Scientific Developer for UFS In-Core Data Assimilation

Overview

We are currently seeking a Scientific Developer for UFS In-Core Data Assimilation (DA) to support the National Weather Service's Office of Modeling and Development (OMD). OMD is responsible for developing complex Earth system models together with its research partners. These numerical weather prediction (NWP) models are deployed on the NWS operational supercomputers to provide weather and seasonal forecast guidance at different timescales from hours to days to months. The NWP models range from global atmospheric models, limited-area (regional) high-resolution convective-scale atmospheric models, hurricane models, global ocean circulation models, cryospheric models, wave models, and land surface models. OMD is involved in all phases of scientific and engineering model development including data ingest, data assimilation, model integration, post-processing, product generation, and validation and verification.

 

The Ideal Scientific Developer for UFS In-Core Data Assimilation will be local to the DC Metro area, and be able to work a hybrid schedule, working 2 days per week in office out of the College Park, MD location, and working from home the remaining 3 days of the week.

Responsibilities

OMD's Unified Forecast System (UFS) Weather Model relies on highly integrated, robust infrastructure to support advanced Earth system prediction. This position is responsible for developing and advancing the core architecture necessary to enable next-generation forecasting capabilities, such as in-core data assimilation, that will support the NWS moonshot goal of a kilometer-scale global model based on the Model for Prediction Across Scales (MPAS) dynamical core. The developer will work with a team of software engineers, physical scientists, and data assimilation specialists to implement dynamic mid-run state management and rewind functionality. The developer will extend existing coupling capabilities to support intra-component (atmosphere-to-atmosphere) exchanges, implement automated regression testing, and leverage modern development practices to ensure the stability, reproducibility, and modernization of the next-generation modeling system.

 

Duties of the Scientific Developer for UFS In-Core Data Assimilation will include the following:

• Lead the development of new UFS Weather Model infrastructure features that enable critical next-generation forecasting capabilities, such as in-core data assimilation (DA) and high-resolution moving nests.
• Incorporate robust state management, including saving and advertising states to the NUOPC coupler, and rewind capabilities, allowing restarts from prior states with no numerical drift.
• Leverage the Earth System Modeling Framework (ESMF) to extend coupling support.
• Design and implement automated tests to ensure that new state management and rewind capabilities remain stable and robust as other changes are incorporated into the UFS Weather Model.
• Work with the DA teams to enable UFS and JEDI builds with in-core DA capabilities.
• Explore and develop the incorporation of incremental analysis updates (IAU) following the rewind step.
• Leverage modern software engineering tools, including AI-assisted coding technologies, to accelerate the development of high-quality infrastructure and comprehensive testing strategies.
• Incorporate modern software development and management practices like Git-based code management, unit testing, continuous integration, and other Agile/DevOps development techniques.
• Rapidly debug complex model infrastructure and integration issues (build failures, runtime errors, coupled system bugs) as they arise.
• Adhere to standards of clear, high-quality documentation to support NWS developers.

 

Qualifications

The Ideal Scientific Developer for UFS In-Core Data Assimilation will have the following:

 

• Background in computer science, software engineering, meteorology/NWP, applied mathematics, or a closely related field with demonstrated numerical modeling and scientific computing experience.
• Demonstrated experience with Earth system coupled modeling frameworks and infrastructure. Expertise with the Earth System Modeling Framework (ESMF) and National Unified Operational Prediction Capability (NUOPC) is strongly preferred. Familiarity with UFS Weather Model and the Model for Prediction Across Scales (MPAS) is highly beneficial.
• Expertise and demonstrated proficiency in modern Fortran and/or C, as well as Python. Experience with C++ is a strong plus, particularly as it relates to Data Assimilation frameworks such as the Joint Effort for Data assimilation Integration (JEDI).
• Proficiency in developing and debugging software in a Linux or similar environment, including shell scripting/automation.
• Experience with data assimilation systems (especially JEDI) and initialization techniques such as Incremental Analysis Updates (IAU) is strongly preferred.Demonstrated working knowledge of modern software engineering practices: collaborative version control with Git, continuous integration, code testing, and documentation.
• Familiarity with AI-assisted code development of high-quality, performant code, or strong willingness to learn.
• Demonstrated experience developing and running Earth system models on high-performance computing (HPC) platforms, including distributed-memory (MPI) parallelization and job schedulers (SLURM, PBS).
• Familiarity with cross-platform package management, build, and test systems, especially CMake and CTest.
• Good verbal and written communication skills in English.
• Ability to work effectively and professionally as part of a diverse, customer-focused team and collaborate with software engineers and scientists across different agencies. The ideal candidate is a self-starter accustomed to independently identifying, prioritizing, and executing development needs.
• Experience working in Agile/Scrum environments is preferred.
• Candidates must be a U.S. Citizen or Green Card Holder to qualify.