Image

About the speaker:

Dr. Ze Jiang is an Australian Research Council (ARC) Early Career Industry Fellow, in partnership with WaterNSW, and a research associate at the University of New South Wales (UNSW) in Sydney, specializing in hydroclimate extremes modelling and forecasting. With a Ph.D. in Water Resources Engineering from UNSW, obtained in 2021, Dr. Jiang's doctoral research led to the development of a novel method called Wavelet System Prediction (WASP), capable of predicting hydroclimate extremes. Dr. Jiang's academic journey began with a bachelor's degree in Environmental Engineering from Hohai University in China in 2012, followed by a Joint M.Sc. program called "EuroAquae - HydroInformatics and Water Management," a partnership between five European universities. Before his doctoral studies, he gained valuable professional experience working for two years at the National University of Singapore (NUS). In 2023, Dr. Jiang further expanded his research horizons as a visiting research fellow at GFZ-Potsdam in Germany, supported by the Helmholtz Information & Data Science Academy (HIDA) grant.

About the seminar:

Forecasting of hydroclimate extremes across a range of timescales is critical for minimizing the socio-economic costs of these events. The El Niño-Southern Oscillation (ENSO), one of the most influential interannual signals in the global climate system, has significant climatic, ecological, and societal effects worldwide. In Australia, the rapid shift from drought to flood poses significant challenges for effective water resource management under a changing climate. Regression-based prediction is commonly adopted even in operational forecasting systems, often necessitating the use of distributional transformations to improve the model specification. One of the issues in such predictions, however, is the marked differences that distinguish the frequency spectrum of the hydrologic response from the predictor variables used. This raises the question of whether there exists an optimal predictor variable transformation that can mimic the frequency spectrum of the observed response. This led to the finding of a new method, the so-called spectral transformation, which modifies predictor spectral representation using wavelet theory for improved system modelling. The method has been successfully applied in different areas, such as hydroclimate downscaling, forecasting, and paleoclimate reconstruction.

In this talk, I will discuss the need to transform predictor variables to improve hydroclimate predictions, with a particular focus on the spectral domain of the variables involved. To allow this method to be used in a variety of different settings, a number of alternatives using wavelet-based approaches are presented as a means of transforming the variance associated with different frequency bands in each predictor variable. The merits of spectral transformations are demonstrated by an example of ENSO forecasting over long lead times. This method can be used for a wide range of applications, not limited to hydro-climatology, as it is a generic method. In addition, an open-source tool, namely Wavelet System Prediction (WASP), has been developed in different languages (R and MATLAB) to assist the use of the method in practice.

Kindly note, Dr. Ze Jiang will be available for one-on-one discussions after the seminar and on 6^th September; please let us know in advance if you'd like to book a slot.