Modelling global riparian traffic networks as a complex human-natural system

Project description

Human society and nature are increasingly intertwined systems in the Anthropocene. The global riparian urban traffic networks are typical cases of increasingly interconnected human-nature systems. The complex interconnected dynamics between human society and nature are best modelled as networks where attributed nodes and weighted edges are embedded in space. This project aims to understand the respective structures and factors that characterize the human and natural networks in global riparian urban traffic systems and how such dynamics co-evolved as an integrated human-nature system to achieve sustainable urban development. Integrating urban traffic networks for the top 100 cities by populations and global river reach networks from Landset images by Google Earth Engine from 1990 to 2021, it involves three objectives:
1) Identifying the key structures and factors that characterize the natural network.
Systematic literature review will first be conducted to identify the biophysical factors that influence the development of river reaches. Network clustering using machine learning algorithms will then be designed to capture the reach substructures that characterize the network and model the changing patterns across rivers in time.
2) Identifying the key structures and factors that characterize the human network.
Literature review and network clustering algorithm design will be conducted to identify the socio-economic factors and substructures that characterize the traffic network, and the changing patterns across cities in time.
3) Understanding the co-evolutionary dynamics of the human-natural network and design management strategies for future.
Overlaying the urban traffic network with river reach network, the emergent network structure will be investigated. The biophysical and socio-economic factors identified previously will be input in statistical clustering models to evaluate the level of influences they have on the network structure. Scenario analysis will be conducted to predict how different riparian city structures will respond to changing climatic and socio-economic conditions in the future.

Outcomes

The successful completion of this project will yield the following outcomes:
1) Enable improved understanding, management, and planning of rivers and cities, leading to more sustainable and resilient human-natural systems.
2) Developing an evaluation framework to assess the performance of the network analysis algorithms and their application to real-world river networks and traffic networks.
3) Building collaboration with experts, researchers, and stakeholders in the fields of hydrology, urban planning, and network analysis.
4) Research publications in international conferences and journals.

Essential capabilities

Excellent cross-disciplinary (natural and social sciences) and critical thinking mindset, excellent mathematical understanding and/or engineering problem-solving skills, demonstrated algorithmic thinking and programming skills in Matlab, Python or R.

Desireable capabilities

Demonstrated experiences in processing large-scale geographical information and hydrological data, reasonable knowledge about network science (e.g. Newman’s book on Network science), and reasonable understanding of machine learning libraries like PyTorch and TensorFlow.

Expected qualifications (Course/Degrees etc.)

Bachelor’s degree in environmental science, engineering, mathematics or statistics, or Master’s degree in environmental science, engineering or technology.