WW2014_Final-Report_MUFFIN_v3.3.pdf

More and more people live in cities and by 2050 two thirds of the global population is expected to live in a city. Because of this urban migration, cities generally need to densify and expand. The densification often leads to more impermeable surface, such as asphalt and concrete, which increases the risk of flash floods in city centres. The expansion may require that historically flood-prone areas need to be exploited, which increases the flood risk also in suburbs and peripheral parts. In parallel with this trend, climate change is expected to intensify the hydrological cycle and for example lead to heavier rainfall. When combined, these changes are likely to significantly increase the risk of flooding in urban and peri-urban areas. There is an urgent need for research towards better understanding, describing and forecasting urban flood risk under these changes. Much of the R&D in this context is focused on how to design (and re-design) cities to become more “climate proof”, in the sense better prepared to handle the expected future climate, e.g. by so-called “nature-based solutions” like green roofs and open dams for stormwater reduction and delay. This is a long-term planning perspective; to gradually modify the cities. There is also a short-term awareness perspective of adapting to the changes, which has been the focus of the MUFFIN project: the need for better tools for real-time urban flood forecasting and decision support (Early Warning Systems). The earlier we can predict a flood risk situation, the higher is the chance of reducing the negative impacts on society. In essence, an Early Warning System for urban flooding consists of three components: 1. Rainfall data. We need both recent rainfall observations, to know how much water is already in the ground, in pipe systems, etc., and rainfall forecasts, that predict the rainfall over the coming hours and days. 2. Flood model. We need a model that translates the rainfall hitting the ground to water flows, both on the surface and underground, and potential flooding, either by the flow exceeding the local drainage capacity or by river overtopping when combined with upstream inflows. 3. Communication tool. We need a tool that presents the rainfall data and the output from the flood model to relevant users. The tool must present these things in a way that is meaningful for the user, it must be fast and it must be easy to access and understand. In MUFFIN, we have worked on all three components. Concerning rainfall data, we have focused especially on rainfall observations by weather radar. We have shown that radar observations of very heavy rainfall are uncertain and that radars with a higher level of detail are preferable. We have explored using citizens’ weather stations and flood reports for improved forecasts. Concerning flood model, we have developed and tested a coupled multi-scale system, essentially with one model describing the upstream inflow (large scale) and another model describing the flooding in the city centre (small scale). We have shown that forecasting with a 1-h time step is useful in an urban context, and a national 1-h model is now operational in Sweden. We have explored the possibility to use open data and open software for innovative urban flood forecasting. Concerning communication tool, we have developed a prototype for radar-based rainfall visualization with several new features, e.g. “seamless” integration of observations and forecasts. Furthermore, the multi-scale model has a graphical interface with continuously updated forecasts to support the user. All work in MUFFIN has aimed at fitting as closely as possible to end-users’ requirements. An international end-user workshop was held in the beginning of the project, and the subsequent development and experiments were designed on the basis of the outcome. Through this approach, we hope that the results and products from MUFFIN will have a real impact. The results and products were disseminated at the successful final conference named CITIES, RAIN and RISK, held in Malmö, Sweden, 13-14/6 2019.
Project web site: http://www.muffin-project.eu/