is transforming meteorology by developing novel techniques for gathering and assimilating weather data into real-time analysis systems. One of the most important tools we leverage to accomplish this is robust statistical analysis of both our observations and the forecast model output we generate in-house and consume from public agencies.
As an Atmospheric Data Scientist, you’ll lead our efforts to build new statistical forecasting systems which combine these observational and model to produce the best forecasts possible for our clients. You have a background in statistical applications in the geosciences, and understand how to extract the signal from the noise of many disparate forecasts. You’re comfortable wielding a diverse toolkit to tackle these problems, including ensemble/timeseries analysis techniques, bias correction procedures, and machine learning. A successful candidate will leverage their knowledge of these tools to prototype new statistical forecasts and analyses applied to massive meteorological datasets.
What You'll Be Doing
- Lead initiatives to develop novel ensemble statistical analysis/post-processing systems to combine unique observations and model data to produce the best possible weather forecast
- Develop novel applications for machine learning to build dynamic, self-correcting forecast systems which iteratively update and refine themselves as new data arrive into ClimaCell’s unique collection of weather observations
- Help develop robust validation procedures and conduct verification studies across the company’s data product portfolio, to ensure that our forecasts are always one step ahead of the changing weather
What You Bring
- Extensive background in statistical and/or machine learning applications to weather forecasting and data analysis
- Experience working with or developing state-of-the-art ensemble forecasting systems and analyses, such as NOAA’s National Blend of Models or NCAR’s DiCast system
- Knowledge of and familiarity with operational ensemble numerical weather prediction systems such as NOAA’s GEFS or ECMWF’s EPS
- Experience building statistical modeling tools using scientific Python (particularly NumPy, pandas, scikit-learn, statsmodels, or related packages) or R
- Familiarity with Linux
- Experience working on cloud computing systems, especially Amazon AWS or Google Cloud
- Experience with other scientific Python libraries or frameworks, especially those used widely in the geosciences (SciPy, sklearn, skimage, xarray, Numba, etc.) or the R “tidyverse” (dplyr, purrr, broom, etc)
- Familiarity with building data processing pipelines and databases to support big data statistical analysis applications
- A Masters or PhD in statistics, mathematics, meteorology, or any other field with corresponding coursework and application in atmospheric science