High-resolution atmosphere, ocean, land, and fully coupled climate modeling with SHiELD, MOM6, and LM4.

Coming soon

A new high-resolution fully coupled configuration

Global storm-resolving atmosphere, eddy-rich ocean, and interactive land; running together in a single, fully coupled system. A first look at what the next generation of GFDL coupled modeling can resolve.

    This animation is from model output only.

    Research

    Earth-system modeling, dynamical cores, and coupled simulation. Click any project for details, figures and animations.

    SHiELD-LM4: High Resolution Coupled Land-Atmosphere Modeling
    Coupled land-atmosphere modeling
    Extending SHiELD into a fully coupled atmosphere-land system to improve precipitation, runoff, and hydrological extremes at kilometer-scale resolution.
    Explore project
    SHiELD-MOM6: High-Resolution Coupled Atmosphere-Ocean Modeling
    Coupled atmosphere-ocean-ice modeling
    Developing the framework that couples SHiELD, MOM6, and SIS2, enabling two-way air-sea feedback during hurricanes at kilometer scales.
    Explore project
    A Minimal, Adiabatic Example of Sudden Stratospheric Warming
    Idealized stratospheric dynamics
    A minimal adiabatic example of sudden stratospheric warming (SSW) using idealized simulations in the GFDL FV3 dynamical core.
    Explore project
    The Duo-Grid and Cubed-Sphere Grid Imprinting
    Eliminating grid imprinting in FV3
    An efficient refinement strategy for the cubed-sphere grid that enables regional simulations in the global FV3 dynamical core without regridding.
    Explore project
    Multiple Same-Level and Telescoping Grid Nesting
    Multiscale simulation in FV3
    A flexible nesting implementation in GFDL’s dynamical core enabling both same-level and telescoping grids for multi-scale simulations.
    Explore project
    Multiple Same-Level and Telescoping Grid Nesting

    Earlier Research — Computational Fluid Dynamics

    Multiphase flow, turbulence, and sub-grid modeling: the numerical foundations behind the Earth-system work above.

    Targeted Particle Delivery via Vortex Ring Reconnection
    Vortex dynamics and particle transport
    A conceptual model for targeted particle delivery using controlled vortex ring reconnection in a duct.
    Explore project
    Induction Heating of Dispersed Metallic Particles in a Turbulent Flow
    Particle-laden turbulence and heat transfer
    Direct Numerical Simulation of inductively heated metallic particles dispersed in a decaying isotropic turbulent carrier gas.
    Explore project
    Induction Heating of Dispersed Metallic Particles in a Turbulent Flow
    Macro-Scale Effects on Sub-Grid Closures in Gas-Solid Riser Flows
    Sub-grid closures for gas-solid flows
    Macro-scale flow topology effects on sub-grid closure models in filtered two-fluid simulations of gas-solid riser flows.
    Explore project
    Macro-Scale Effects on Sub-Grid Closures in Gas-Solid Riser Flows

    Contact

    Interested in high-performance computing, numerical modeling, Earth-system simulation, or scientific software?

    I am always glad to discuss research collaborations, scientific software development, and computational modeling.