The Rotterdam experiments


The overall objective of the Rotterdam experiments is to conduct an intensive monitoring campaign in the city centre of Rotterdam. The goal is to bring together meteorologists, hydrologists, urban drainage engineers and end-users to bridge the gap between research and practitioners and address some of the current challenges related to urban flood forecasting. Some research questions to be addressed include:

  • How accurately can we measure rainfall in urban areas?
  • What spatial and temporal resolutions are necessary to adequately capture and predict rainfall patterns responsible for urban flooding?
  • How accurately can we predict the location and timing of heavy rainfall based on state-of-the art numerical weather prediction and radar nowcasting techniques?
  • How can we integrate rainfall information into the decision making process for city managers and emergency management services – before, during and after the flood?


Rotterdam is Netherlands' second largest city and most important link to the North sea. It is situated in an area of peatland along the river Maas and largely lies below sea level. The city is characterized by an extensive network of drainage canals and man-made ponds. In most places, groundwater is found at around 70 cm below the surface. Because the weak peat soil compresses easily, the surface level of the city sinks by about 1 to 2 cm every year.  Water supply and drainage for the city consists of surface waters, a sewerage system, pumping stations, pressure pipelines, overflows and several sewage treatment plants. Most of the rainfall drains through the sewer along with domestic wastewater to the sewage treatment plants. From there, pumps send the treated water into the river Maas. If the sewer capacity is exceeded during heavy rain, the untreated sewage enters the surface waters through a series of overflow outfalls and from there, flows into the river Maas. Because Rotterdam has little surface water, options for storing excess stormwater are limited. The city depends on the regional water authorities of Schieland, Delfland and Hollandse Delta for protection against floods. During prolonged dry periods, the city also depends on fresh water sources from outside. Like many other cities in the Netherlands, Rotterdam must adapt its water management in the face of new hydro-climatic risks. Sea levels are rising and more water flows through the river than before. Summer rain bursts are becoming shorter and more intense, increasing the risks of urban flooding. The annual precipitation over Rotterdam is about 36 million m3 of water, increasing by about 1% per year. In the long term, the city wants to expand its excess water storage facilities by using new surface waters, underground storage facilities, green roofs and water plazas.

Experiment design
Accurate rainfall observations in cities are of critical importance for reliable prediction of water levels and flows, overflow from sewers and predicting urban flood events. In recent years, a high resolution X-band rainfall radar has been installed in the centre of Rotterdam, as part of the RainGain project ( The radar has dual-polarimetry and Doppler capability and can measure just above the urban canopy, at approximately 150 m altitude with a resolution of 100 meters and 1 minute and a range of about 30-40 km. The radar offers unique opportunities to study the structure and dynamics of rainfall in an urban setting, to analyse micro-climate effects and turbulence induced by the urban fabric. In addition, the radar provides unique insight into the link between small-scale rainfall variability and urban hydrological response. Table 1 summarizes the different instruments and sensors available in the area. In addition to the radar, there is also additional independent meteorological information provided by a dense network of 10 professional weather stations on the ground, several disdrometers, citizen rain gauge networks, GNSS water vapour retrievals and vertical profiling radar. Hydrological data will be available at 40 flow sensors and about 60 water level sensors at pumping stations and in sewers throughout the city.


Table 1. Instruments deployed during the summer experiment

Instrument Measured variables Resolution
Weather stations Rainfall, wind, temperature, humidity, radiation 5 min, 10 stations
Polarimetric X-band radar Reflectivity, differential reflectivity, phase shift, Doppler spectrum, rainfall 1 min, 30-100 m
MRR vertical profiling radar Vertical Doppler spectra, Drop Size Distributions 1 min, 30-100 m
Disdrometers Drop size distribution, rainfall estimates 5 min, 6 disdrometers
Flow sensors Flow in outgoing main of sewer pumping stations 1 min, 40 flow sensors
Water level sensors Water levels in chambers at pumping stations 1 min, 40 water level sensors
Water level sensors Water level at sewer overflow weirs 1 min, 21 water level sensors
GNSS sensors for water vapour retrieval GNSS water vapour estimates from signal delays 24h, 9 stations
Citizen weather stations (weather underground) Precipitation, temperature, wind, humidity 10 min, variable density, up to 4 stations per km2

Numerical Modelling

During the experimental campaigns numerical simulations of atmospheric variables including precipitation, air temperature, mixing ratio and wind speed and direction from the WRF-ARW (Weather Research and Forecasting – Advanced Research WRF) mesoscale meteorological model will be evaluated. WRF-ARW is a non-hydrostatic mesoscale numerical weather prediction model developed at the National Center for Atmospheric Research. WRF-ARW offers the choice between a large number of microphysics and land surface parameterization schemes. Every microphysics scheme uses a unique method of parameterizing the atmospheric heat and moisture tendencies and microphysical rates, while each land surface scheme parameterizes land surface characteristics such as snow, soil, and vegetation. The WRF simulations of precipitation will be validated against rainfall observations from the Rotterdam X-band radar, micro-rain radar and measurements from in-situ meteorological sensors. One of the objectives is to assess the sensitivity of simulated rainfall to different microphysics and land surface parameterization schemes. In parallel, other objectives include an evaluation of the impact of data assimilation from the X-band radar, and the potential use of WRF model ensemble runs for probabilistic rainfall forecasting.

Expected Outcomes

The Rotterdam experiments will contribute towards the general objectives of MUFFIN with new rainfall measurement and numerical prediction methods and help improve the current early warning and flood forecasting methods through a better understanding of how atmospheric and hydrological processes interact with the urban fabric. Special emphasis is given to the reliability and usefulness of rainfall forecasts for end-users depending on the considered application and on the efficient coupling of rainfall and urban flood models. The following list summarizes the main expected outcomes of the experiments:

  • High resolution rainfall maps and space-time variability of rainfall in an urban environment.
  • Information about urban modification of rainfall patterns and differences between rainfall above the urban canopy and at ground level.
  • Short-term ensemble rainfall forecasts for lead times up to a couple of hours based on state-of-the-art combination of stochastic radar nowcasting techniques and high-resolution numerical weather prediction models.
  • Variability in hydrological response patterns based on flows and water level observations (runoff ratios, peak flows, response times).
  • Hydrological/hydrodynamic response modelling: validation of forecasted water levels and flows, accuracy and precision of hydrological response characteristics, contributions of error sources to overall uncertainty.
  • Post-processing and flood assessment: tailored information for improved visualization, awareness and communication depending on the requirement of different end-users.

End-users and Partners

Other Links
Urban Green-Blue Grids:
Rotterdam Climate Initiative:
Rain Gain:
Geoscience & Remote Sensing:
Water Management: