AI for Climate Action- Development & ML Engineering Support

Job title:

AI for Climate Action- Development & ML Engineering Support

Company

United Nations

Job description

Details:Mission and objectives:Mission
To create a unique nexus where civil society meets and engages the academy, industry, and government to collectively mitigate risks and make smarter decisions toward more resilient, democratic, and prosperous communities. As a centre of excellence, GCRI excels in research, innovation, and capacity building across enterprise risk and innovation management, addressing the societal impacts of technological disruptions and financial evolution. Uniting experts from economics, finance, policy, and technology, we are committed to building interdisciplinary tools, capacities, and communities for risk mitigation, resilience building, and sustainable developmentObjectives
GCRI’s objectives include:
1. Global Knowledge Exchange: Creating a platform for interdisciplinary discussion and research on risk and resilience, fostering global collaboration among industry leaders, academics, policymakers, and practitioners
2. Standardization and Metrology: Developing and promoting standards and measurement practices to enhance the reliability and interoperability of risk management strategies.
3. Systems Innovation: Building tools, capacities and communities for the development and application of risk management technologies, standards, and innovation ecosystems.
4. Scholarly and Industry Publications: Publishing standards, reports, and scholarly articles to contribute to academic and practical discourse in risk management.
5. Liaison with Global Entities: Engaging with civil society and maintaining technical interactions with international entities to promote a unified approach to risk management challenges.
6. Advanced Study: Investigating fundamental and applied domains related to risk management, including natural and human-induced risks, and developing advanced tools for risk intelligence and emergency managementContext:Climate change is increasingly recognized as a significant driver of extreme weather events and environmental instability, affecting critical systems such as water, food, energy, and health (WFEH). The AI for Climate Action and Risk Management initiative is a specialized project designed to harness the power of advanced artificial intelligence (AI) and machine learning (ML) to tackle climate-related risks. Using state-of-the-art techniques in data science, multi-hazard modelling, and predictive analysis, this project aims to create tools that aid in climate resilience and adaptive response through effective early warning systems and risk assessment models.Key Objectives:Develop AI/ML models that accurately assess climate risks, predict extreme events, and model cascading hazards, covering all relevant sectors within the WFEH Nexus.
Integrate a full spectrum of data requirements, including satellite imagery, environmental sensors, and real-time weather data, to enhance the accuracy and reliability of risk assessments.
Support a multi-hazard approach that captures the complexity of climate-related events, building a layered risk model capable of analyzing interdependencies and interactions across various climate risks.
Enable early warning systems that support timely, data-informed decision-making for communities, policymakers, and stakeholders facing climate risks.Project Scope:This initiative is part of the Global Centre for Risk and Innovation (GCRI), working within the Nexus Ecosystem framework to advance climate resilience and adaptation strategies. The project’s outputs are intended to benefit communities worldwide, providing practical tools for climate adaptation, disaster forecasting, and resilient infrastructure planning.Advanced Methodologies:Multi-source data integration from diverse climate, economic, and environmental datasets.
Complex matrix-based risk modelling, enabling a comprehensive understanding of individual and cumulative hazard impacts.
Machine learning techniques increase predictive accuracy, such as ensemble models, neural networks, and probabilistic modelling.
Stakeholder-specific insights for populations, infrastructure, and ecosystems, tailoring risk analysis to sectoral needs and vulnerability profiles.Technical and Data Requirements:Climate and weather data (historical and real-time) from trusted sources like remote sensing and meteorological databases.
Environmental and economic datasets to assess resource vulnerabilities and economic impacts.
Stakeholder-specific data to map exposure and sensitivity for targeted resilience planning.
The initiative will produce decision-ready insights for climate risk mitigation, adaptation financing, and long-term sustainability by fulfilling these data and modelling requirements.Task type:Technology DevelopmentTask description:6 Online Volunteers to support on Multi-Source Data Collection & PreprocessingDeliverables: Curated, structured datasets ready for AI/ML model inputs across various hazard types
Responsibilities:
Aggregate data from diverse sources, including GCRI datasets, open-source environmental databases, satellite imagery, and real-time sensors.
Preprocess data for complex matrix-based risk modelling, addressing all hazards (floods, droughts, storms, wildfires, etc.).
Implement data normalization, cleaning, and quality assurance for consistency in AI inputs.
Collaborate on metadata documentation, ensuring data accessibility and transparency for team-wide use.
Frequency: Weekly data updates, monthly summaries on data integration status and quality checks.6 Online Volunteers to support on Advanced Model Development (Multi-Hazard AI/ML Models)Deliverables: Multi-hazard AI/ML models capable of climate risk prediction, hazard interaction analysis, and trend forecasting
Responsibilities:
Develop machine learning models to predict climate risks and assess hazard interactions within complex matrices, supporting early warning systems.
Focus on adaptive model architecture in TensorFlow, PyTorch, or similar frameworks to optimize for high-impact, multi-hazard scenarios.
Incorporate advanced techniques, such as ensemble learning, neural networks, and probabilistic modelling, for enhanced predictive accuracy.
Work with risk layering to model interdependencies and cascading effects between hazards.
Frequency: Bi-weekly updates on model development progress, with an end-of-cycle deliverable for a fully functional model.6 Online Volunteers to support on Stakeholder Impact & Vulnerability InsightsDeliverables: Metrics and insights focused on stakeholder vulnerabilities and exposure across hazard layers
Responsibilities:
Develop and integrate stakeholder-specific metrics to identify and quantify vulnerabilities for populations, infrastructure, and resources within the Nexus Ecosystem.
Analyze risk model outputs to assess stakeholder exposure to multi-hazard scenarios.
Create and validate insights reports highlighting critical vulnerabilities and supporting targeted resilience measures.
Ensure feature extraction and parameter settings are aligned with sector-specific needs, such as water, food, energy, and health.
Frequency: Bi-weekly insights reports, with updates on stakeholder metrics and vulnerability analysis.6 Online Volunteers to support on Model Testing, Validation & CalibrationDeliverables: Rigorous, validated models ready for deployment in predictive risk assessments
Responsibilities:
Perform comprehensive model testing and validation across multiple hazard categories and risk layers.
Implement calibration techniques to enhance model precision and accuracy for high-impact applications, especially in forecasting complex, cascading events.
Develop and maintain testing scripts, documenting performance metrics such as accuracy, robustness, and computational efficiency.
Generate and share validation protocols, supporting best practices for disaster forecasting and early warning reliability.
Frequency: Weekly testing summaries and monthly calibration reports, ensuring model reliability and accuracy.6 Online Volunteers to support on Documentation, Integration & DeploymentDeliverable: Complete documentation and deployment protocols for models, data processes, and risk frameworks
Responsibilities:
Document all stages of data processing, model development, and validation, ensuring transparent and reproducible workflows.
Develop comprehensive deployment workflows that integrate models within GCRI’s systems, focusing on seamless data and model interoperability.
Oversee the integration of AI/ML outputs into decision-support tools and early warning platforms, coordinating with team members for smooth operationalization.
Maintain communication channels and provide progress updates to synchronize all roles for timely project delivery.
Frequency: Continuous documentation updates, with consolidated monthly integration and deployment reports.Requirements:Data Collection & Integration – Proficiency in gathering and integrating diverse datasets, including environmental, meteorological, and economic data, from multiple sources such as remote sensors, satellites, and public databases.Data Preprocessing & Cleaning – Advanced skills in data wrangling, cleaning, and preprocessing, ensuring high-quality, consistent data inputs for modeling. This includes handling missing data, normalization, and feature engineering.Machine Learning & Deep Learning Frameworks – Expertise in using AI/ML libraries like TensorFlow, PyTorch, and Scikit-learn for model building, fine-tuning, and deployment.Multi-Hazard Risk Modeling – Ability to design complex, matrix-based risk models that capture interactions across multiple hazards (floods, droughts, storms, etc.), using advanced statistical and probabilistic modeling techniques.Time Series Analysis – Proficiency in analyzing time-series data, especially for climate and environmental patterns, using libraries such as Pandas, NumPy, and specialized tools for temporal data modeling.Geospatial Data Analysis & GIS – Knowledge of geospatial analysis tools (e.g., QGIS, ArcGIS) and experience with spatial data, critical for creating high-resolution risk maps and visualizing climate impact layers.Model Testing & Validation – Skills in model evaluation, validation techniques, and calibration to ensure robust and accurate predictions, including cross-validation, bootstrapping, and hyperparameter tuning.Advanced Statistical Analysis – Strong foundation in statistics for risk assessment, including probability distributions, hypothesis testing, and regression analysis for understanding and predicting climate impacts.Stakeholder Mapping & Vulnerability Assessment – Familiarity with social vulnerability mapping and exposure analysis, creating metrics to assess stakeholder risks and needs within climate impact frameworks.Documentation & Version Control – Proficiency in documentation best practices and version control (Git, GitHub) to maintain transparent, reproducible workflows and effective collaboration in model development.Work Hours: 11 – 15

Expected salary

Location

Canada

Job date

Fri, 15 Nov 2024 02:09:38 GMT

To help us track our recruitment effort, please indicate in your email/cover letter where (un-jobs.net) you saw this job posting.