Senior Data Scientist – Numerical Weather Prediction

Job not on LinkedIn

November 22

Apply Now

Receive Emails with Similar Jobs

UrtheCast

WebsiteLinkedInAll Job Openings

Artificial Intelligence • Science • Agriculture

UrtheCast is a company that specializes in Earth observation data analytics, offering advanced change detection systems and insights for a variety of sectors including agriculture, finance, insurance, and environmental management. Utilizing satellite imagery and artificial intelligence, UrtheCast provides valuable data that supports decision-making, risk mitigation, and strategic planning for government policies, business strategies, and scientific research. The company's solutions help in measuring impacts of climate change, managing natural resources, and optimizing financial risk strategies, thereby enabling its clients to make informed, data-driven decisions.

0 - 1 employees

🤖 Artificial Intelligence

🔬 Science

🌾 Agriculture

💰 $2M Post-IPO Equity on 2020-02

📋 Description

• Perform complex statistical and comparative analysis on large-scale weather datasets including BRIB2, NetCDF, and NEMSIO formats from multiple data sources • Develop comprehensive validation frameworks to assess the accuracy and skill of weather forecasting models against publicly available operational models and observational data • Conduct forecast verification studies using standard meteorological metrics (RMSE, ACC, bias, skill scores) across multiple atmospheric variables and forecast lead times • Analyze and document the strengths, limitations, and performance characteristics of major operational weather models (GFS, GEFS, CFS, ECMWF IFS, ERA5) • Identify forecast biases, systematic errors, and areas for improvement in weather prediction systems • Evaluate the impact of different initialization times, resolutions, and parameterizations on forecast quality • Write robust, production-quality Python code following software engineering best practices for weather data processing, analysis, and model evaluation • Develop and maintain scalable data pipelines to ingest, process, and analyze meteorological data from multiple sources in various formats (GRIB2, NetCDF, NEMSIO) • Integrate analysis scripts and machine learning models into existing production codebase using modern development workflows • Deploy cloud-based solutions to AWS using AWS CDK (Cloud Development Kit) and infrastructure-as-code principles • Implement MLOps best practices including model versioning, experiment tracking, monitoring, and automated retraining pipelines • Build CI/CD pipelines for continuous integration and deployment of forecasting models and data processing workflows • Optimize code performance for handling large-scale meteorological datasets efficiently • Design, develop, and deploy AI-based weather forecasting models using machine learning and deep learning techniques • Research and implement state-of-the-art approaches in AI weather prediction including neural networks, graph neural networks, transformers, and generative models • Evaluate emerging AI weather models (e.g., ECMWF AIFS) and assess their applicability to business use cases • Develop hybrid forecasting approaches that combine physics-based numerical weather prediction with data-driven machine learning methods • Train models on large historical weather datasets (ERA5, HRRR, GFS archives) using distributed computing resources • Implement probabilistic and ensemble forecasting techniques using machine learning to quantify forecast uncertainty • Optimize model architectures for computational efficiency and forecast skill • Develop derived meteorological features and indices from raw weather data that provide value for industry-specific applications

🎯 Requirements

• Master's degree or Ph.D. in Atmospheric Science, Meteorology, Climate Science, Computational Science, Data Science, Physics, or closely related quantitative field with focus on weather/climate applications • 5-8 years of professional experience in atmospheric science, weather forecasting, climate modeling, or closely related fields • 3+ years of hands-on experience working with operational numerical weather prediction models (GFS, GEFS, CFS, ECMWF IFS, ERA5, or similar) • 3+ years of production-level Python development for scientific computing, data analysis, and machine learning applications • Demonstrated experience processing and analyzing large-scale meteorological datasets in GRIB2, NetCDF, or NEMSIO formats • Proven track record of developing and deploying machine learning models or data science solutions in production environments • Experience with forecast verification, model evaluation, and statistical analysis of weather prediction systems • Strong portfolio demonstrating weather data analysis, visualization, and modeling projects.

🏖️ Benefits

• Competitive compensation • Flexible time off