Flood risk dynamics and adaptation analyses for coastal cities based on internet big data and hydrology-hydrodynamic models

Abstract

Climate change, sea-level rise, land subsidence, and rapid urbanization are likely to increase flood risk in low-lying coastal cities in the future, posing serious challenges to urban sustainability. To quantitatively assess the future trends of flood risks and to carry out effective adaptation measures are among the hot topics for flood risk management. A sound answer needs fine-scale data to support an integrative analysis of hydrology-hydrodynamic processes, socio-economic impacts, and adaptation measures. In this regard, this paper puts forward a set of multidisciplinary and comprehensive methods to address the frontier scientific issues. First, internet big data and machine learning methods were adopted to map the building footprint, height, and its economic values, providing a fine-scale spatial and economic information of elements at risk. The residential building footprint was mapped with an accuracy of 92.8%, and the housing price was derived via web crawler from different online real estate websites. Second, the hydrology-hydrodynamic models of TOMAWAC, TELEMAC and MIKE 1D/2D were integrated to simulate four storm-surge flooding scenarios. The scenarios are based on the real Typhoon event TC9711 on August 18, 1997, whereby the wind, wave, tide, 1D flood, and 2D flood processes are simulated and flood scenarios of four different return periods are extrapolated. Especially, for future scenarios in 2050, sea level rise and land subsidence are also considered. Third, the expected annual damage (EAD) is estimated via risk analysis technique with combination of flood maps in the four return periods and the fine-scale exposure data. Finally, a cost-benefit analysis was done to evaluate the cost-effectiveness of two flood adaptation measures, i.e., dry flood-proofing and wet flood-proofing measures, for residential buildings. Taking Shanghai as an example, the above methods were applied to comprehensively assess the storm-caused flood risk of residential buildings, its future trends to 2050, and the costs and benefits of two adaptation measures. The results show that: (1) Sea-level rise and land subsidence will significantly exacerbate the flood risks in Shanghai. (2) Under the three Representative Concentration Pathways (RCP4.5, RCP8.5 and RCP8.5 High-end), the area of exposed residential buildings in Shanghai by 2050 may reach 31−49 km2 (4%–6% of total building area) for a 1/200-year storm flood event and 180–206 km2 (14%–16%) for a 1/5000-year event. (3) The EAD of residential buildings in 2017 is 440 million CNY (65.4 million USD), which can increase 1.6–2.0 times and reach to CNY 700 million–900 million (104.0 million–133.8 million USD) in 2050. (4) The implementation of both dry and wet flood-proofing measures could potentially reduce the EAD by 76%–79% and 88%–90%, respectively. Both measures are considerably cost-effective: the net present values of dry and wet flood-proofing measures are 2.84 billion–5.53 billion CNY (0.4 billion–0.8 billion USD) and 5.85 billion–9.94 billion CNY (0.9 billion–1.5 billion USD), and the benefit/cost ratios are 1.9–2.4 and 7.9–9.8 respectively. This indicates that the cost-effectiveness of dry flood-proofing measures is less than that of wet measure. Our findings can provide both scientific understanding and effective adaptation measures for mitigating the future flood risk in Shanghai. These multidisciplinary methods could also be applied to assess flood risk and adaptation strategies in other coastal cities.