The rural landscape is ever changing and some changes in land use and land management practices can affect the flood risk downstream. Many individuals, groups and organisations want to predict the impact of historical and proposed changes, but the spatial scale for patterns of change lies awkwardly somewhere between the small and large scales at which there is established science.

At the small scale, for example, there are equations for flow in porous media and open channels, and at the catchment scale and beyond there are lumped rainfall-runoff models which can be combined with national data sets such as HOST (Hydrology of Soil Types). This talk described new methods that use the established science but work at an intermediate scale.

The effects of change on runoff are predicted on a grid, and maps created that show how such effects accumulate and attenuate as they propagate downstream through the river network. The methods have been applied to the Upper Severn and to an extreme event in the Upper Hodder (117mm rainfall in 90 minutes, measured in August 1967).

Exploring extreme events using joint probability analyses - Dr Cecilia Svensson (767KB)

Quantifying flood risk using ensemble predictions and a new digitised rainfall archive - Dr Patrick McSharry, University of Oxford

Flood events in the UK are usually associated with extreme rainfall and can last from minutes to weeks. Efficient management and mitigation of flood risk, especially surface water flooding in urban areas, requires accurate and reliable precipitation forecasts as inputs to flood risk models. Houses in flat areas are particularly at risk, and meeting the shortage of houses in the south-east of the UK requires building on these areas.

We constructed an archive of British Rainfall data from 1860 to the present day to identify the rainfall mechanisms underlying extreme rainfall events. The connection between record rainfall and flooding is highly nonlinear, implying that probabilistic forecasts of rainfall are required. The main outcomes of the project are:

digitisation of a long sought-after historical rainfall data source;
demonstration of the benefits of producing probabilistic forecasts by combining statistical techniques with the outputs of ensemble predictions from a NWP model; and
an objective method for classifying extreme UK rainfall patterns based on the new historical rainfall archive.

Further details including four peer-reviewed publications are available from the Smith School of Enterprise & the Environment website.

Uncertainty Assessments of Flood Inundation Impacts - Dr Fredrik Wetterhall, SMHI Sweden

This study set up novel techniques for tracking uncertainties through a modelling framework of flood inundation under a future climate change. The study was carried out over a catchment in the river Severn in bordering between Wales and England.

The modelling framework consists of statistically and dynamically downscaled meteorological input from an ensemble of GCMs and RCMs. The climate input is further driving a set of rainfall-runoff models, in this case LISFLOOD-RR and HBV. The hydrological models provide modelled discharges which are fed through two flood inundation models, LISFLOOD-FP and HECRAS.

The uncertainties are cascaded through the modelling chain and it is important to rigorously estimate this uncertainty at all levels. The overall aim of this study was to incorporate all these uncertainties at the very end of the chain in a flood risk map.

Overview and main outcomes on the COST 731 Action-Propagation of uncertainty in advanced meteo-hydrological forecast systems - Dr Pier Paolo Alberoni

The COST 731 Action, of the latest in a series of COST Actions related to radar meteorology, can be seen as the expression of the will of a large number of European meteorological and hydrological services to further both understanding and, even more so, application of systematic uncertainty information.

The Action hence focuses on hydro-meteorological forecasting and how to deal with the uncertainties inherent in the entire forecast chain. The COST 731 Action was proposed within the ESSEM Domain and launched mid 2005 for a five-year period as an offspring of a series of COST Actions related to radar meteorology.

While the COST Actions 72, 73, and 75 dealt with pure scientific issues related to single weather radars, radar networking, and advanced capabilities (Meischner et al., 1997), Action 717 focused on the application of radars in hydrological and NWP models.

Efforts like the COST 731 Action suggest that a systematic treatment of uncertainty is a prerequisite is information produced by advance hydrometeorological forecast systems are to be used in operational contexts by practitioners and decision makers.

The main outcome may be headlined as follows. COST 731: acted as a 'door-opener' for dealing with uncertainty in a systematic way, especially in operational settings; witnessed the development and evaluation of several integrated, end-to-end forecasts systems; provided information and hands-on training on visualisation platforms of probabilistic information based on huge amounts of observational and forecast data; learnt to express uncertainty in directly usable forms, ie QPE ensembles; promoted convection-permitting NWP and convective-scale radar data assimilation.

The Occurrence of Rainfall in a Changing Climate - Dr Tim Osborne (2·8MB)

Linking meteorological, ocean and coastal models using ensemble methods - Dr Qingping Zou

Within the UK, assets worth an estimated £132·2 billion are at risk from flooding by the sea and £7·8 billion more from coastal erosion. In contrast, assets at risk on river flood plains are valued at £81·7 billion.

In the UK coastal flood defences are usually designed to withstand extreme events with a return period of between 50 to 200 years, taking account of sea level rise. Currently, there is a lack of a robust and integrated 'cloud-to-coast' framework for assessing coastal flood risk. The interactions between the atmosphere, oceans and coasts are poorly understood. There are large uncertainties in the performance of sea defences and predictions of coastal flood risk in extreme conditions.

This talk presented an integrated 'cloud-to-coast' ensemble modelling framework of coastal flood risk due to wave overtopping and toe scour at sea defences. It consists of three key components: meteorological forecasts, tide-surge-wave predictions and surf zone models. Using this integrated modelling approach, the flood risk in coastal areas from the extreme events can be accurately predicted and the associated uncertainties are assessed by creating ensembles of possible future storm events.

FREE

Resources

Project reporting

Presentations from FREE meeting at Imperial College October 2010

About the programme