A tale of two cities: flood protection for Shanghai using the Thames Estuary 2100 approach

When it comes to flooding, the financial centres of Shanghai and London have a lot in common. Both cities are located on tidal estuaries, at risk from tidal inundation and vulnerable to sea level rise, so lessons we learned in London are perfectly suited to help protect China’s second city.

London

The flood defence system for London is dominated by the Thames Barrier, a major flood control structure across the river that prevents high tidal water levels passing upriver and flooding the city. The Thames Estuary also has flood defences along its entire length, with those downriver of the Thames Barrier, about two metres higher than the defences upriver of the barrier.

In 2012, the Environment Agency published TE2100, a flood management plan for the Thames estuary to the year 2100. Developed to ensure that London and the Thames Estuary are protected from future flooding as sea levels rise, the plan includes a sequence of interventions to improve the flood defence system. 

At HR Wallingford, we spent two years developing the approach for TE2100, and our technical report of 2009 went on to form the basis of the 2012 TE2100 plan. The methodology used to prepare the plan includes consideration of a range of future climate change scenarios and the development of options that can be adapted as the sea level rises. These options follow adaptation pathways, where interventions are required as particular sea level rise thresholds are reached.
 

Shanghai

Tidal inundation during typhoons is the main flood risk to the Huangpu River in Shanghai, although the city is also at prone to flooding caused by typhoon rainfall. The city is protected from tidal flooding by defences along the river, which are designed to ‘1 in 1000 years’, the same design standard of protection as London. However, sea level rise and ground subsidence has led to the need for improved defences and the city is proposing to construct a flood control barrier downriver of Shanghai near the river mouth.

HR Wallingford was asked to research and model options for the project, leading the technical work, while all of the modelling and analysis was carried out in China, by HR Wallingford and its partners. Our research took the TE2100 approach and applied the method to Shanghai. The project was funded by a Newton Fund research grant to support collaborative research between the UK and other countries, in this case between our UK and Shanghai offices, the Shanghai Climate Change Research Centre and SOAS (University of London).

Over the course of 2018 and 2019, we met the with the stakeholders both in China and the UK to discuss the project and to collect data. These data included information on the sources of flooding (high tides; river flows and heavy rainfall); survey data for the Huangpu River; the flood protection system for Shanghai; information on the flood risk area including the topography and land use; and flood damage data.

The present day and future flood scenarios we developed for Shanghai included: tidal water levels; river flows; rainfall; and drainage inflows to the Huangpu River. The team used sea level rises of up to 3 metres in order to consider the flooding that could occur in the very long term (more than 100 years), or in the case of a rapid rise in sea level that exceeds current projections.

Our team then developed two models using InfoWorks ICM software. The first was a 1-dimensional (1D) model of the Huangpu River and the second took the 1D model and combined it with a 2D model of the entire floodplain area, including the city of Shanghai and the surrounding areas, to create a 1D/2D model. The 1D model was used to predict water levels in the Huangpu River with no overtopping of defences in order to determine the flood defence levels that would be needed to prevent flooding and model flood protection options. The 1D/2D model was used to assess flooding and flood damages.

The team then applied the1D/2D model to a range of scenarios including combinations of high tides, river inflows and rainfall in order to determine present and future flooding and flood damages. Whilst the estimates are approximate, the results clearly show that a very high level of flood damage could occur in Shanghai.

The 1D model was used to determine flood water levels along the river for present and future scenarios of tidal water levels combined with river flows and drainage inflows. It soon became apparent that the present-day flood defence crest levels are below the required levels at some locations.

Options for Shanghai

Shanghai, like London, is highly constrained by development on the floodplains, meaning that major engineering works, such as raising of flood defences and a flood control barrier, are the best options to protect the city from flooding. We established design criteria for the flood protection system including standards of protection, and used modelling to develop adaptation pathways for flood protection options, similar to those developed for London.  

The options investigated included raising all the flood defences and developing a strategy for raising the defences in stages. The average increase in defence crest levels needed to provide protection to the year 2080 is about 1.1 m, with further defence raising needed after this date, although this would result in high river walls on both sides of the river.

The city of Shanghai’s preferred option is a new flood control barrier, so we also explored this option. There are a set of key parameters must be established as part of the design process. These included: the defence level provided by the structure; the high tide level at which the barrier must be closed;  the timing of a barrier closure; management of upriver fluvial and drainage inflows during a barrier closure; and the number of closures per year.

The number of times a barrier closes in a year is important because it affects the risk of failure. When barriers are closed frequently, the time available for maintenance is reduced and the annual probability of failure during a closure increases. The number of closures per year can be controlled by raising the upriver defences, so we developed an adaptation pathway for the potential barrier that included raising of the upriver defences to reduce both the number of closures and to take account of water level increases caused by potential future drainage inflows. 

When the sea level rises, barriers are closed more frequently and the impacts on navigation become more serious. For this reason we also investigated a barrier with locks and a tide excluding barrage. In addition, the team considered options for reducing fluvial inflows on the Huangpu River and managing urban flooding from rainfall.

After exploring a number of options, we concluded that raising of the initial defence followed by the construction of a barrier and the raising of associated defences appears to be the most suitable option for Shanghai. Interestingly, this approach was also identified as the best option for London and the Thames Estuary, where the existing barrier will be replaced or improved in about 2070 based on current sea level rise projections. So, it seems that the two cities will continue to tell similar tales about flooding and flood control well into the future.

References

Environment Agency (2012), Managing flood risk through London and the Thames estuary: The TE2100 Plan, Environment Agency, November 2012.

HR Wallingford (2009), The Estuary Wide Options, Thames Estuary 2100, Technical Report - Appendix D, HR Wallingford on behalf of the Environment Agency, November 2009.