Human activity signal detection method and device for flood risk and medium

Abstract

The invention relates to the field of earth science, and discloses a human activity signal detection method and device for a flood risk and a medium, and the method comprises the steps: collecting data; flood event identification; carrying out flood event occurrence probability trend quantification; quantifying a flood event attribution risk score; identifying human activity signals in the flood event based on a signal-to-noise ratio criterion; evaluating population and land exposure of flood risks; quantifying future flood magnitude and variation rate; and estimating the influence of the flood magnitude and the change rate on the flood occurrence probability. According to the method, understanding on potential influence of flood can be deepened, and a scientific basis is provided for flood control and disaster reduction planning, urban layout adjustment and climatic change coping strategies in the future.

Description

Technical Field

[0001] The present invention relates to the field of earth science, and in particular to a method, device and medium for detecting human activity signals of flood risks. Background Art

[0002] Global warming and human activities have intensified the terrestrial water cycle and land-atmosphere interaction, changed the underlying water resources, increased the risk of floods, and directly affected the social economy and human health. Observation sites are unevenly distributed around the world, and the time span of runoff data recording is limited. The quality of time series is uneven, making it difficult to accurately assess flood risks based on observational data. In addition, existing studies are usually limited to the response of flood risks to human activities at the regional scale and lack global representativeness. Therefore, understanding the response of global flood risks to the combined effects of climate change and human activities and future risk estimation and attribution are crucial for future disaster prevention strategy formulation, urban flood control planning, and sustainable development. Summary of the invention

[0003] The purpose of the present invention is to explore the relationship between flood risk, climate change and human activities, and to provide a method for detecting human activity signals of flood risk.

[0004] This method quantified the probability of occurrence and changing trend of flood events, constructed a signal identification framework for the joint impact of climate change and human activities on flood risks, captured the signals of human activities in the probability of flood occurrence under future high socioeconomic development path simulation experiments, evaluated the number of people and total land area exposed to flood risks at different future periods and global temperature rise levels, calculated the flood magnitude and variability, clarified the impact of flood magnitude and variability on the probability of flood events, and provided theoretical support for decision makers to make decisions on adaptation and mitigation of climate change.

[0005] The present application provides a method for detecting human activity signals of flood risk, which specifically includes the following steps:

[0006] Step S1: Data collection; collect daily runoff data from the Global Runoff Data Center (GRDC); collect daily runoff data from the historical period based on the observational meteorological model simulation of the third phase of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP 3a), and runoff and average near-surface temperature data from the pre-industrial climate-related experiments, historical experiments and future high socio-economic development path simulation experiments (SSP585) based on the simulated meteorological model simulation of ISIMIP 3b; collect population data under historical and future high socio-economic development path simulation experiments; collect data on dams and corresponding reservoirs for human activities, household and industrial water intake and use, urban area and irrigated land area in the historical period;

[0007] Step S2: flood event identification: combining the site observation data obtained in step S1 and the historical daily runoff data of ISIMIP3a and 3b, using a certain percentile of the daily runoff in the reference period as a threshold, to identify flood events that occurred each year in the study period;

[0008] Step S3: Quantify the trend of flood event probability; Combine the site observations obtained in steps S1 and S2 with the annual flood events simulated by ISIMIP3a and 3b, calculate the proportion of days with flood events in a given period simulated by the site and the model divided by the proportion of days with flood events in the reference period, record it as the probability ratio, and then calculate the time series trend of the probability ratio;

[0009] Step S4: quantifying the flood event attributable risk score; combining the daily runoff data of the ISIMIP 3b global future high socioeconomic development pathway simulation experiment (SSP585) obtained in step S1 and the flood event occurrence probability obtained in step S3, calculating the flood event attributable risk scores at different regional scales;

[0010] Step S5: Identification of human activity signals in flood events based on the signal-to-noise ratio criterion; combining the daily runoff data of the pre-industrial climate-related experiment, historical experiment and future high socio-economic development path simulation experiment SSP585 obtained in step S1, and the probability of flood events obtained in steps S2 and S3, the signal-to-noise ratio framework is used to detect and quantify the signal time when the impact of climate change and human activities on the probability of global flood events exceeds the natural variability for the first time and the corresponding global temperature rise level;

[0011] Step S6: Population and land exposure assessment to flood risk: Combine the population data obtained in step S1 and the climate change and human activity signals in the flood event obtained in step S5 to assess the total population and land area exposed to flood risk at different periods and global temperature rise levels;

[0012] Step S7: quantify the magnitude and variability of future floods; combine the daily runoff data of the ISIMIP 3b historical scenario and the future high socioeconomic development pathway simulation experiment (SSP585) obtained in step S1, use the annual maximum seven-day flow and the Gini coefficient to quantify the relative changes in flood magnitude and variability in different future climate states compared to the historical reference period (1985-2014);

[0013] Step S8: Estimation of the impact of flood magnitude and variability on flood probability; combining the probability of flood events from the future high socioeconomic development pathway simulation experiment (SSP585) obtained in step S3 and the relative changes in flood magnitude and variability obtained in step S7, for areas with increased flood risk, evaluate the dominant factors of increased risk based on the Student's t-test.

[0014] A storage medium stores instructions and data for implementing a method for detecting human activity signals of flood risks.

[0015] A flood risk human activity signal detection device comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a flood risk human activity signal detection method.

[0016] The beneficial effects provided by the present invention are:

[0017] (1) The present invention provides a comprehensive flood risk prediction method that takes into account the combined impact of climate change and human activities on flood risk. The present invention accurately identifies and quantifies the probability of observed and simulated flood events, and deepens the understanding of the response mechanism of flood risk to climate change and human activities. In addition, by evaluating the probability of flood events under simulation experiments of future high socioeconomic development paths, the present invention provides a scientific basis for evaluating long-term changes in flood risk.

[0018] (2) The present invention innovatively introduces the signal-to-noise ratio criterion to identify the time when the climate change and human activity signals in flood events exceed the natural variability for the first time, and to evaluate the total population and land area exposed to flood risks in different periods and global temperature rise levels, providing strong technical support and reference for in-depth exploration of the driving factors and impacts of flood risks.

[0019] (3) This paper quantifies the relative changes in flood magnitude and variability during different future climate states, and provides key information for decision makers to formulate effective flood risk management strategies by assessing the main driving factors in areas with increased flood risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of a method, device and medium for detecting human activity signals of flood risk provided by the present invention;

[0021] Figure 2 It is a schematic diagram considering the historical underlying surface water and soil management factors and changes;

[0022] Figure 3 It is a comparison chart of flood probability trends for different runoff datasets from 1971 to 2019;

[0023] Figure 4 It is a graph showing the probability of flood events occurring;

[0024] Figure 5 is a graph showing the relationship between the global area-weighted average flood event probability and attributable risk score for different percentile thresholds for identifying flood events and the global temperature rise level and different periods;

[0025] Figure 6 It is the probability distribution map of global flood events identified by different percentile thresholds at 1.5, 2 and 3℃ temperature rise levels;

[0026] Figure 7 It is a map of the occurrence time and temperature rise magnitude of climate change and human activity signals in flood events, as well as the total population and land area exposure;

[0027] Figure 8 It is a graph of relative changes in flood magnitude and variability in the future high socio-economic development pathway simulation experiment (SSP585);

[0028] Fig. 9 It is a working schematic diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0029] To make the purpose, technical solution and advantages of the present invention clearer, the following will further describe the implementation of the present invention in conjunction with the accompanying drawings. The present invention provides a method, device and medium for detecting human activity signals of flood risk, which specifically include the following steps:

[0030] Step S1: Data collection; collect daily runoff data from observational sites of the Global Runoff Data Center GRDC; collect daily runoff data from the historical period based on the observational meteorological model simulation of the third phase of the Inter-Sectoral Impact Model Intercomparison Project ISIMIP 3a, and daily runoff and average near-surface temperature data from pre-industrial climate-related experiments, historical experiments and future high socio-economic development path simulation experiments (SSP585) based on simulated meteorological models of ISIMIP 3b; collect population data under historical and future high socio-economic development path simulation experiments; collect data on dams and corresponding reservoirs for human activities, household and industrial water intake and use, urban area and irrigated land area in the historical period;

[0031] Step S2: When quantifying the probability of flood events from 1971 to 2019, the 90th percentile of daily runoff from 1981 to 2010 is used as the threshold; when quantifying the probability of flood events from 1865 to 2065, the 99th percentile of daily runoff from 1601 to 1849, the entire period of pre-industrial climate-related experiments, is used as the threshold.

[0032] Step S3: Quantify the trend of flood event probability; Combine the site observations obtained in steps S1 and S2 with the annual flood events simulated by ISIMIP3a and 3b, calculate the proportion of flood event days in a given period for the site and simulation divided by the proportion of flood event days in the reference period, record it as the probability ratio, and then calculate the time series trend of the probability ratio;

[0033] In step S3, the probability of occurrence of flood events is calculated by using the observed and simulated daily runoff data to calculate the probability of occurrence of flood events (PR) year by year. The calculation formula is as follows:

[0034]

[0035] Among them, P1 is the probability of occurrence of a flood event in a given period, and P0 is the probability of occurrence of the flood event in the historical reference period.

[0036] According to the obtained time series of flood event probability from 1971 to 2019, it is homogenized (divided by the average occurrence probability), and the series change trend is calculated based on the Sen slope.

[0037] Step S4: quantifying the flood event attributable risk score; combining the daily runoff data of the ISIMIP 3b global future high socioeconomic development pathway simulation experiment (SSP585) obtained in step S1 and the flood event occurrence probability obtained in step S3, calculating the flood event attributable risk scores at different regional scales;

[0038] In step S4, based on the probability of occurrence of flood events, the flood event attributable risk scores (FAR) at different regional scales are calculated, and the calculation formula is as follows:

[0039]

[0040] Among them, P1 is the probability of occurrence of a flood event in a given period, and P0 is the probability of occurrence of the flood event in the historical reference period.

[0041] Step S5: Identification of human activity signals in flood events based on the signal-to-noise ratio criterion; combining the daily runoff data of pre-industrial climate-related experiments, historical experiments, and future high socioeconomic development path simulation experiments (SSP585) obtained in step S1, and the probability of flood events from 1971 to 2065 obtained in steps S2 and S3, the signal-to-noise ratio framework is used to detect and quantify the signal time when the impact of climate change and human activities on the probability of global flood events exceeds the natural variability for the first time and the corresponding global temperature rise level;

[0042] In step S5, the noise of natural variability is defined as half the standard deviation of the probability of flood events in the baseline period of 1601-1849 in pre-industrial climate-related experiments, and the change signal is defined as the difference between the mean probability of flood events between every 30-year sliding window and the historical period 1971-2019 starting from 1970. When the change signal exceeds the natural variability noise and lasts for no less than 20 years, it is considered that there is a signal of the impact of climate change and human activities, and the corresponding appearance time is the central year of the 30-year sliding window (for example, 1975 represents the climate state level from 1956 to 1985), and the corresponding global temperature rise is the multi-model average temperature rise level.

[0043] Step S6: Population and land exposure assessment to flood risk; combining the population data obtained in step S1 with the climate change and human activity signals in the flood event obtained in step S5, assess the total population and land area exposed to flood risk at different periods and global temperature rise levels;

[0044] In step S6, when quantifying the total population and land area exposed to flood risks, for a specified year or temperature rise level, for the grid points where climate change and human activity impact signals appeared before this year or temperature rise level, the total population and land area in this year or this temperature rise level are calculated.

[0045] Step S7: quantify the magnitude and variability of future floods; combine the daily runoff data of the ISIMIP 3b historical experiment and the future high socioeconomic development path simulation experiment (SSP585) obtained in step S1, use the annual maximum seven-day flow and the Gini coefficient to quantify the relative changes in flood magnitude and variability in different future climate states compared to the historical reference period (1985-2014);

[0046] In step S7, the maximum 7-day arithmetic mean flow (MAX7) and the Gini coefficient (Gini) are used to quantify the flood magnitude and variability under the combined influence of climate change and human activities, respectively. The Gini coefficient calculation formula is as follows:

[0047]

[0048] Among them, y i is the MAX7 of the i-th year, and n is the total number of years. The Gini coefficient ranges from 0 to 1. The lower the Gini coefficient, the more uniform the time variation of MAX7 and the smaller the variability.

[0049] After calculating the global annual MAX7 and variability time series, we selected 1985-2014 as the historical reference period and calculated the relative changes of MAX7 and Gini coefficient in different 30-year climate states every 10 years from 2015 to 2099. The calculation formula is as follows:

[0050]

[0051] Among them, P1 is the historical reference period and P2 is the specified climate state period.

[0052] Step S8: Estimation of the impact of flood magnitude and variability on flood probability; combining the probability of flood events of the future high socio-economic development path simulation experiment (SSP585) obtained in step S3 and the relative changes in flood magnitude and variability obtained in step S7, the dominant factors of increased risk are evaluated based on the Student's t-test;

[0053] In step S8, in order to clarify the dominant factor of the increased probability of flood events (grid cells with probability of occurrence ≥ 1), the flood magnitude and variability in the historical period (1985-2014) and the future period (2070-2099) are subjected to a Student's t test. When the flood magnitude or variability shows an upward trend and the significance test p<0.05, this indicator is considered to be the dominant factor leading to the increased probability of flood events.

[0054] As an implementation case, the present invention is further described. The implementation case is used to illustrate the present invention, but this case is not used to limit the scope of application of the present invention. It is also applicable to different regions and other time periods.

[0055] The implementation flow chart of a flood risk human activity signal detection method, device and medium of the method of the present invention is as follows: Figure 1 As shown, the specific steps are as follows:

[0056] (1) Collection of test data;

[0057] In this implementation case, long-term daily flow observation data from 1971 to 2019 were collected from 10,831 stations of the Global Runoff Data Center (GRDC) version 3 and 375 stations of the Ministry of Water Resources of China. The criteria for selecting meteorological stations are as follows: 1) the number of days without observations in a year shall not exceed 60 days; 2) the qualified years shall not be less than 30 years. Finally, 4,366 stations around the world were selected for calculation.

[0058] Daily runoff data for ISIMIP 3a and 3b were also collected, as shown in Tables 1 and 2 below.

[0059] Table 1 ISIMIP 3a models selected under the direct human activity scenario of historical experiments (obsclim_histsoc) and corresponding simulation time

[0060]

[0061] Table 2 Selected ISIMIP 3b pre-industrial climate-related experiments, historical experiments, and future high socioeconomic development path simulation experiments (SSP585) and corresponding simulation times

[0062]

[0063]

[0064] At the same time, population data under simulation experiments of historical and future high socioeconomic development paths were collected, as well as data on dams and corresponding reservoirs, household and industrial water withdrawal and use, urban area and irrigated land area of ​​human activities in historical periods, and the changes and global distribution of different human activity factors. Figure 2 shown.

[0065] (2) Flood event identification and quantification of occurrence probability trends

[0066] like Figure 3 As shown, the methods of steps S2 and S3 are adopted, and the 90th percentile of daily runoff from 1981 to 2010 is used as the threshold. The proportion of days with flood events occurring annually in a given period of time for the site and simulation is calculated and divided by the proportion of days with flood events occurring in the reference period to quantify the probability of flood events from 1971 to 2019. The probability is then homogenized (divided by the average probability of occurrence), and the series change trend is calculated based on the Sen slope.

[0067] The spatial pattern of flood event probability trends from 1971 to 2019 identified based on runoff records at observation stations showed significant regional differences ( Figure 3 a). Above 50°N and in Central Africa, the probability of flood events shows a significant upward trend, while below 50°N, especially in South America and East Asia, the probability of flood events shows a downward trend. These flood event probability distribution patterns also exist in the hydrological model driven by meteorological observation data (ISIMIP 3a) ( Figure 3 b) and hydrological model driven by meteorological simulation data under climate change and human activities (ISIMIP 3b) ( Figure 3 c) Simulation. Meanwhile, the probability of global flood events decreased from 1971 to 1987, and gradually increased after 1987 ( Figure 3 d). There is no significant change trend in the probability of flood events in areas below 50°N (p>0.05, Figure 3 f), while the trend of flood event probability in ISIMIP 3a (3.10% / 10 years, p<0.05) and 3b (1.68% / 10 years, p<0.05) in areas above 50°N increased significantly ( Figure 3e) The trend of the probability of historical flood events in ISIMIP 3a even exceeds the 90th percentile confidence interval of the probability of flood events in pre-industrial climate-related experiments (2.90% / 10 years). Figure 3 The global spatial patterns of the changing trends in the probability of flood events in b and c correspond to each other, indicating that above 50°N, climate change and human activities in the historical period have gradually amplified the increasing trend of the probability of flood occurrence, while their impact signals have not yet appeared in the global flood probability during this period.

[0068] (3) Estimation of the probability of future flood events and quantification of attributable risk scores

[0069] like Figure 4 As shown in Figure 1, the method of steps S3, S4 and S5 is used. This case uses the 99th percentile of daily runoff from 1601 to 1849 in the ISIMIP 3b pre-industrial climate-related experiment as the threshold to quantify the probability of flood events from 1865 to 2065 and calculate the attributable risk score (FAR) of flood events at different regional scales. The ISIMIP 3b simulation data set shows that the probability of flood events and the attributable risk score in the global land area are positively correlated with global temperature rise, that is, as the temperature rise level increases, the global flood risk increases ( Figure 4 Compared with a temperature rise of 1.5°C, the probability of flood events increases by 10.1% under a temperature rise of 2°C ( Figure 4 a), the corresponding attributable risk score is about 0.70, that is, at a temperature rise of 2°C, about 70% of global flood events can be attributed to climate change and human activities ( Figure 4 c).

[0070] As climate warming intensifies, the probability of floods in more and more regions is increasing, and this trend is gradually amplified under the influence of climate change and human activities. As the global temperature rises from 1.5℃ to 3℃, the proportion of areas with increased probability of flood events increases from 59.6% to 69.4% of the global land area, among which the probability of flood events near the equator and above 50°N increases the most ( Figure 4 Specifically, the probability of flood events in high-latitude regions and Central Africa, South Asia and southern Australia increased, while the probability of flood events in southern North America, northern South America, the Mediterranean and Europe decreased, and the probability of flood events in parts of southern Russia and Central Asia showed a trend of changing from decreasing to increasing. These findings indicate that more and more regions around the world, especially high-latitude regions, will face serious flood risks, and climate change and human activities will further exacerbate this risk.

[0071] Under the global warming level of 1.5-3℃, the basins with positive flood event attributable risk scores account for 71.9% to 87.5% of the total basins ( Figure 4 h, j and l), which shows that climate change and human activities have a positive effect on the probability of floods in most basins and are significantly affected by climate change and human activities (signal-to-noise ratio ≥ 1). Among them, the flood risks in the Niger and Ob River basins are most affected by climate change and human activities. At the climate zone scale, the flood attributable risk scores of arid areas are the highest under different global temperature rise levels, and they face the greatest flood risks ( Figure 4 i, k and m).

[0072] like Figure 5 and Figure 6 As shown in the figure, the 90th, 95th, 99.5th and 99.9th percentiles of daily runoff from the entire period of the ISIMIP 3b pre-industrial climate-related experiment from 1601 to 1849 are used as thresholds to quantify the probability of flood events from 1865 to 2065. As the threshold increases, the probability of global flood events and the attributable risk score increase, and are positively correlated with temperature rise and year ( Figure 5 ). In terms of spatial distribution, the larger the threshold selected for identifying flood events, the greater the trend of the probability of flood events. That is, in areas where the probability of flood events shows a decreasing / increasing trend, the decreasing / increasing trend will be more drastic when the threshold is larger ( Figure 6 ).

[0073] (4) Identification of human activity signals in flood events and assessment of population and land exposure to flood risk based on the signal-to-noise ratio criterion

[0074] like Figure 7As shown, this case adopts the method of step S5, combined with the daily runoff data of the ISIMIP 3b pre-industrial climate-related experiment, historical experiment and future high socio-economic development path simulation experiment (SSP585) obtained in step S1, and the probability of flood events from 1971 to 2065 obtained in steps S2 and S3, and uses the signal-to-noise ratio framework to detect and quantify the time when the impact of climate change and human activities on the probability of global flood events first exceeds the signal of natural variability and the corresponding global temperature rise level. The noise of natural variability is defined as half of the standard deviation of the probability of flood events in the baseline period of pre-industrial climate-related experiments from 1601 to 1849, and the change signal is defined as the difference between the mean probability of flood events between each 30-year sliding window and the historical scenario from 1971 to 2019 starting from 1970. When the change signal exceeds the natural variability noise and lasts for at least 20 years, it is considered to be a signal of climate change and the impact of human activities. The corresponding appearance time is the central year of the 30-year sliding window (for example, 1975 represents the climate state level from 1956 to 1985), and the corresponding global temperature rise is the multi-model average temperature rise level.

[0075] In this case, based on step S6, combined with the population data obtained in step S1 and the climate change and human activity signals in flood events obtained in step S5, for a specified year or temperature rise level, for the grid points where climate change and human activity impact signals appeared before this year or temperature rise level, the total population and land area in this year or this temperature rise level are calculated.

[0076] Globally, by 2065, under the future high socioeconomic development path simulation experiment (SSP585), the areas where climate change and human activities influence signals in the probability of flood events are detected exceed 38.8% of the total land area ( Figure 7 a). In most parts of Central Africa, Russia and Central Asia, the signals of climate change and human activities in the probability of flood events appear earlier (before 2050), while in eastern Brazil, southern Russia, Kazakhstan and Indonesia, the impact signals are later than 2050. At the same time, the signals of climate change and human activities in the probability of flood events in these regions correspond to lower global temperature rise levels (earlier than the global warming level of 2°C) ( Figure 7 c). At the basin level, 43.8% of the basins detected signals of climate change and human activities in the probability of flood events, among which the signals appeared earliest in the Mekong River and Congo River basins ( Figure 7 b) The Mekong River Basin signal appeared even before global warming reached 1.5°C ( Figure 7d), which indicates that climate change and human activities have a more significant impact on the probability of flood occurrence in these basins. Among the 32 basins, the Congo Basin has the earliest signal of climate change and human activities in the probability of flood events, which is earlier than the year when the signal appears under the global temperature rise of 2°C ( Figure 7 e dot corresponds to time).

[0077] By 2060 (or a global temperature rise of 3°C), it is expected that there will be 12.9×10 8 (14.6×10 8 ) and 3.5×10 7 (3.8×10 7 )km 2 of land is affected by the probability of flood events ( Figure 7 f and g). However, before the global temperature rise reached 1.5℃, the proportion of population and area affected by flood events was less than 10% of the total affected population and area. This huge difference shows that if global warming is limited to 1.5℃ instead of 2℃, it is very likely that 2.7×10 7 km 2 of land area and 10 9 People avoid exposure to flood events.

[0078] (5) Quantification of future flood magnitude and variability and prediction of their impact on flood probability

[0079] like Figure 8 As shown, this case adopts the method of step S7, combined with the daily runoff data of the ISIMIP 3b historical scenario and the future high socioeconomic development pathway simulation experiment (SSP585) obtained in step S1, using the annual maximum seven-day flow (MAX7) and the Gini coefficient (Gini) to quantify the relative changes in flood magnitude and variability in different 30-year climate states every 10 years from 2015 to 2099 compared with the historical reference period (1985-2014).

[0080] In this case, based on step S8, combined with the probability of flood events in the future high socioeconomic development path simulation experiment (SSP585) obtained in step S3 and the relative changes in flood magnitude and variability obtained in step S7, a Student's t-test is performed on the flood magnitude and variability in the historical period (1985-2014) and the future period (2070-2099) for areas with increased flood risk (grid cells with probability of occurrence ≥ 1). When the flood magnitude or variability shows an upward trend and the significance test p<0.05, this indicator is considered to be the dominant factor leading to the increased probability of flood events.

[0081] From 2070 to 2099, the magnitude and variability of global floods are both increasing. On a global scale, 75.0% and 60.9% of land areas are experiencing an increasing trend in flood magnitude and variability, respectively ( Figure 8 a), while the areas where flood magnitude and variability showed a decreasing trend were mainly distributed in the northern United States, northern South America, Argentina, Morocco, southern Africa, the Mediterranean, western Europe and southern Australia, which is highly consistent with the areas where the probability of flood occurrence decreased ( Figure 4 At the basin scale, flood magnitude (27 / 32) and variability (25 / 32) in most basins showed an increasing trend ( Figure 8 b). Over time, the relative changes in flood magnitude in the four climate zones all showed an increasing trend, with the largest increases in flood magnitude in the temperate and arid zones, while the relative changes in flood variability in the four climate zones showed a slight increasing trend over the 80-year period (2015-2094) ( Figure 8 c). From 2070 to 2099, the relative changes in flood magnitude and variability in the four climate zones all showed an increasing trend, among which the relative change in flood magnitude in the arid zone was the largest (31.1%), the relative change in flood variability in the temperate zone was the largest (15.4%), and the relative change in flood magnitude and variability in the frigid zone was the smallest (the relative changes in flood magnitude and variability were 11.5% and 3.4%, respectively) ( Figure 8 d).

[0082] During the period 2070-2099, among the regions where the probability of flood events is predicted to increase, the regions where only flood magnitude, only flood variability, and both flood magnitude and variability show significant increasing trends account for 57.4%, 5.5%, and 12.7%, respectively. Figure 8 e). In addition, flood magnitude, variability and relative changes of floods in 43.8%, 9.4% and 31.2% of the basins showed a significant increase ( Figure 8 f). This suggests that increases in flood magnitudes in most regions and river basins are the main driver of the increased probability of flood events.

[0083] See also Fig. 9 , Fig. 9 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically comprises: a human activity signal detection device 401 for flood risk, a processor 402 and a storage medium 403.

[0084] A human activity signal detection device 401 for flood risk: The human activity signal detection device 401 for flood risk implements the human activity signal detection method for flood risk.

[0085] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the human activity signal detection method for flood risk.

[0086] Storage medium 403: the storage medium 403 stores instructions and data; the storage medium 403 is used to implement the human activity signal detection method for flood risk.

[0087] The beneficial effects of the present invention are:

[0088] (1) The present invention provides a comprehensive flood risk prediction method that takes into account the combined impact of climate change and human activities on flood risk. The present invention accurately identifies and quantifies the probability of observed and simulated flood events, thereby deepening the understanding of the response mechanism of flood risk to climate change and human activities. In addition, by evaluating the probability of flood events under simulation experiments of future high socioeconomic development paths, the present invention provides a scientific basis for evaluating long-term changes in flood risk.

[0089] (2) The present invention innovatively introduces the signal-to-noise ratio criterion to identify the first impact signal time of climate change and human activity signals in flood events, and to evaluate the total population and land area exposed to flood risks in different periods and global temperature rise levels, providing strong technical support and reference for in-depth exploration of the driving factors and impacts of flood risks.

[0090] (3) This paper quantifies the relative changes in flood magnitude and variability during different future climate states, and provides key information for decision makers to formulate effective flood risk management strategies by assessing the main driving factors in areas with increased flood risk.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting human activity signals of flood risk, characterized in that: The following steps are involved: Step S1: Data collection; collect daily runoff data from the Global Runoff Data Center (GRDC); collect daily runoff data from the historical period based on the observational meteorological model simulation of the third phase of the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP 3a), and daily runoff and average near-surface temperature data from the pre-industrial climate-related experiments, historical experiments and future high socio-economic development path simulation experiment SSP585 based on the simulated meteorological model simulation of ISIMIP 3b; collect population data under historical and future high socio-economic path scenario simulation experiments; collect data on dams and corresponding reservoirs for human activities, household and industrial water intake and use, urban area and irrigated land area in the historical period; Step S2: flood event identification: combining the site observation data obtained in step S1 and the historical daily runoff data of ISIMIP3a and 3b, using a certain percentile of the daily runoff in the reference period as a threshold, to identify flood events that occurred each year during the study period; Step S3: Quantify the trend of flood event probability; Combine the site observations obtained in steps S1 and S2 with the annual flood events simulated by ISIMIP3a and 3b, calculate the proportion of flood event days in a given period in the site observations and model simulations divided by the proportion of flood event days in the reference period, record it as the probability ratio, and then calculate the time series trend of the probability ratio; Step S4: quantifying the flood event attributable risk score; combining the daily runoff data of the ISIMIP 3b global future high socio-economic development path simulation experiment SSP585 obtained in step S1 and the flood event occurrence probability obtained in step S3, calculating the flood event attributable risk scores at different regional scales; Step S5: Identification of human activity signals in flood events based on the signal-to-noise ratio criterion; combining the daily runoff data of the pre-industrial climate-related experiment, historical experiment and future high socio-economic development path simulation experiment SSP585 obtained in step S1, and the probability of flood events obtained in steps S2 and S3, the signal-to-noise ratio framework is used to detect and quantify the signal time when the impact of climate change and human activities on the probability of global flood events exceeds the natural variability for the first time and the corresponding global temperature rise level; Step S6: Population and land exposure assessment to flood risk: Combine the population data obtained in step S1 with the climate change and human activity signals in the flood events identified in step S5 to assess the total population and land area exposed to flood risk at different periods and global temperature rise levels; Step S7: quantify the magnitude and variability of future floods; combine the daily runoff data of the ISIMIP 3b historical experiment and the future high socioeconomic development path simulation experiment SSP585 obtained in step S1, use the annual maximum seven-day flow and the Gini coefficient to quantify the relative changes in flood magnitude and variability in different future climate states compared with the historical reference period 985-2014; Step S8: Estimation of the impact of flood magnitude and variability on flood probability; combining the probability of flood events of the future high socio-economic development pathway simulation experiment SSP585 obtained in step S3 and the relative changes in flood magnitude and variability obtained in step S7, for areas with increased flood risk, evaluate the dominant factors of increased risk based on the Student's t-test.

2. A method for detecting flood risk human activity signals according to claim 1, characterized in that: In step S2, when quantifying the probability of flood events from 1971 to 2019, the 90th percentile of daily runoff from 1981 to 2010 is used as the threshold; when quantifying the probability of flood events from 1865 to 2065, the 99th percentile of daily runoff from 1601 to 1849 for the entire period of pre-industrial climate-related experimental scenarios is used as the threshold.

3. The method, device and medium for detecting flood risk human activity signals according to claim 1, characterized in that: Step S3 is as follows: S31: Calculation of flood event probability, using observation and simulation data to calculate the annual flood probability (PR), the calculation formula is as follows: Among them, P1 is the probability of occurrence of a flood event in a given period, and P0 is the probability of occurrence of the flood event in the historical reference period; S32: Calculation of time series trend. Based on the obtained time series of flood event probability from 1971 to 2019, it is homogenized and the series change trend is calculated based on the Sen slope.

4. The method for detecting human activity signals of flood risk according to claim 1, characterized in that: In step S4, based on the probability of flood events, the flood event attribution risk scores FAR at different regional scales are calculated as follows: Among them, P1 is the probability of occurrence of a flood event in a given period, and P0 is the probability of occurrence of the flood event in the historical reference period.

5. The method, device and medium for detecting flood risk human activity signals according to claim 1, characterized in that: In step S5, the noise of natural variability is defined as half of the standard deviation of the probability of flood events in the baseline period of pre-industrial climate-related experiments from 1601 to 1849, and the change signal is defined as the difference between the mean probability of flood events in every 30-year sliding window starting from 1970 and the historical period from 1971 to 2019; when the change signal exceeds the natural variability noise and lasts for at least 20 years, it is regarded as the presence of a signal of the impact of climate change and human activities, and the corresponding appearance time is the central year of the 30-year sliding window, and the corresponding global temperature rise is the multi-model average temperature rise level.

6. The method, device and medium for detecting flood risk human activity signals according to claim 1, characterized in that: In step S6, when quantifying the total population and land area exposed to flood risks, for a specified year or temperature rise level, for the grid points where climate change and human activity impact signals appeared before this year or temperature rise level, the total population and land area in this year or this temperature rise level are calculated.

7. The method for detecting human activity signals of flood risk according to claim 1, characterized in that: Step S7 is specifically as follows: S71: Two indicators, namely the annual maximum 7-day arithmetic mean flow MAX7 and the Gini coefficient Gini, are used to quantify the magnitude and variability of floods under the combined influence of climate change and human activities. The calculation formula of the Gini coefficient is as follows: Among them, y i is the MAX7 of the i-th year, and n is the total number of years; the Gini coefficient ranges from 0 to 1. The lower the Gini coefficient, the more uniform the temporal variation of MAX7 and the smaller the variability; S72: After calculating the global annual MAX7 and variability time series, select 1985-2014 as the historical reference period, and calculate the relative changes of MAX7 and Gini coefficient in different 30-year climate states every 10 years from 2015 to 2099; the calculation formula is as follows: Among them, P1 is the reference period and P2 is the specified climate state period.

8. The method for detecting human activity signals of flood risk according to claim 1, characterized in that: In step S8, in order to clarify the dominant factor for the increased probability of flood events (grid cells with probability of occurrence ≥ 1), a Student's t-test is performed on the flood magnitude and variability in the historical period 1985-2014 and the future period 2070-2099; when the flood magnitude or variability shows an upward trend and the significance test p < 0.05, this indicator is considered to be the dominant factor leading to the increased probability of flood events.

9. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a method for detecting human activity signals of flood risk as described in any one of claims 1 to 8.

10. A flood risk human activity signal detection device, characterized in that: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a method for detecting human activity signals of flood risk as described in any one of claims 1 to 8.

Images