• Charles Sturt Associate Professor Dr Azizur Rahman research published in prestigious and top-ranked journal, ‘Environmental Research’
• The research investigates ways of measuring the trophic status of water ecosystems using machine intelligence and artificial intelligence which is an evolution of manual measures
• The technique will support data-driven, evidence-based decision-making processes in water resource management and environmental conservation

A Charles Sturt University academic is helping improve the sustainability and health of marine ecosystems through his latest published research.

Associate Professor in Mathematics and Statistics in the Charles Sturt School of Computing, Mathematics and Engineering Dr Azizur Rahman published his international collaborative research paper titled, ‘Data-driven modelling for assessing trophic status in marine ecosystems using machine learning approaches’ in the Environmental Research journal in November 2023.

Dr Rahman said the achievement was particularly exciting, giving the journal’s ranking of 12^th best from 371 journals covering environmental sciences.

“It has been a great privilege to see my article in such a top-ranking journal, with an h-index of 164 which measures both the productivity and citation impact of the publication,” he said.

The research investigates ways of using machine learning (ML) and artificial intelligence (AI) algorithms to develop new model to enhance the ability to assess the trophic status of aquatic ecosystems.

“The trophic status of a water ecosystem essentially measures any high, natural variability with organic carbon matter produced per unit of time and surface in the body of water,” Dr Rahman said.

“It also considers the productivity of an aquatic ecosystem, which not only depends on the nutrients input, but also links to a rise in residence time, a drop in turbidity or grazing pressure.”

Dr Rahman said his research is essential to understanding the health of such ecosystems which society relies on to function.

“We need to improve the way we monitor trophic status to be able to improve this overall sustainability,” he said.

“Traditional modelling outputs suffer from many issues, including imprecision, multicollinearity, insufficient parameter integration and a lack of robustness to adapt to diverse aquatic ecosystems.

“Our proposed ML and AI-based technique is more robust in terms of its effectiveness in assessing trophic status, reliability, accuracy and contemporary data-oriented applicability.”

One example in Dr Rahman’s research considered the Bay water in Ireland.

“Case study results with Irish data demonstrate that the quality of the Bay's water decreases due to various pressures including raw wastewater pressure, domestic waste pressure, agricultural pressure and nutrient loads,” he said.

By using Prof Rahman’s technique, he said it could enhance the reliability and accuracy of trophic status assessments for other types of water bodies.

“If we develop our model with other water bodies, we can expand this to assess a range of water sources in Australia such as Byron Bay or Sydney Harbour, through to inland water bodies such as tourist hotspots like Lake Albert in Wagga Wagga or the Murrumbidgee River, which farmers rely heavily on,” he said.

“We will be able to support data-driven, evidence-based decision-making processes in water resource management and environmental conservation efforts by various stakeholders including researchers, government or non-government policymakers and agencies or organisations who advocate for conservation and societal sustainability.”