• Flooding is a major world-wide problem which seems to have increased in severity within the last decade. Many scientific studies have focused on the "major" recent floods and on methods for forecasting and managing floods at such large scales. In many respects the smaller-scale "localised" severe flood events, called flash floods, can be very damaging and often require much more complex forecasting techniques than the larger scale floods. The project focuses on floods in small, medium and urban catchments and will develop integrated techniques for understanding their meteorological genesis and forecasting their occurrence. The crucial role-played by rainfall accumulation will be examined in order to understand the flood forecasting uncertainty linked with the present state of the art rainfall estimation and forecasting techniques.

    The accuracy of extreme precipitation forecasts and consequent floods forecasts is of extreme importance in preventing loss of life, health and psychological damage and infrastructure and property loss or damage. Extreme rainfall events are also important in the generation of landslides and benefits in terms of risk assessment will follow from the increased understanding of the generating mechanisms and spatial distribution of heavy rainfalls.

  • The prediction and management of rainfall-induced flood events requires a complex strategy. Many different information sources are available, e.g. radar, satellite remote sensing, telemetering raingauges and river level gauges. In the common practice, forecasting tends to rely on only one (or at best only a few) of these sources of information. Recent research has demonstrated the importance of numerical weather prediction models, especially high resolution limited area models, in forecasting extreme precipitation at catchment scales. However one of the principal limitations on their use is the availability of real-time data of suitable scale and coverage to initialise the NWP model. In this project, we seek to identify how conventional raingauges, Doppler radar data, together with other remote sensing data, may be best linked to NWP models so that the prediction and the estimation of severe rainfall can be qualitatively and quantitatively improved.

    Hence the principal objective will be to assimilate in the currently available NWP methods information obtained from Doppler moments and reflectivity along with other remote sensing data. The scientific strategy would be to convert Doppler data into atmospheric parameters suitable for linking to NWP algorithms.

    Another innovative aspect of the proposed work is to exploit NWP results to improve the interpretation of radar measurements. This exercise, complementary to the previous one, not only enables a general improvement in extracting information from radar but also a reduction in inherent radar errors, thus allowing to quality control radar measurements.

    Finally, a major component will be the hydrological assessment of the benefits in terms of reduced uncertainty obtainable with the improved rainfall field, forecasted and estimated, coupled with an assessment of the sensitivity of hydrological models to different errors in their input. The joint use of all the information available will increase the overall quality of the actual flood forecasting.

    CARPE DIEM project will advance the present state of the art, mainly by increasing the co-operative use of high-resolution radar information and NWP. A key aspect can be identified in the assimilation of Doppler radar moments. The use of different numerical techniques increases the possibility of successfully attaining the project objectives.

  • CARPE DIEM's main goal is to improve the overall quality of flood and flash-flood forecasting. The improvements in severe storm precipitation deposit both estimated and forecasted, are of primary importance, also, for users community beyond that the "flood users". To demonstrate the possibility of exploiting the results in other areas we have involved, as part of the End-Users panel, the Finnish National Road Authority and CLABSA from Spain.

    Ancillary advantages are also expected in other fields.

    • A better identification of the rain phase, as well as a better snow forecast, is important in all Northern European countries for snow clearance procedure.
    • Better precipitation phase identification is crucial for forecasting winter floods, which have been severe in Central Europe during the 90s. In heavy winter precipitation the areal distribution of phase determines whether preventive actions should be started against accumulated snow.
    • An improved estimation of rain deposit in urban catchments has a direct feedback into the sewage management.
    • A better understanding of how the Doppler radar data could be used in operational suites of NWP models. In this context weather radar can be used in a secondary role to fill gaps in a network of wind profilers and autosondes.
    • The software tools developed to identify anomalous propagation echoes could be easily adapted and adopted for exploitation by a wide range of end users, particularly weather bureaux employing radar to quality-control radar data.