I'm a physical oceanographer from Mexico, finishing my M.Sc. in Earth Sciences (Meteorology & Physical Oceanography) at UNAM. I build and run numerical ocean models, from idealized process studies in MITgcm to an operational regional forecast with HYCOM.

🔭 Seeking a PhD starting in 2027 in polar ocean modelling: ice–ocean interaction, internal waves and mixing in polar oceans, and machine learning for ocean models. I'd love the PhD to include Antarctic or Arctic fieldwork.

Advisor: Dr. Ana Karina Ramos Musalem · Presented at EGU General Assembly 2025

Idealized MITgcm experiments of a bay the size of Sebastián Vizcaíno Bay (Baja California), forced by a 2.8 h wind pulse, under linear, realistic and two-layer stratification, each with a no-bay control run.

• The 3.8 h seiche period does not depend on stratification, but its decay does (8.3, 9.0 and 9.9 h).
• Barotropic-to-baroclinic conversion at the mouth explains 18, 10 and 1 % of the seiche decay.
• The seiche is not the main pathway: most of the baroclinic energy leaving the bay is its near-inertial response (1.01–1.07 f), which exits as a sloping beam and as coastal internal Kelvin waves.

➡️ Model configuration · Analysis & figures

• Built and operate a 1/25° HYCOM forecast for the Mexican Pacific (1.8 M grid points, 36 hybrid layers), forced by WRF winds, rivers and 8-constituent TPXO10 tides with self-attraction and loading. It provides the total-water-level component of CENAPRED's early-warning system for storm surge and extreme events.
• Maintain the daily HYCOM forecast for the Gulf of Mexico on UNAM's HPC cluster, delivering SST, SSH and heat-content products to PEMEX, CONAGUA and civil-protection authorities.
• Operate a PyGNOME Lagrangian dispersal system for oil-spill response, and wrote a tested Python toolkit (265 tests) that compares HYCOM, ESPC and Copernicus Marine forecasts.
• Designing a two-way WRF–HYCOM coupled system (ESMF/NUOPC) for tropical cyclones in the eastern Pacific.

(Operational code is private to the institution.)

MITgcm simulations of rotating-tank upwelling over two identical canyons. The upstream canyon always upwelled more than the downstream one, so the canyons interact dynamically; the difference shrinks as the Rossby and Burger numbers grow.

Research cruise on R/V Justo Sierra, Gulf of Mexico (May 2024): hydrographic, biological and sediment sampling, data QC.

Ocean & climate models: MITgcm · HYCOM · WRF forcing · PyGNOME · ESMF/NUOPC · TPXO10 · ERA5 · GEBCO Machine learning: ECMWF Anemoi (graph neural networks on ERA5) · scikit-learn Programming & HPC: Python · Fortran · Bash/C-shell · MATLAB · MPI · SLURM · Linux clusters Data: xarray · NetCDF4 · Dask · NumPy · Matplotlib · Cartopy · Jupyter

🎙️ Creator and host of Pronóstico, a Spanish-language podcast on climate science.

🌎 Spanish (native) · English (professional) · French (basic)