Professor Gab Abramowitz
ABOUT ME
Biography
I'm interested in models of natural systems and especially trying to understand when models are useful for making inferences about a natural system. I work with climate, hydrology and ecology models, primarily in evaluation, benchmarking and uncertainty assessment. I work in the Climate Change Research Centre at UNSW and am a Chief Investigator with the ARC Centre of Excellence for Climate Extremes.
RESEARCH
Research Goals
- To understand the scope and extent of inference we can make with climate model projections
- To quantify the expectations of performance we should have from models of natural systems
- To develop meaningful model evaluation strategies for models of systems with significant internal variability
Research in Detail
Together with students and other researchers, I use a range of climatological, hydrological and ecological observations to try to understand the circumstances in which model simulations are useful, assess the uncertainty in model predictions and help develop international standards for model evaluation and benchmarking. My two main foci at the moment are land surface model benchmarking and model dependence assessment in multi-model ensemble climate prediction.
I am currently a member of the GEWEX Global Land-Atmosphere System Study panel and the management committee for the Community Atmosphere Biosphere Land Exchange (CABLE) model, Australia's community land surface model. I also lead development of modelevaluation.org, a web application, which provides automated land surface, hydrological and ecological model evaluation and benchmarking tools as well as observational data sets.
Climate scientists use ensembles of climate models to get a collection of independent estimates of a prediction problem. Yet climate research teams share literature, data sets and even sections of model code. To what extent do different climate models constitute independent estimates? What is the most appropriate statistical framework with which to define independence? What are the implications of ignoring model dependence?
Supervision Opportunities/Areas
If you are interested in persuing a PhD or Honours project and you think this type of area might interest you, I'd love to hear from you. I'm particularly interested if you have ideas of your own that you'd like to pursue, but there are many potential projects in areas like uncertainty quantification in natural systems modelling, land surface model benchmarking and evaluation or an application of machine learning to climate science that we might discuss. If you have a background in maths, physics or computing there are likely plenty of options.
Students who are successful obtaining PhD scholarships (e.g. an Australian Postgraduate Award or International Postgraduate Research Scholarship) may be offered additional "top-up" funding, on a case-by-case basis.
Advice for prospective students
While you may not have a clear idea of your research topic yet, a PhD project will ultimately be yours to investigate, develop and communicate. Your supervisor will help you make sure that you arrive at a topic that is achievable in 3-4 years, scientifically rigorous, and most importantly, interesting for you. If the topics I've talked about above don't seem quite right, there is a great collection of other academics in the CCRC who might be able to help you. We're all friendly, so please come and talk to us.
The Climate Change Research Centre is a great place to do your PhD. We have a large PhD student cohort that is academically and socially engaged, with students from a wide range of academic and cultural backgrounds. We have an induction process that includes being assigned a student "buddy", to make sure you're aware of everything that might be relevant for you, including a range of social events run by students. As part of the Kensington campus we are only 10 minutes from the centre of Sydney.
TEACHING & OUTREACH
Courses I teach
CLIM1001: Introduction to Climate Change
- Publications
- Media
- Grants
- Awards
- Research Activities
- Engagement
- Teaching and Supervision