Undergraduate Scholarships

The characteristics of numerically simulated clouds and convection depend on the resolution of weather and climate models. Subgrid clouds are parameterized in coarse-resolution models but are often resolved at higher resolutions. Such clouds are essential in understanding shallow convection and can significantly affect the radiation budget if unaccounted for in our current models. This project aims to quantify the characteristics of subgrid clouds by comparing several associated cloud and radiation fields simulated at different resolutions from a numerical weather prediction model to ground-based and available satellite observations. The main objective of the project is to understand how well sub-cloud variability is captured to varying resolutions in model simulations.

Requirements: The project requires Python programming skills in analysing data.

Supervisors: Dr. Abhnil Prasad and Prof. Steven Sherwood

Large-scale climate modes such as El Niño-Southern Oscillation, Southern Annular Mode, and Indian Ocean Dipole significantly influence weather and climate variability across Australia. These modes are typically quantified using Sea Surface Temperature (SST) data from specific regions of the ocean. This project aims to compute these climate indices using simulations from Australia's seasonal forecasting system, ACCESS-S2. ACCESS-S2 is a fully coupled dynamical model operated that provides seasonal climate forecasts. Specifically, the project will derive large-scale climate mode indices from historical SST outputs of ACCESS-S2 and compare these results with satellite-derived SSTs.

The project is expected to commence in July.

Requirements: The successful applicant should have strong programming skills in Python.

Supervisors: Dr. Mandy Freund and Dr. Sanaa Hobeichi

Global climate models (GCMs) are essential for projecting long-term climate trends, but their coarse spatial resolution limits their ability to capture regional climate variability and extremes. To address this, high-resolution regional downscaling has been carried out over Australia, in collaboration with the Bureau of Meteorology, CSIRO, the NSW Department of Climate Change, Energy, the Environment and Water (DCCEEW), the University of New South Wales, and the University of Queensland.

This project will investigate the added value of the four regional downscaled simulations compared to global CMIP6 models, with a focus on identifying improvements in the representation of climate extremes and regional patterns. The student will perform inter-model comparisons using output from the downscaled simulations and GCMs, helping to evaluate the reliability and relevance of these projections for climate adaptation and planning.There will be an additional focus on whether bias correction will influence the outcomes of the added value datasets.

Through this work, the student will develop skills in analysing high-resolution climate datasets, interpreting model outputs, and working with scientific programming tools commonly used in climate science.

Requirements: Basic proficiency in Python or another scientific programming language is required. Familiarity with climate data formats (e.g., NetCDF) or experience working with large datasets is beneficial but not essential.

Supervisors: Prof. Jason Evans, Dr. Ulrike Bende-Michl, Dr. Christian Stassen

This project aims to develop an agentic AI tool for climate applications using NVIDIA AI platforms and tools, OpenAI’s agent builder, or similar frameworks. The focus will be on creating an intelligent chatbot capable of engaging, informative climate-related conversations. The student will experiment with approaches such as fine-tuning large language models or implementing retrieval-augmented generation (RAG) pipelines, exploring how different platforms perform in this context. We are looking for a self-driven student with a strong interest in artificial intelligence and climate science.

Requirements: We are looking for a self-driven student with a strong interest in artificial intelligence and climate science. A background in computer science, machine learning, or related fields is preferred, though enthusiasm and initiative are equally valued.

Supervisors: Dr. Sanaa Hobeichi and Prof. Alex Sen Gupta

Build a tool that benchmarks and visualizes how common scientific array operations scale with dataset size and chunking (NumPy vs xarray eager vs xarray+Dask), producing interactive “scaling curves” and “chunking maps” that teach users how performance changes.

Requirements: A student who is comfortable writing Python, curious about performance, willing to run controlled experiments, and interested in understanding how scientific code behaves as data sizes grow.

Supervisors: Dr. Sam Green and Dr. Sanaa Hobeichi