In WP 4 multi scale flood forecasting experiments are set up, tested and compared across different case study catch- ments, spatial and temporal scales, and with different input and forcing data, catchment characteristics, etc. The main objective is to evaluate high-resolution multi-basin flood forecasting against local fine scale catchment flood forecasting through joint case studies. By comparing accuracy and uncertainty in critical runoff/discharge, predicted flood prone areas, potential lead-time, etc. the two model scales are evaluated with regards to the resolution required of inputs and outputs and their potential in real-time urban flood forecasting. Key issues include the following:

  • Rainfall forcing. What is the impact of (temporal and spatial) resolution in rainfall observations and forecasts? How can different types of observational sensors be used and combined? What is the relation to weather situation?
  • Hydrological simulation. What is the impact of land-use information? What is the value of high-resolution multi- basin hydrological forecasting in an urban context? Can large-scale information support local flood models?
  • End-user communication. Which information is relevant for which end-user? How can we increase the relevance and ”sharpness” of the information? How should ensemble forecasts be used and presented for optimum value?

Table 1 below shows an overview of the joint forecasting experiments, scheduled to be performed during 2018.


Table 1 Characterization of the study basins and the joint experiments





Large scale catchment (HYPE setup)

Delfland (~400 km2)

Catchment type: Rural/urban

Haaganpuro (11km2)

Catchment type: Urban

Østerå catchment (150 km2)

Catchment type: Rural/urban


Local scale catchment

Rotterdam (~22 km2)

Model: 3Di

Catchment type: Urban

Länsi-Pakila (1 km2)

Model: SWMM Catchment type: Urban

Kærby (1 km2)

Model: MIKE Catchment type: Urban

Hydrological forcing data for calibration and validation

Water level sensors

Flow and pump data from drainage system

Flow measurements at both catch- ments

Water level and GRW gauges CSO registrations

Pump data



Rainfall observations

Rain gauges

KNMI C-band radar data TUD X-band

Stochastic rainfall simulator

Rain gauges

University of Helsinki composite radar data

FMI radar data

SVK Rain gauge network

Local rain gauges + disdrometer DMI C-band radar

AAU X-band radars



Rainfall forecasts

KNMI C-band nowcasts HARMONIE forecasts WRF-forecast

Aalto radar nowcast FMI HIRLAM HARMONIE

AAU ensemble nowcast on X/C-band radar








Special focus experiments

·    Space/resolutions of models and inputs

·    Evaluation of accuracy between different rainfall products and fore- cast products

·    Display of flood risk maps

·    Detailed investigation of rainfall / flood response using artificial rain- fall fields

·    Predictive skills of large scale model local catchment

·    Impacts of land-use in flooding

·    Impacts of temporal resolution on small catchment simulations.

·    Space/resolutions of models and inputs

·    Comparisons of return periods of rainfall and flood response

·    Predictive skills of large scale model local catchment

·    Use of HYPE model as boundary condition for local model

·    Development of fast and distribut- ed flood model based on pre- simulated multiple system states