Flooding is becoming an increasingly common problem across Europe, where floods from rivers, estuaries and the sea threaten millions of people as well as causing significant economic impacts. The magnitude and nature of flood hydrographs is strongly influenced by the spatial variation and accuracy of rainfall and varies as a function of catchment scale and type, rainfall properties and catchment antecedent conditions.
In the content of flood management, my PhD dissertation investigated the relationship between spatial rainfall and runoff production, for a range of scales and catchment types aiming to an effective design of rain gauge and radar networks. The study was based on 12 years of hourly rainfall and runoff data, 17 rain gauges (1x1 km2 spatial resolution), and 14 flow stations from the 1040 Km2 Upper Lee catchment, UK.
Continuous-time, semi-distributed rainfall-runoff modelling was considered due to the potential to represent the effects of spatially variable inputs such as rainfall and also to simulate the changing antecedent moisture conditions that determine the storm runoff response for a given rainfall event.
The project objectives were:
- Evaluate the potential of spatial-temporal rainfall properties (with emphasis on spatial rainfall variability) to explain hydrological catchment response
- Develop a calibration strategy for parsimonious, conceptual, semi-distributed hydrological models capable of representing the catchment heterogeneity and predicting runoff at ungauged sites
- Investigate the significance of spatial variability of rainfall on flood runoff generation as a function of catchment scale and type, spatial-temporal rainfall properties, and antecedent catchment conditions
- Investigate the different sources of uncertainty (lack of model structure and parameter identifiability, data uncertainty) and their propagation to the results
- Provide guidance on the effective design of rain gauge and radar networks and spatial resolution of rainfall for hydrological applications given the catchment scale and type