Machine Learning and its relationship to climate change mitigation

 

Content provided by David Moret from DPM Consulting Group

Flood impact and risk assessments are critical reviews which need to be analysed to decrease and mitigate damages to people, community, and infrastructure. The recent flooding events in New South Wales and Queensland have shown how our environment is becoming increasingly susceptible to the impacts of climate change.

Heavy rainfalls and severe weather conditions between February and April 2022 led to loss of life, significant economic costs, fuel, food, and water shortages.

The latest report from the Intergovernmental Panel on Climate Change shows that although greenhouse gas emissions have slowed down in recent years, but have continued to grow, leading to a continuous rise of the average temperature of the planet which has an impact on the rainfall intensity.

As demonstrated by recent events in Queensland and New South Wales, our climate is becoming warmer, characterised by a decrease in the frequency of rainfall events and, simultaneously, experiencing an increase in the intensity of rainfall. Based on the data collected in the past and during the recent flooding events, professionals are able to determine the potential flood impacts of proposed developments and results show the potential change in variables, such as velocities, flood depths etc.

Due to the increased risk of flooding related to climate change and to population growth near coastal or riverine areas, it has also become fundamental to develop quickly and accurately estimate the probability of flooding in real time. It has been demonstrated that machine learning is at the forefront of providing a tool for hydrologists, modellers, and the industry to accurately analyse large amounts of data and provide results which can be easily used by practitioners to better inform hydrological and hydraulic models.

But what is machine learning? As defined by Sara Brown from Massachusetts Institute of Technology “Machine learning is a subfield of artificial intelligence, which is broadly defined as the capability of a machine to imitate intelligent human behaviour. Artificial intelligence systems are used to perform complex tasks in a way that is similar to how humans solve problems”.

Different machine learning models can be used depending on the situation and outcome that the user requires. For example, real-time river height and gauges are fundamental in determining flood forecasting during a rainfall event. However, this information is often inaccurate and affected by different issues (such as loss of data, sensor noise etc.). Machine learning is able to analyse and clean this data, find hidden patterns and make flood predictions which can then be provided to the public as well as emergency bodies.

Machine learning offers a wide range of opportunities, solving complicated issues more efficiently, with less costs, and demonstrating an excellent performance in flood prediction and flood risk assessment1.

Machine learning provides a fundamental tool which should be considered within the development process to not only accurately providing real time warning and predict flooding which could potentially save lives and money, but also offers an instrument to further develop and investigate various aspects of flood impact and risks. Combining machine learning with the traditional methods and techniques of the stormwater industry, allows professionals to offer accurate and robust results which can be used for emergency purposes, minimising costs and saving lives.

1Ighite, EH, Shirawaka, H & Tanikawa, H 2022, ‘Application of GIS and Machine Learning to Predict Flood Areas in Nigeria’, Sustainability, vol. 14, no. 5039.

 

Related resources

Delivering a zero-emission neighbourhood

Songbird Oxley, an EnviroDevelopment certified project in Brisbane, Queensland, is being delivered as one of Queensland’s first zero-emission communities. What does this mean?

Inflation and Sustainability

Energy is a major input into nearly every product or service we purchase. Rising energy prices are influenced by myriad factors many of which are unrelated, for example COVID supply chain friction and the invasion of Ukraine.