Outputs

PhD Students

Students, who will be the next generation of urban flood managers, have played an important role in our project. The students have academically benefitted from involvement in the project, having been introduced to new concepts.

  • Luis Duque-Yaguache (Newcastle University) developed our case study, that demonstrates how probabilistic information can be used in a real-time flood forecasting system
  • Mingming Song (NHRI) is developing a model of our study site using the SWMM model
  • Qian Li (Loughborough University) performed the hydraulic simulations for the study site

Publications

L. Duque (submitted). A Monte Carlo Simulation Study of the Factors Influencing the Performance of Flood Early Warning Systems. PhD thesis, Newcastle University, UK.

Li, Q., Liang, Q. and Xia, X., 2020. A novel 1D-2D coupled model for hydrodynamic simulation of flows in drainage networksAdvances in Water Resources137, p.103519.

Virtual Flood Game

We have developed a game, Virtual FloodPlease play our game and see if the use of probabilistic information leads to better decisions. This provides a good basis for understanding the use of probabilistic information in real-time flood forecasting.

Data Products

Flood Maps

A number of flood maps have been generated in this project for our study site, including the July 2016 event, which caused widespread inundation, and several synthetic storms.  Details of the hydraulic simulations are provided here. Each map provides the maximum simulated flood depth (millimeters) recorded during a hydraulic simulation. These maps may be downloaded below:

  1. July 2016 event 
  2. The 100-year rainfall events
  3. The 50-year rainfall events
  4. The 25-year rainfall events

The rainfall used for the return period events can be downloaded here [Excel].

How to Use: The data are in ESRI ASCII Grid format, which can be read using most GIS software.

Limitations: The sources of uncertainty are discussed in the Hydraulic Modelling section.

Survey of Inner Rivers

A field survey of the inner river widths within the polder was performed to improve the topographic representation in the hydraulic model simulations.

Survey Inner RiversSurvey Inner RiversSurvey Inner Rivers is available for download.

Rainfall Data

One-hundred years of synthetic 1-hour rainfall data have been generated for the Dongshan rain gauge, using RAINSIM. These data may be used in continuous simulation to assess sewer performance and flood inundation. The hourly data have also been disaggregated to 5-minute values. The data can be downloaded here:

  1. Hourly data
  2. 5-minute data

How to Use: The data are in a single column and the values are accumulated totals over the time interval (i.e. hourly or 5 minutes). The first line is for January 1st, 2021. (Note, the data are for current climatic conditions, a start date is specified to ensure the sequence of leap years is correct.)

Limitations: The RAINSIM model is fitted to reproduce the daily and hourly statistics of observed rainfall records. We had access to a long observed daily record (65 years), so we have good observed daily statistics for model fitting. Ideally, a 20+ year record would also be used to estimate the sub-daily statistics, however, only 5-years of data were available. Limited availability of sub-daily data is a common problem. An analysis of the selected storms did not identify any major issues with using this short record for fitting.