Modeling Scientist

Job Title: Modeling Scientist (Uncertainty Quantification)

Department:
Modeling & Analytics

Reports to:
Lead Modeling Scientist

Location:
Remote

Base Salary Range: $100k - $160k base salary

The Modeling Scientist is responsible for improving model traceability, uncertainty quantification, and predictive trustworthiness in Arva's ecosystem model predictions. This role is central to advancing Arva's monitoring, reporting, and verification platform for greenhouse gas emission reductions and removals.

Working at the intersection of statistics, machine learning, and process-based ecosystem modeling, this role works closely with ecosystem modelers and data engineers to design robust model traceability and uncertainty frameworks that support transparent, decision-ready outputs for customers, partners, and environmental markets. The Modeling Scientist plays a critical role in translating scientific rigor into real-world impact through credible, auditable modeling systems.

Primary Job Responsibilities

Uncertainty Quantification and Model Evaluation

• Generate and apply model traceability framework for ecosystem and biogeochemical models to enable rigorous model testing and improvements
• Design and implement uncertainty quantification framework for the models, including parameter, structural, aleatory, and epistemic uncertainties
• Apply sensitivity analysis, multivariate testing, and cross-validation to evaluate model robustness and generalizability across space and time
• Quantify and communicate model confidence, uncertainty bounds, and performance metrics

Statistical and Probabilistic Modeling

• Develop hierarchical and Bayesian approaches to support distributed and iterative model optimization
• Apply probabilistic methods to integrate data, models, and uncertainty across scenarios
• Analyze model outputs to diagnose limitations and inform model improvement strategies

Machine Learning and Model Integration

• Integrate machine learning techniques with process-based or mechanistic models to improve predictive performance and scalability
• Partner with data engineers to implement reproducible, scalable modeling pipelines
• Contribute to the design of model evaluation and optimization workflows

Scientific Communication and Documentation

• Communicate uncertainty, confidence intervals, and model performance clearly to internal teams and external stakeholders
• Contribute to scientific reports, transparent model documentation, and peer-reviewed publications as appropriate
• Support defensible, auditable model outputs suitable for regulatory and credit market review

Key Competencies / Requirements

• 5+ years demonstrated experience in uncertainty quantification, probabilistic modeling, and data model integration
• Advanced proficiency in Python and scientific computing, with experience building reproducible modeling pipelines
• Strong software engineering practices, including writing modular, testable, and well-documented code
• Deep commitment to scientific rigor, transparency, and integrity
• Experience integrating machine learning with process-based or mechanistic models preferred
• Familiarity with ecosystem or Earth system models such as Day Cent or CESM preferred
• Familiarity with cloud platforms and data systems, including AWS and relational or spatial databases, preferred
• Master's or PhD degree or equivalent experience in Statistics, Applied Mathematics, Environmental Science, Earth System Science, Biology, or a related quantitative field

Responsibilities:

• Generate and apply a model traceability framework for ecosystem and biogeochemical models to enable rigorous model testing and improvements.
• Design and implement an uncertainty quantification framework, including parameter, structural, aleatory, and epistemic uncertainties.
• Apply sensitivity analysis, multivariate testing, and cross-validation to evaluate model robustness and generalizability.
• Quantify and communicate model confidence, uncertainty bounds, and performance metrics.
• Develop hierarchical and Bayesian approaches for distributed and iterative model optimization.
• Apply probabilistic methods to integrate data, models, and uncertainty across scenarios.
• Analyze model outputs to diagnose limitations and inform…