Study and Application of Flood Control Risk Trend Analysis Model

: In order to analyze the comprehensive risks of natural disasters quantitatively and improve the accuracy of natural disaster management and control, this paper expands the F indicator, Forecast, which is about real-time monitoring and early warning data of natural disasters, and forms the flood control risk trend analysis model framework based on PSR. The framework is named FPSR, i.e. Forecast-Pressure-State-Response, composed of static data and dynamic data. By establishing the four-level index system of flood control risk trend analysis in Fangshan District of Beijing, screening factors, and using analytic hierarchy process, AHP, and experts scoring to determine the weights of each factor, it constructs the flood control risk trend analysis model, FCRTAM. At last, using the real-time monitoring and early warning data of natural disasters in Beijing and the information such as disaster-causing factors, historical natural disasters, major hidden dangers, disaster-bearing bodies, disaster reduction resources (capacities), etc., from National Natural Disaster Comprehensive Risk Census in Fangshan, it analyzes the flood control situation of each town in Fangshan. The results show that the results flood control risk index calculated according to FCRTAM is


INTRODUCTION
Due to its complex geographical environment, there break out various natural disasters in Beijing, especially floods, hail disasters and forest fires, and so on.The natural disasters have caused great threats and losses to the safety of people's lives and property.It is helpful for management natural disaster risk to objectively understand natural disaster risk and accurately grasp the hidden danger and evolution direction of natural disaster.However, natural disaster is characterized by lots of outbreak points and wide influence area and unpredictability, it is difficult to control natural disaster risk with limited manpower.
Quantifying the risk of natural disaster, mastering urban risk, and identifying natural disaster risk levels and preparing countermeasures ensure the pertinence and efficiency of the implementation of natural disaster risk response (Xie, 2021).So, scholars at home and abroad have adopted a variety of methods to assess natural disaster risk.
For the risk assessment of rain and flood disasters, Crichton (Crichton, 2011) proposed the flood risk triangle model, and Aleksandra (Aleksandra, 2011) used 26 indicators to assess the vulnerability of flood risk in Manchester, UK.Benito (Benito, 2004) and Nott (Nott, 2004) assessed urban flood risk by using historical disasters data.By using historical data, Guo The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLVIII-5/W1-2023 International Conference on Geomatics Education -Challenges and Prospects (ICGE22), 10-12 May 2023, Hong Kong SAR, China Tao(Guo, 1991) and Xu Yueqing (Xu, 2001) analyzed the risk of urban flood.Wang Qianwen(Wang, 2021) studied the risk effect of rainstorm and flood disaster based on the system of "3 relationships-2 characteristics-20 indicators".Du Juan(Du,2006) established a prediction model based on flood-related indicators to evaluate the loss of urban flood disasters.

Flood Control Risk Index
The Flood Control Risk Trend Analysis Model, are the values and weights of 2nd-level of F, P, S, and R. Their values' and weights' ranges are from 0 to 1.
The values of the 2nd-level F,P,S and R are combined by 3rd-level indicators' values and its weights, and the 3rd-level indicators' values are combined by the 4thlevel indicators' values and its weights, as follows. (3) In formula ( 3)-( 5), the subscripts of 2, 3 and 4 mean the

Weights Settings
The key of FCRTAM is to determine the indicators of each level and their weights.The method is as  In the table, "+" in indicator directionality means the larger the indicator value is, and the higher the flood control risk is.And "-" means the larger the indicator value is, the lower the risk is.

Data Processing
When In formula (6), ,   and   mean the current value, the minimum value, and the maximum value respectively.And if the indicator's directionality is "+",

Classification Of Risk Grades
According to the historical data and historical trend analysis data of Fangshan Natural Disaster Survey, and comparing between the results of FCRI many times and the actual values, the relationship between FCRI value and risk grade is determined.Table 2 shows the flooding risk grades of each town in Fangshan District with the FCRI((Song, 2022a).

Overview Of The Study Area
The total area of Fangshan District in Beijing is 2,019 square kilometers, with 28 towns.Its annual average The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLVIII-5/W1-2023 International Conference on Geomatics Education -Challenges and Prospects (ICGE22), 10-12 May 2023, Hong Kong SAR, China temperature is 10.8°-11.7°C,and Its annual average precipitation is 602.8-645.3mm.The meteorological disasters often occur in Fangshan, including drought, rainstorm, gale, hail and temperature anomaly.Each area of Plain's, hill's and mountain's in Fangshan is onethird.There are six types of geohazards in Fangshan, including collapse, landslide, mudslide, unstable slope, ground collapse and ground subsidence.

FCRTAM Validation
In 2020, Fangshan District was completed the natural disaster comprehensive risk survey ((Song, 2021b)

ACKNOWLEDGMENTS
assessed mountain torrents in Shidu Town of Beijing.For the risk assessment of typhoon disaster, Hong Yifeng(Hong, 2014) studied the tropical cyclone disaster risk in eastern Zhejiang based on AHP and weighted comprehensive evaluation method.Zhang Yongheng(Zhang, 2009) studied typhoon disasters in Zhejiang Province by constructing typhoon comprehensive evaluation index and disaster level index.Ma Qingyun(Ma, 2008) analyzed and evaluated typhoon disaster level based on weighted average method.Huang Chunzhi(Huang, 2018) constructed and quantitatively analyzed typhoon disasters in Fujian province based on AHP and fuzzy comprehensive evaluation method.For the risk assessment of urban rainstorm disaster, Zhang Yuhua(Zhang, 2019) proposed a fuzzy comprehensive index evaluation system of urban rainstorm based on fuzzy mathematics.Fu Hongen(Fu, 2021) predicted the risk of rainstorm and flood disaster in Shenzhen based on GA-SVR-C.Li Zihua(Li, 2012), Liu Yao(Liu, 2014) and Li Tao(Li, 2016) studied the risk zoning and defense of lightning disaster in City based on GIS and natural disaster risk assessment method.Jin Juliang(Jin, 1998) used cloud model to analyze the temporal and spatial distribution characteristics of drought in Anhui Province.These researches attempt to quantitatively analyze the single disasters such as rainstorm, typhoon, lightning, landslide, geohazard and drought, and so on.But these are few people or no people to study on the comprehensive risk assessment of natural disasters and the development trend of natural disasters.In view of Beijing's comprehensive risk survey of natural disasters carried out from 2020 to 2022, this paper combines the real-time monitoring and early warning and results of natural disasters survey to comprehensively analyze the flood control trend in Beijing(Song, 2022a), especially the flood control risk at the town, to support urban flood prevention and mitigation in Beijing.

FCRTAM
, based on FPSR is composed of four-level indicators, whose standard principles for selecting indicators are scientific and accessible, universal and regional, dynamic and static, etc.The first-level indicator of FCRTAM is Flood Control Risk Index, FCRI.The 2nd-level indicators of FCRTAM consist of F, P, S and R, etc.The 3rd-level indicators and the 4thlevel indicators are selected from the real-time monitoring and early warning data of natural disasters and the results of Fangshan comprehensive risk survey of natural disasters(Song, 2021b), including flooding disaster, rainstorm disaster, earthquake, geohazards, such as landslide, mudslide, collapse, and major safety risks, hazard-affected bodies, disaster mitigation resources (capacities).FCRI is as follows: followings.After selecting and determining the indicators of each level, the judgment matrix of indicators is constructed by Saaty Scaling Law(Saaty, 1980), then the weights of indicators of each level are calculated by AHP.The steps are: (1) Select targeted and reasonable indicators to construct a hierarchical model.(2) Construct the judgment matrix of indicators for each level by comparing indicators in pairs.(3)Calculate the maximum eigenvalue   and eigenvector of the judgment matrix V = { 1 ,  2 , ⋯ ,   , ⋯ ,   } .(4) Check the Consistency of judgment matrix.(5) If the consistency of the matrix is not satisfied, repeat step (2).If the consistency is satisfied, the eigenvector corresponding to the maximum eigenvalue is the weight of the indicator.The process of checking consistency of judgment matrix is as follows.According to the maximum eigenvalue   , calculate the matrix consistency index  = (  − )/( − 1)(n is the order of the matrix).Compare CI with the random consistency index RI of the same order and get the ratio  = /.RI can be treated as a constant with some matrix.If CR<0.1, it indicates that the judgment matrix satisfies the consistency.Considering that FCRTAM is mainly to reflect the impact of natural disasters on urban security, and analyze various natural disasters trend situation, when setting the indicators' weights of FCRTAM, the P indicator's weight of natural disasters is the largest, the F indicator's weight is the second, and the S and R indicators are the third and the fourth.So, in order to highlight F and P indicators, the order of 2nd-level indicators' weights is:   >   >   >   .Table. 1 shows the weights of 2nd-level, 3rd-level and 4th-level indicators of F.
, and the disaster information of the Twhole district was got.So, in the FCRTAM, the values of the 4th-level indicators' values of P, S and R are got from the results of the pilot project of the Fangshan census in 2020, and the values of F are from the monitoring and early warning data from Beijing Meteorological Service, Beijing Water Authority, Beijing Municipal Commission of Planning and Natural Resources, and so on.The results of FCRI and various indicators values of each town in Fangshan District in July 2021 show in Table 4.And the monitoring and early warning data are as followings.
Beijing.In comprehensive consideration of the flood disaster for many years in Beijing, the risk of flood disaster caused by heavy rainfall in July is high.It is necessary to pay attention to urban waterlogging, floods in mountains, floods in small and medium-sized rivers.And Beijing Municipal Commission of Planning and Natural Resources estimates that July will be the peak period of geohazards in Beijing, and the number of geohazards caused by rainfall will increase significantly.It is necessary to pay attention to the prevention of rain-induced collapse, landslide, debris flow, and ground collapse disasters, especially along the traffic line, scenic spots, front and back of houses, debris flow ditches, and goaf.And according to the values, the 4 th -level F indicators' values are in Table 3.According to the rule of FCRI, the flood control risk grade of each town in Fangshan in July 2021 is shown in Figure 2. In Figure 2, there are 10 high-risk towns, which are mainly in mountainous areas, including Z Hebei, Dashiwo, Shidu, Qinglonghu, Hancunhe, Xiayunling, Nanjiao,Fozizhuang, Daanshan and Puwa.There are 14 meddium-risk towns, which are in city center, including Liangxiang, Zhoukoudian, Liulihe, Yancun, Doudian, Shilou, Changgou, Zhangfang, Shijiaying, Xinzhen, Dongfeng, Xingcheng, Xilu and Gongchen, and so on.There are 4 low-risk towns, including Changyang, Chengguan, Xiangyang and Yingfeng.According to the results, it shows that the high-risk areas are mainly the towns with large pressure P indicators, such as rainstorms, floods, debris flows and collapses, including Daanshan, Dashiwo, Fozizhuang, Hancunhe, Nanjiao, Puwa, Qinglonghu, Xiayunling, and the towns with large F indicators values, such as flood, high-impact weather and geohazards, including Hebei and Shidu.The low-risk towns are mainly with high S and R indicators' values, including Changyang, Chengguan and Xiangyang.

Figure 2 .
Figure 2. FCRI map of Fangshan District in July 2021 in each town divided by the number of disasters is basically consistent with the result of the flood control risk situation analysis model.The verification shows that the FRIC results of FCRTAM based on FPSR are basically consistent with the actual situation.When verifying the model, the validity of the model needs to be further verified to support the promotion and application of the model to the whole city, because only the information about geohazards and waterlogging points can be collected, while the loss caused by disasters cannot be collected.4.CONCLUSIONSBy studying the research on the analysis and evaluation of natural disaster risk situation at home and abroad, this paper adds the F indicator, Forecast, which is about real-time monitoring and early warning data of natural disasters, and innovatively puts forward the flood control risk trend analysis model framework, named FPSR, i.e.Forecast-Pressure-State-Response, which is composed of static data and dynamic data and integrates real-time monitoring and early warning of natural disasters, natural disasters, major hidden dangers, disaster bearing body, disaster reduction resources (capacity), and so on.According to the importance of flood control risk trend analysis factors, it establishes the indicator's system for flood control risk trend analysis in Fangshan District of Beijing, selects the indicators and its weights in each level by combining AHP and expert scoring, and builds the flood control risk situation analysis model based on FPSR.Finally, the pre-processed real-time monitoring and early warning data of natural disasters in Beijing and the experimental nation risk survey results of Fangshan are imported into the flood control risk trend analysis model to calculate the FRCI of each town of Fangshan District.The FCRI values are the comprehensive flood control risk trend in Fangshan, which can provide reference for analysis and disaster prevention and mitigation decision-making.

Table 2 .
Flood control risk grades in Fangshan District

Table 3 .
4th-level F indicators' values of each town in Fangshan District in July 2021

Table 4 .
FCRI and various indicators values of each town in Fangshan District in July 2021 (%) The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLVIII-5/W1-2023 International Conference on Geomatics Education -Challenges and Prospects (ICGE22), 10-12 May 2023, Hong Kong SAR, China

Table 5 .
Statistics on the actual number and proportion of disasters in July 2021 In July 2021, the actual weather and disasters in Fangshan District are as follows.It rains for 22 days in July, i.e. 4 days of heavy rain, 3 days of moderate rain, 14 days of thunderstorms, and 1 day of light rain.There are 8 geohazards and 49 waterlogging points, 12 of them are newly added.5 geohazards are in high-risk regions, 3 in medium-risk region.The proportion of

Table 5 .
By analyzing the actual weather and the number of waterlogging points and geohazards in each town in Fangshan districts in July 2021, the risk grade