Machine learning in Urban Hydrological Modelling
Urbanization has significantly changed hydrological processes. However, the physically-based urban hydrological models require a large amount of data and mills computation resources.
The available data in Oslo, as well as in many other cities, do not support an accurate simulation of hydrological processes using these models. With the development of computer science, machine learning has developed fast in previous years. The machine learning methods, such as artificial neural network and vector machine are used in hydrological simulation and uncertainty analysis.
This master thesis will apply machine learning in urban hydrological modeling. The students will use different machine learning methods to predict runoff in urban catchments in Oslo.
The model results will be compared with in situ measurements and simulations by physically-based models, such as the Storm Water Management Model. The students will affiliate under two ongoing projects-New Water Ways (https://newwaterways.no/) and Sustainable Urbanization in the Context of Economic Transformation and Climate Change: Sustainable and Liveable Cities and Urban Areas.
The student will contribute to these two projects and gain knowledge about stormwater management at both national and international levels.