Half a degree warming might cause doubled economic loss and intensified affected population of flood in China

Based on future scenario data and an improved quantitative assessment model of natural disaster risk, in this paper, we analyze the response of the characteristics of flooding events in China to 1.5°C and 2°C of global warming, quantitatively 10 assess social and economic risks of the floods, and determine the integrated risk levels. The results indicate that for the RCP4.5 and RCP8.5 scenarios, the hazard and distribution area of the floods increase with increasing temperature and the influence range of floods in different levels expands more rapidly under RCP4.5 scenario. The floods mainly affect the social economy in the regions with lower altitudes and smaller slopes in eastern China. With intensification of temperature rise, the affected population and the direct economic losses would be aggravated. For 2°C of global warming, under RCP8.5 scenario, affected 15 population by floods would increase by 2 million and the economic risk would nearly double compared with 1.5°C of global warming. The economic risk under RCP4.5 scenario would even reach three times that for 1.5°C of global warming, but its proportion to GDP is lower than that of RCP8.5 scenario. Under both scenarios, the ranges of the medium-high flood risk zones would gradually expand westward and northward.


Introduction 20
In recent decades, changes in climate have caused impacts on natural and human systems on all continents and across the oceans (IPCC, 2012;IPCC, 2014). At the global and regional scales, climate-related hazards have caused enormous damage (Kundzewicz et al., 2014;Johnson et al., 2016;Luo et al., 2018;Paprotny et al., 2018), resulting in a significant increase in economic losses (Parmesan and Yohe, 2003;Patz et al., 2005). In 1998-2017, countries influenced by climate-related disasters reported direct economic losses of $2.245 trillion USD, of which floods were the most frequent type of disaster, making up 25 43% of all recorded events (Swiss Re Group, 2017;UNISDR and CRED, 2018). Over the last 20 years, China has suffered the second largest loss due to the frequent occurrence and serious impacts of climate-related hazards (Ding et al., 2006;Huang et al., 2007;UNISDR and CRED, 2018). Among all of the meteorological and hydrological disasters, floods are characterized by sudden and frequent occurrence, and the direct economic losses caused by these events account for the largest proportion of the total economic losses of climate-related hazards (Writing Committee for Third National Assessment Report on Climate R D E P =    Bates et al. (2008) pointed out that floods are affected by a variety of climatic and non-climatic processes. Precipitation is the most important climatic process. Flood disasters in China are mostly caused by intense precipitation, and the intensity and losses of the floods are closely related to the corresponding precipitation. Therefore, the amount of precipitation can be used to classify the intensity of floods (Zhai et al., 1999;Gong and Wang, 2000;Zhai et al., 2005;Ma et al., 2018). From the perspective of historical flood disasters, the maximum accumulated 3-day precipitation has an important impact on the 95 occurrence of floods during the continuous precipitation process. Generally, mild, moderate, and severe floods correspond to maximum accumulated 3-day precipitation values of 30 (35) -150 mm, 150 -250 mm and ≥ 250 mm, respectively (Li et al., 2012). Among non-climatic processes, altitude and slope have the greatest impact on the formation and development of floods, so they can be used as environmental correction parameters for the underlying surface (Zhou et al., 2000;Thompson and Clayton, 2002;Li et al., 2012). 100 Based on the above theories, the probability of floods can be expressed as: ( 2) where P is the probability or hazard of flood, F indicates the possibility of intense precipitation, and I is the environmental correction parameter for the underlying surface.
Correspondingly, the quantitative assessment model of the social and economic risks of pluvial floods can be expressed as 105 follows: (3) where R is the social and economic risk of the floods, D is the destructive power of different magnitudes of floods, and E is the exposure of the risk bearing bodies. Therefore, the steps of the social and economic risk assessment of pluvial floods in China are as follows: 110 The maximum accumulated 3-day precipitation values of 30 (35) -150 mm, 150 -250 mm and ≥250 mm were taken as the maximum occurrence of mild, moderate, and severe floods. These values were converted into a probability (probability = frequency/time period × 100%, set to 1 if the probability is greater than 100%). Furthermore, considering the heterogeneity of the regional geographical environment, the environmental correction parameter of the underlying surface, which is based on the altitude and slope, was used to correct this probability. Thus, the hazard of flooding was obtained; 115 Flood disasters can have serious impacts on the economy and society. GDP and population were selected as the exposure of the economy and society. For the destructive power of floods, a quantitative relationship was established between the different magnitudes of the floods and the loss rates of GDP and population according to the classification criterion of the maximum accumulated 3-day precipitation based on statistical data for 1001 flood disasters from 1990 to 2008 in China, which was obtained from Li et al. (2012); 120 The exposures of the economy and society were obtained from SSPs data given in the data sources section; The social and economic risks of floods were quantitatively evaluated based on the hazards and destructive power of floods and the exposures of risk bearing bodies; Finally, the integrated risk of floods was obtained using the superposition analysis method and was graded using the multiple of the standard deviation method. 125    (Fig. S2c, d). The changes of mild hazard index would be mainly 155 concentrated in the Northwest China, with an increase in the eastern and a decrease in the western under RCP4.5 and RCP8.5 scenarios (Fig. S2e, f).

Flood risks for the population and economy
Floods have serious impacts and cause damage to the social economy, food production, natural ecosystem, and infrastructure.
In this study, the population and economy were selected as the risk bearing bodies to quantitatively assess the risks of floods 160 in the future. China Plain, the Northeast Plain, the Middle-Lower Yangtze Plain, the Sichuan Basin, and the Pearl River Basin. The distribution of the population and economic risks under RCP4.5 scenario would be basically consistent with RCP8.5 scenario, but the scope would shift westward. The affected population and the direct economic loss in the southern part of the Yangtze River region would be greater than that for RCP8.5 scenario ( Fig. 3; Fig. 4a, c, e; Fig. 5; Fig. 6a, c, e; Fig. S3a, c, e; Fig. S5a, c, e). 175 increase. Moderate floods would affect nearly 29 million people, the direct economic loss would be close to $78 billion USD, and the impact range would expand northward and westward. The population affected by mild floods would increase by more than 2 million, and the direct economic loss would nearly double, but the impact range would be basically the same as that for 1.5°C of global warming. Under RCP4.5 scenario, the population affected by floods of different levels would be lower, and the 185 direct economic loss would be significantly higher. However, the proportion of GDP would be less than that for RCP8.5 scenario. The impact ranges of the two scenarios would be basically the same (Fig. 3; Fig. 4b, d, f; Fig. 5; Fig. 6b, d, f; Fig.   S3b, d, f; Fig. S5b, d, f). in Central China and South China (Fig. S4a, b). The changes in population risks of moderate floods would be mainly distributed in eastern of the Heihe-Tengchong Line. The distribution patterns under both scenarios would be similar, and the areas of increased risk would be mainly concentrated in the North China Plain and the Yangtze River Basin (Fig. S4c, d). The change in population risks of mild floods to the eastern of the Heihe-Tengchong line would be small, while the change in the western of the Heihe-Tengchong line would be large, with the most significant increases mainly in northern North China,Northwest 200 China and the Sanjiangyuan (Fig. S4e, f). would increase by more than 500% (Fig. S6a, b). Under RCP8.5 scenario, the economic risks of moderate floods would decrease on both sides of the Heihe-Tengchong Line and would nearly double on the eastern side of the Line, with the risk increasing 210 more than fivefold in the Northeast Plain and the Qinling Mountains. Under RCP4.5 scenario, the risks would decrease on both sides of the Heihe-Tengchong Line and eastern Northeast China. The risks in most areas on the eastern of the Line would increase by about 2 times, and the risks in the southern Southwest China would increase by more than 10 times (Fig. S6c, d).
Under RCP8.5 scenario, the economic risks of mild floods would increase within 100% in most parts of the country, and the risk increase in the northern Northwest China would be more than twice. Under RCP4.5 scenario, the risks in most parts of the 215 country would increase by more than 200%, and the risk in the northern Northwest China and the Sanjiangyuan would increase by more than 10 times.

Integrated risks of floods
Based on the above quantitative assessment of the population and economic risks of floods, the standard deviation method was used to assess the integrated risks of floods. 220  For 2°C of global warming, under RCP8.5 scenario, the distribution areas of the high, medium, and low risk zones would increase to different extents. The expansion of the medium risk zone would be most significant, and mainly affects the eastern part of Northwest China and the central and western parts of Southwest China. The medium-high risk zones would account for more than 48% of the national total land area in China. Under RCP4.5 scenario, the distribution areas of the high risk and 235 medium risk zones would increase slightly, mainly in North China and the eastern part of Northwest China, while the distribution area of the low risk zones would decrease slightly. The medium-high risk zones would account for more than half of the national total land area in China (Fig. 7b, d).

Discussion
In the future, with the continuous intensification of global warming, the impact and distribution area of severe and moderate 240 flood events with more serious impacts would continue to expand, and the social and economic risks of floods would become more serious. In terms of climate scenarios, the distribution area of floods under RCP4.5 scenario would be broader; however, under RCP8.5 scenario, the floods would affect more people, the direct economic losses would account for a larger proportion