Loading Events
This event has passed.

The impacts of urbanisation on long and short-term water quality changes are of central concern in many estuaries and coastal waters. However, nutrient data are often sparse in time and space and have non-linear responses to environmental factors, making it difficult to systematically analyse spatial and temporal trends and therefore quantify their controls.

This research has demonstrated that random forest (RF) and gradient boosting machine (GBM) models can be successfully used for interpolation and simulation of nutrient and oxygen concentrations, across a range of hydrological systems and hydrological conditions.

Improved prediction accuracy and lower error were achieved using hybrid RF and GBM methods, compared to stand alone or more traditional interpolation/extrapolation methods. This work has demonstrated that machine learning models are an essential tool to fully utilise all available water quality and hydrological data for water quality modelling, to facilitate the exploration of spatial and temporal signals in groundwater and surface water nutrient data, and ultimately the management of receiving waters.
Benya Wang is a PhD candidate at the School of Engineering and School of Agriculture and Environment. He graduated with an environmental science degree in 2014 from Zhejiang University, China. Benya’s main research interest is using data-driven models to improve our understanding nutrient transport between surface landscapes, surface water, and groundwater.