STAT41140

Learning Outcomes:

On completion of STAT41140, students should be able to:
- Explain clearly the foundational concepts of neural networks and deep learning architectures used in weather and climate forecasting
- Apply and implement modern AI methods, such as transformers and graph neural networks, to real-world meteorological and climatological datasets
- Critically evaluate the strengths and limitations of state-of-the-art AI-driven forecasting systems, including ECMWFs AIFS
- Effectively interpret and communicate the results of AI-based weather and climate models to diverse audiences, both technical and non-technical
- Independently develop and validate AI solutions addressing practical challenges in weather forecasting and climate analysis

Indicative Module Content:

STAT41140 spans foundational concepts and cutting-edge developments in artificial intelligence applied to meteorology and climate science. Initially, students will explore essential topics including basic neural network architectures, their training methodologies, and validation approaches, supported by intuitive learning methods inspired by Tom Yeh's "AI by Hand" worksheets. From these foundations, the module will cover sophisticated AI architectures such as transformers, graph neural networks (GNNs), and attention mechanisms, examining how these powerful techniques improve predictive accuracy. Students will gain in-depth
knowledge of operational forecasting models, notably ECMWF's Artificial Intelligence Forecasting System (AIFS), analysing
their implementation details, strengths, and limitations through real-world case studies. Practical hands-on sessions will teach students to implement these advanced AI techniques using Python-based frameworks, such as PyTorch, and apply them directly to meteorological and climate datasets. Throughout, there is an emphasis on critical evaluation and clear communication, ensuring that students understand not only how to construct robust AI models, but also how to rigorously assess and effectively present their predictive capabilities.

UNESCO highlights a set of key competencies that support individuals in their development and support society in achieving the UN Sustainable Development Goals (SDGs). UCD has adapted these competencies and is combining them with others to form a general framework of learning competencies. This module will help you develop the competencies below to the levels specified. 

The United Nations identified seventeen Sustainable Development Goals (SDGs) as core to the 2030 Agenda for Sustainable Development, and UCD contributes in general to SDG 4 Quality Education. Further SDGs explored within this module if relevant are listed below. A scale of 1 - 5 indicates the extent to which the SDG is covered.


 

Student Effort Hours:

Approaches to Teaching and Learning:

The teaching and learning approaches in STAT41140 emphasise active, practical, and collaborative learning. Students will engage with interactive lectures that introduce theoretical concepts in an accessible, applied context. Guided computer sessions will offer extensive hands-on practice, enabling students to directly implement AI methods on real meteorological data using Python-based AI frameworks. Each component of the module will feature structured individual and group tasks, fostering peer learning and teamwork as students collaboratively address real-world forecasting problems. Reflective learning is encouraged through critical
evaluations of model performance and limitations, while student presentations will assist in developing effective scientific communication skills. The module incorporates enquiry-based learning, where students actively investigate case studies from operational forecasting systems, deepening their understanding and preparing them to tackle practical meteorological and climate-related challenges.

Requirements, Exclusions and Recommendations

Learning Requirements:

Students need to have a foundational understanding of linear algebra and calculus, with modules taken at undergraduate level in each of these subjects. Some familiarity with Python programming environments such as VS Code or Google Colab is recommended.

Module Requisites and Incompatibles

Not applicable to this module.
 

Assessment Strategy

As part of UCD's student support, under the Additional Consideration policy, extensions may be available for the following assessments in the module: Assignment (including essay/poster), Portfolio, Reflective Assignment, Report(s), and Individual Project.

 

Please see Student Jargon Buster for more information about remediation types and timing. 

Feedback Strategy/Strategies

• Group/class feedback, post-assessment

How will my Feedback be Delivered?

Feedback in STAT41140 is designed to be continuous and developmental. Students will receive formative feedback on draft work and coding activities through in-class discussions, written comments, and 1:1 support. Peer review will be used in class sessions to encourage collaborative reflection on coding quality and modelling approaches. Short written assignments will be discussed in class, with model answers and class-wide feedback provided. Throughout, students will be encouraged to use self-assessment checklists aligned with the modules learning outcomes.

Timetabling information is displayed only for guidance purposes, relates to the current Academic Year only and is subject to change.