Growing populations and expanding impervious area combined with climate changes have increased many water related risks in cities including flooding, water supply shortages, and deteriorating surface water quality. To help mitigate these problems in China, the ‘Sponge City’ concept has been introduced as a new paradigm in stormwater management which involves city-wide implementation of more sustainable and multifunctional stormwater infrastructure including green roofs, rain gardens and permeable pavement. By providing storage, evaporation and infiltration where rainfall occurs, this infrastructure can reduce flooding and help to recharge groundwater resources within cities, while at the same time supporting biodiversity and improving urban livability and climate. Over three studies, this thesis has explored the hydrologic impacts of Sponge City implementation through the use of remote sensing and hydrologic modelling methods.
The first study employed remote sensing methods to produce detailed land cover information from satellite imagery to evaluate the potential for Sponge City implementation within a case study area. This study indicated that city-wide land cover could be identified with a reasonable degree of accuracy based on satellite imagery especially if combined with surface elevation and land parcel boundary data. In a second study, a detailed hydrologic model of the case study area was developed to evaluate a variety of Sponge City scenarios under historic rainfall and temperature. Results indicated that the target of capturing 85 % of annual rainfall is likely achievable even at the city or catchment scale with a feasible extent of new Sponge City infrastructure. The third study focused on a more local scale and evaluated the ability of a hydrologic model to replicate observations from a permeable pavement installation. This study found that discharge and evaporation from permeable pavement is highly dependent on the specific pavement used and is not well simulated for most types of individual rainfall events with current modelling tools.