Milan Curcic
Email | GitHub | Twitter | Medium | Google Scholar
Welcome to my page! I'm an academic scientist at the University of Miami's Rosenstiel school. I study ocean waves and their role in Earth System prediction. I enjoy making software, especially when there's an interesting overlap with science. I'm a proud dad of two boys, Nolan (3-year old) and Liam (deceased at birth on August 9, 2022).
Here are some projects that I've been working on:
- SUSTAIN laboratory: I do research and science support in the lab. I study wave growth and air-sea fluxes using theory, measurements, and numerical models. I aldo help set up lab experiments, perform data quality control and management, and write software in support of the lab.
- Coastal Land Air-Sea Interaction (CLASI): An ONR-DRI project that executes targeted field campaigns toward developing coast-aware air-sea parameterizations for operational weather prediction models. I've been actively involved in the assembly of ASIS spar buoys, their deployment and recovery from resarch vessels, data quality control and science algorithms.
- neural-fortran: A parallel deep learning framework for Fortran applications. It supports training and inference for arbitrarily deep dense and convolutional networks. Neural-fortran has been developed to solve the two-language problem of machine learning (mostly Python) for Earth System models (mostly Fortran). It's been successfully used to accelerate various models from chemistry and combustion to weather and climate.
- CloudDrift: An open-source Python package that accelerates the use of Lagrangian ocean data and its analysis. It provides common Lagrangian analysis functions and readers for custom Lagrangian datasets. CloudDrift is led by Prof. Shane Elipot at the Rosenstiel School and funded by the NSF Earthcube program.
- Cloudrun: A SaaS for running WRF in the cloud without the need for a supercomputer or weather modeling expertise. Cloudrun has provided short-fuse WRF forecasts for high-end competitive sailing teams, government research labs, universities, and startups.
- Henet: Developed the first observational platform for objective and accurate wave height measurement for XXL Surf competitions.
- Modern Fortran: Building Efficient Parallel Applications: A programming book that teaches modern Fortran programming to beginners. Rather than a language reference manual, it focuses on hands-on, practical examples to teach the most importrant Fortran concepts. As of Q1 2023, it's sold more than 3000 copies and is the second highest rated Fortran book on Amazon. Source code for all the projects in the book is available on GitHub.
- Fortran-lang: I co-founded Fortran-lang in December 2019 with Ondřej Čertík (creator of SymPy and LFortran) to provide an open-source community and space for people to collaborate on modern Fortran tooling and library ecosystem. Fortran-lang has since been grown by more than 100 open-source developers, has participated in the Google Summer of Code program, and has provisionally been accepted as a supported project of NumFocus. Nowadays I mainly run Fortran-lang's Twitter account and occasionally help out with administrative tasks and mentoring Google Summer of Code students.
- Daily Dune: The best Dune quotes reader on the internet (seriously).
- University of Miami Wave Model (UMWM): A 3rd generation spectral wave model that I developed with Prof. Mark Donelan during my PhD at the University of Miami. UMWM has been succesfully used to simulate global swell and windsea, waves in coastal and hurricane environments, wave-induced transport (Stokes drift), ancient Martian seas and methane lakes on Titan, as well as laboratory waves.
- Fortran libraries: I developed and maintain datetime-fortran and functional-fortran libraries.