Science and Technology Center (STC), https://leap.columbia.edu/ , a multi-institutional center effort meant to improve climate projections using novel artificial intelligence for better climate adaptation. This position will be based at LEAP’s offices in New York City.
The goal of this project is to develop a new multi-scale representation for Earth system modeling that can be used in fully differentiable generative modeling frameworks (e.g., flow matching, diffusion, and stochastic interpolants).
The first objective will be to expand currently available multi-scale representations in two directions: 1) Considering wavelet-based representations that are optimized for the geometry of weather and climate maps, including e.g. spherical wavelets; 2) Considering other approaches to multi-scale modeling.
These will be tested on well known partial differential equations (PDE) systems that regulate fluid dynamics (for example, turbulent radiative layer and Rayleigh-Benard convection), and compared to pixel-space representations in both deterministic and generative approaches.
Next, the ARS will use the new tools and apply them to case studies in weather and climate modeling, considering 1) emulations of ocean and atmospheric systems, focusing on global scales; 2) generative downscaling.
This ARS will work with LEAP scientists to build the workflows that allow for rapid production, analysis, and emulation of data products and to disseminate findings to research group leaders in LEAP and the wider research community.
The ARS will closely collaborate with graduate students, postdocs, and other staff within LEAP, as well as the TerraD2I Lab at CUNY led by Dr. Viviana Acquaviva.
We are committed to building a community of scientists with a range of academic and professional backgrounds, and believe that a variety of perspectives and experiences is essential to advancing our research and mission.
Minimum Qualifications:
- A Ph.D. in Data Science, Computer Science, Physics, Earth System Science or a directly related discipline is required by the start of the appointment.
- Strong programming skills are a requirement.
Preferred Qualifications:
- Post-doctoral experience and demonstrated experience in Earth System Science, Data Science, or similar.
- Fluency in Python, including deep learning frameworks such as PyTorch.
- Advanced experience in generative AI modeling, including flow matching and diffusion models.
- Experience with climate data formats (netcdf, zarr, xarray) and with designing and running data analysis pipelines for global climate simulations.
- Experience with signal processing tools is desirable.
- Experience with distributed training frameworks on high performance computing and cloud-based systems.
- Excellent command of the English language (verbal and written) and strong communication skills are desired.
Applications must include: (a) curriculum vitae (b) statement of research (optional) (c) names of at least three references who may be asked to provide letters.
Columbia University is an Equal Opportunity Employer / Disability / Veteran
Pay Transparency Disclosure
The salary of the finalist selected for this role will be set based on a variety of factors, including but not limited to departmental budgets, qualifications, experience, education, licenses, specialty, and training. The above hiring range represents the University’s good faith and reasonable estimate of the range of possible compensation at the time of posting.
