Postdoctoral Fellow I: Convective Prediction and Machine Learning Scientist

  What You Will Do:

The selected postdoctoral fellow will contribute to ongoing research activities within NCAR’s Mesoscale and Microscale Meteorology (MMM) and Computational and Information Systems (CISL) laboratories related to improving predictions of high-impact convective weather phenomena (e.g., hailstorms, tornadoes, lightning) using machine learning (ML) and artificial intelligence (AI). The fellow will be expected to identify novel ML and statistical approaches and apply the techniques to generate post-processed predictions of convective weather hazards on timescales of hours to days over the contiguous U.S., as well as use these predictions to better quantify the practical predictability of convective hazards.

The fellow will have flexibility to pursue research aligned with the broad goals noted above. Specific research foci may include: 1) identifying new and existing ML techniques to produce probabilistic convective weather predictions and quantifying their strengths, limitations, and biases compared to contemporary methods, 2) integrating new observational datasets of convective storms, e.g., from satellite-based sensors, within ML systems to generate and validate predictions of convective hazards, 3) testing new ML training strategies, such as transfer learning, using archived and real-time NWP forecasts, and 4) applying explainable AI methods to increase fundamental understanding of ML model behavior and identify new forms of post-processed NWP guidance.

The postdoctoral fellow will interact extensively with NCAR scientists, staff, and other postdoctoral fellows to achieve the research goals over the position’s two-year term. The fellow will have the opportunity to collaborate with NOAA/GSL scientists and assist with the demonstration of promising ML-based hazard guidance. The fellow will be encouraged to present at scientific conferences, publish scientific papers, and pursue professional development activities. Opportunities will exist to contribute to diversity, equity, and inclusion (DEI) activities, including mentoring students through summer research programs.

Responsibilities:

Conduct independent and collaborative research related to applications of ML to the prediction of convective-scale hazards. This may include exploring the potential of ML algorithms to improve post-processed severe hazard guidance, testing new ML training techniques, integrating severe hazard observational datasets for ML training and forecast validation, and applying ML interpretation techniques to improve understanding of ML model predictions.

Design and conduct retrospective and real-time experiments to demonstrate the utility of ML predictions using archived and operational NWP forecast datasets. Assist with NOAA testbed activities, as needed.

Prepare research results for presentation at conferences and publication in peer-reviewed journals. Assist with the preparation of summary reports or project progress reports as needed.

Participate in professional development activities, such as interacting with other NCAR scientists and postdoctoral fellows. Contribute to DEI activities, such as assisting with the mentorship of a SOARS protégé.

What You Need:

Education and Years of Experience

• Ph.D. degree within the last 5 years or expected within the next 6 months in atmospheric science, computer science, statistics, or a related area.

Knowledge, Skills, and Abilities

• Basic understanding of ML and other statistical learning algorithms.
• Basic understanding of convective-scale hazards (e.g, tornadoes, hailstorms) or other environmental hazards and their prediction.
• Advanced skills in scripting languages (e.g., python/bash/NCL), especially related to processing/analyzing model output.
• Demonstrated ability to work independently and collaboratively as part of a research team.
• Excellent time management and organization skills.
• Demonstrated and effective written and oral communication skills.
• Shared values with NCAR’s commitment to diversity, equity and inclusion

Desired:

• Previous experience applying ML techniques to atmospheric science problems, as demonstrated by prior research.
• Advanced understanding of ML and statistical learning algorithms.
• Working knowledge of the scientific python stack (e.g., numpy, scipy, etc.) and python machine learning-related packages (e.g., scikit-learn, tensorflow, etc.).
• Familiarity with state-of-the-art convective-scale prediction tools, including convection-allowing models, ensembles, and convective-scale forecast verification methods.

Applicant notes:

• A cover letter is required.
• An Inclusion Statement will be required for all applicants applying to this position. This statement should address past efforts, as well as future vision and plans to advocate for and advance diversity, equity, and inclusion in the organization and/or field of work.
• A pre-employment screening is conducted in conjunction with an offer for employment. This screening may involve verifying or reviewing any of the following relevant information: restricted parties screening, employment verification, performance records of internal candidates, education verification, reference checks, verification of professional licenses, certifications, and Motor Vehicle Records.  UCAR complies with the Fair Credit Reporting Act (FCRA).
• Please note that while the position description details both minimum requirements as well as desired skills and experience, we want to remind applicants that you do not need to have all the desired skills and experience to be considered for this role. If you have the passion for the work along with experience in a related field, you are encouraged to apply. We can provide on-the-job training for the rest.
• For more information about our commitment to diversity, equity, and inclusion, here is the link to the Office of Diversity, Equity & Inclusion Strategic Plan and to the ODEI landing page.