urban areas and conglomerates are often subject to flooding;
this has several reasons closely related to urbanization that contribute to this problem:
- dramatic increase in sealed surfaces in the urban and per-urban areas, that
strongly affect runoff (and infiltration) processes;
- changes in the land use of urban watersheds, with both inreases in sealed surfaces but also
land use change from forest to other forms that are less capable of water retrention;
- damages increase with increasing values of properties and infrastructure affected,
but also the economic costs of temporary interruptions , primarily of the tranbsportation system;
- the potential increase in the frequency and intensity of precipitation events
due to climate change, increased temperature and thus a faster hydrological cycle.
Early warning using model-based medium and short-term forecasts (numerical weather forecasts,
data assimilation from monitoring and weather radar) can help to control the effects,
(direct control of SCADA systems) mitigate damage, but also contribute to a better planning
and design of urban land use and drainagae systems.
The modeling of urban floods is based on a veryu high resolution digital elevation model (DEM)
at 30 m horizontal resolution of better (LIDAR based), as well as data on the drainage system
(open channels, canals, sewers), using the
EPA Storm Water management Model model system (SWMM) for the drainage network.
Hourly precipitation (starting with forecasts of several daya; if a storm is predicted,
the forecasting frequency can be increased from 20 to 12 and ulimately 6 hours,
with hourly updates of the initial conditions for nowcasts based on monitoring data assimilation.
The implementation can combine a semi-distributed (multiple coupled catchments)
and a fully distributed (high resolution raster) that can be run at high temporal resolutions
(hourly forecasts to several minutes for dynamic nowcasts).
The model supports the full range of strategic planning, early warning,
and operational control (including the possibility for real-time optimization).
The model also support mobile clients for first responders.