Multi-temporal-spatial-scale urban rainfall attribution and influence evaluation method, equipment and medium

Abstract

The invention relates to the field of carbon neutralization, and discloses a multi-temporal-spatial-scale urban rainfall attribution and influence evaluation method, device and medium, and the method comprises the steps: collecting data; identifying the urbanization degree; analyzing the influence of urbanization on the long-term change trend of extreme rainfall; carrying out urbanization to carry out unsteady-state frequency analysis on extreme rainfall; the cause of urbanization on the unsteady influence of extreme rainfall; performing attribution analysis on the contribution of climate change and urbanization to the extreme rainfall trend; the method has the beneficial effects that the influence and attribution of urbanization on the multi-temporal-spatial-scale precipitation mode can be understood, and some more valuable reference bases and methods are provided for urban planning and climate adaptation strategies in China.

Description

Technical Field

[0001] The present invention relates to the field of carbon neutrality, and in particular to a multi-temporal and spatial scale urban precipitation attribution and impact assessment method, equipment and medium. Background Art

[0002] As global carbon emissions continue and intensify, anthropogenic climate change increases precipitation variability and precipitation magnitude, leading to more frequent and intense extreme precipitation. In particular, China, located in the East Asian monsoon region, is constantly facing severe extreme precipitation under the influence of anthropogenic climate change, causing cascading effects such as floods and landslides. At the same time, China, as the most populous country in the world, is undergoing rapid urbanization, with the urbanization rate rapidly increasing from 17.92% in 1980 to 65.2% in 2022. This intensive and rapid urbanization process has a significant impact on regional precipitation patterns by changing surface characteristics and atmospheric boundary layer structure, changing energy balance and regional water cycle processes, resulting in more intense precipitation in urban areas relative to rural areas. Changes in precipitation may cause an increase in extreme precipitation, posing a serious threat to people's lives and property. Therefore, in the past few decades, many studies have evaluated the impact of urbanization on precipitation based on empirical measurements and numerical model simulations.

[0003] Given that the increase in extreme precipitation in urban areas is caused by both anthropogenic climate change and urbanization, it is important to separate and quantify the contributions of anthropogenic climate change and urbanization to precipitation change. However, there is currently very limited research on the impact of urbanization on precipitation at multiple spatiotemporal scales in China. In addition, due to the complex interactions between climate change and urbanization, their respective contributions to precipitation change are still unclear. Therefore, quantitatively assessing the impact and attribution of urbanization on precipitation at multiple spatiotemporal scales in China will help to further deepen our understanding and fully recognize the intrinsic connection between urbanization and precipitation change. It can also provide a scientific basis for urban planning, water resources management, and flood prevention and disaster reduction, enhance the ability of cities to adapt to extreme weather events under the background of climate change, and provide support for the sustainable development of cities in my country. Summary of the invention

[0004] The purpose of the present invention is to propose a multi-temporal and spatial scale urban precipitation attribution and impact assessment method, equipment and medium to solve the technical problem of insufficient research methods on the impact of climate change and urbanization on multi-temporal and spatial scale precipitation in China.

[0005] The present invention provides a multi-temporal and spatial scale urban precipitation attribution and impact assessment method, comprising the following steps:

[0006] S1: Data collection: collect daily precipitation observation data, land use coverage data, population density data and large-scale meteorological data;

[0007] S2: Identification of urbanization degree: combining daily precipitation observation data, land use coverage data and population density data obtained in step S1 to identify rural, urbanized and urban sites in China;

[0008] S3: Analysis of the impact of urbanization on the long-term trend of extreme precipitation; combining the rural, urbanized and urban sites obtained in step S2, quantifying the annual and seasonal trends of extreme precipitation, conducting mutation point analysis, and detecting the non-steady state of extreme precipitation;

[0009] S4: Analysis of the non-steady frequency of extreme precipitation due to urbanization; Combined with the non-steady stations analyzed in step S3, non-steady frequency analysis is performed to evaluate the non-steady state of extreme precipitation;

[0010] S5: Causes of the non-steady-state impact of urbanization on extreme precipitation; combining the daily precipitation observation data and large-scale meteorological data obtained in step S1, analyzing the extreme precipitation changes caused by large-scale circulation or local changes;

[0011] S6: Attribution analysis of the contribution of climate change and urbanization to extreme precipitation trends; Combine the rural, urbanized, and urban sites obtained in step S2 to clarify the individual contributions of climate change and urbanization to extreme precipitation trends at the national, regional, and local scales.

[0012] A storage medium stores instructions and data for implementing a multi-temporal and spatial scale urban precipitation attribution and impact assessment method.

[0013] A multi-temporal and spatial scale urban precipitation attribution and impact assessment device comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a multi-temporal and spatial scale urban precipitation attribution and impact assessment method.

[0014] The beneficial effects provided by the present invention are:

[0015] (1) The impact of urbanization on precipitation in the present invention is a multi-temporal and spatial scale analysis, which can comprehensively analyze the trend of extreme precipitation from different time scales and spatial scales (time: annual and seasonal; space: national, regional and local). The impact of urbanization on precipitation varies with the change of research area and time, and the results are extensive and generalizable.

[0016] (2) At present, there are many studies on the impact of urbanization on precipitation, but the research on the impact of urbanization on precipitation at multiple spatiotemporal scales in China is very limited. In addition, due to the complex interactions between climate change and urbanization, their respective contributions to extreme precipitation changes are still unclear. The proposed method helps to understand the impact and attribution of urbanization on precipitation patterns at multiple spatiotemporal scales, and provides some more valuable references and methods for urban planning and climate adaptation strategies in my country. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic flow chart of the method of the present invention;

[0018] Figure 2 are the time series of mean annual and seasonal extreme precipitation at rural, urbanized, and urban stations in China;

[0019] Figure 3 It is a comparison of the steady-state and non-steady-state return periods of the latitudinal mean annual and seasonal extreme precipitation in the northern climate zone (Zone VI);

[0020] Figure 4 is the change in daily precipitation quantile in the first cluster obtained by clustering with machine learning algorithm in each region;

[0021] Figure 5 is the change in daily precipitation median in the second cluster obtained by clustering with machine learning algorithm in each region;

[0022] Figure 6 is the relative individual contribution of climate change and urbanization to the trends of the nine precipitation indices at the national level;

[0023] Figure 7 is the relative individual contribution of climate change and urbanization to the trends of the nine precipitation indices within urban areas;

[0024] Figure 8 It is a working schematic diagram of the hardware device of the present invention. DETAILED DESCRIPTION

[0025] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0026] Before formally describing the present invention, the scheme of the present invention is first generally described for easy understanding.

[0027] Please refer to Figure 1 , Figure 1 It is a schematic diagram of the process of the present invention.

[0028] The present invention provides a multi-temporal and spatial scale urban precipitation attribution and impact assessment method, comprising the following steps:

[0029] S1: Data collection: collect daily precipitation observation data, land use coverage data, population density data and large-scale meteorological data;

[0030] It should be noted that, in the method of the present invention, in step S1, the daily precipitation observation data is the daily precipitation records of 1857 stations covering the period from 1961 to 2014; the land use coverage data should include surface type data (such as impervious surface) representing the city; the population density data is data with a spatial resolution of 1km×1km; the large-scale meteorological data includes temperature, humidity, air pressure, zonal wind, meridional wind and vertical wind speed.

[0031] S2: Identification of urbanization degree: combining daily precipitation observation data, land use coverage data and population density data obtained in step S1 to identify rural, urbanized and urban sites in China;

[0032] It should be noted that in step S2, according to the daily precipitation observation data, land use data and population density data obtained in step S1, when the site is located in an urban area and the population density is greater than 1000 people / km 2 When the site is assigned, it is defined as an urban site. The sites within a radius of 50 km with the urban site as the center are defined as urbanized sites, while other unassigned sites are defined as rural sites.

[0033] S3: Analysis of the impact of urbanization on the long-term trend of extreme precipitation; combining the rural, urbanized and urban sites obtained in step S2, quantifying the annual and seasonal trends of extreme precipitation, conducting mutation point analysis, and detecting the non-steady state of extreme precipitation;

[0034] It should be noted that in step S3, extreme precipitation is defined as the maximum annual and seasonal (four seasons) precipitation at each station. The nonparametric Mann-Kendall test is used to quantify the long-term trend of annual and seasonal extreme precipitation, and the Pettit mutation point analysis is applied to identify non-stationary points in the time series to reveal the period of change in the extreme precipitation time series. In addition, the bootstrap method is used to sample the changes in the significance level of the extreme precipitation time series to detect the non-stationary state of extreme precipitation.

[0035] S4: Analysis of the non-steady frequency of extreme precipitation due to urbanization; Combined with the non-steady stations analyzed in step S3, non-steady frequency analysis is performed to evaluate the non-steady state of extreme precipitation;

[0036] It should be noted that in step S4, the non-stationary frequency analysis uses 66 frequency analysis models based on the location, scale and shape of the generalized additive model, and only the stations that show non-stationary extreme precipitation use these 66 non-stationary models. The generalized additive model assumes that the independent observation variable Y i (i=1,2,3,…,n)Follow F y (Y i / θ i), where n is the total number of observed variables, θ i =(θ i1 ,θ i2 ,…,θ ip ) is a vector of p parameters containing the characteristics of the location, scale and shape random variables. k is the index of the distribution parameter. When k = 1, 2, ..., p, the distribution parameter is obtained through the monotonic link function g k (·) Associate the distribution parameters with the explanatory variables. The generalized additive model involves multiple models and uses the semiparametric additive formula as follows:

[0037]

[0038] In the formula, θ k is a parameter vector of length n; X k is an n×m-order explanatory variable matrix, β k is a parameter vector of length m, h jk Represents the distribution parameter for the explanatory variable x jk The functional dependency of x jk is the matrix X k A column.

[0039] S5: Causes of the non-steady-state impact of urbanization on extreme precipitation; combining the daily precipitation observation data and large-scale meteorological data obtained in step S1, analyzing the extreme precipitation changes caused by large-scale circulation or local changes;

[0040] It should be noted that in step S5, in order to further verify the cause of the non-steady-state impact of urbanization on extreme precipitation, the eight climate zones within China are first considered to calculate the latitudinal average precipitation in each region. Secondly, the six driving factors (temperature, relative humidity, air pressure, latitudinal wind, meridional wind and vertical wind speed) after standardization are considered in each region, and the dimension is reduced without reducing the variability of the original data set through principal component analysis. Third, the calculated principal components are used as input for cluster analysis through machine learning clustering algorithm. Finally, the Dunn index is used to optimize the total number of clusters. After completing the above steps, the time series of the latitudinal average precipitation in each region is divided into two equal 27 years (i.e., 1961-1987 and 1988-2014), and the quantiles (from 75% to 99%) of each group of climate zones are estimated respectively.

[0041] S6: Attribution analysis of the contribution of climate change and urbanization to extreme precipitation trends; Combine the rural, urbanized, and urban sites obtained in step S2 to clarify the individual contributions of climate change and urbanization to extreme precipitation trends at the national, regional, and local scales.

[0042] It should be noted that in step S6, the attribution analysis quantifies the individual contributions of climate change and urbanization to extreme precipitation trends at national, regional and local scales in China. Changes in precipitation in rural areas represent the impact of climate change and can be used as a reference to isolate precipitation changes caused by urbanization in urban areas. At the national scale, the contribution is calculated using a weighted aggregation method:

[0043] ConNclimate=(100-ConNurban)×100%

[0044]

[0045] Where, Con Nclimate and Con Nurban is the contribution of climate change and urbanization to precipitation trends at the national scale; S sub is the proportion of urban stations with positive or negative precipitation trends to all stations; P urban and P rural are precipitation trends in urban and rural areas, respectively; P N It is the average precipitation trend of all stations with positive or negative precipitation trends in China.

[0046] At the regional and local scales, climate change and urbanization jointly affect the precipitation trend at a station, and their contribution can be calculated as:

[0047]

[0048] Where, Con Rclimate and Con Rurban is the contribution of climate change and urbanization to precipitation trends at the regional or local scale; P urban and P rural are precipitation trends in urban and rural areas, respectively; P R It is the regional or local precipitation trend of a certain city station, which is positive or negative.

[0049] The embodiment of the present invention takes the daily precipitation observation data of 1857 stations in China from 1961 to 2014 as an example to further describe the technical solution of the method for attribution and impact assessment of urban precipitation in China at multiple spatiotemporal scales under the effects of climate change and urbanization. The embodiment is used to illustrate the present invention, but is not used to limit the scope of application of the present invention, and is also applicable to different regions or time periods.

[0050] The method of the present invention is an evaluation method for attribution and impact of precipitation at multiple spatiotemporal scales in China under the effects of climate change and urbanization. The implementation flow chart is as follows: Figure 1 As shown, the specific steps are as follows:

[0051] (1) Collection of basic data;

[0052] In this embodiment, daily precipitation observation data of 1857 stations in China from 1961 to 2014 were collected, of which 1336 stations had no missing data, and 445 stations had a data missing rate of less than 1%, accounting for 24% of the total. The missing precipitation values ​​of 1-2 days were filled with the average precipitation of the adjacent days, and the missing data of consecutive days were interpolated by the long-term average of the same number of days in other years; the 1km resolution land use coverage data set of 2015 from the Resource and Environmental Science Data Center of the Chinese Academy of Sciences was collected, including provincial and national cultivated land, forest land, grassland, water body, construction land and unused land; the population density data of 1995 and 2015 from the Resource and Environmental Science Data Center of the Chinese Academy of Sciences was collected; the large-scale meteorological data of the reanalysis data of the National Center for Environmental Prediction and the National Center for Atmospheric Research were collected, including temperature, humidity, air pressure, zonal wind, meridional wind and vertical wind speed, and the data set time covers 1961-2014. Table 1 is the selected data information:

[0053] Table 1 Main data information

[0054]

[0055] (2) Identification of urbanization level;

[0056] In this embodiment, the land use coverage data is divided into urban areas and non-urban areas. When the site is located in an urban area and the population density is greater than 1000 people / km 2 It is defined as an urban site. 1000 people / km 2 The population density threshold is defined as the 98.5th percentile value in China, and a total of 14.6% of the sites are determined as urban sites. With the urban site as the center, the sites within a radius of 50km are defined as urbanized sites, and a total of 20.6% of the sites are determined as urbanized sites. Other unassigned sites are defined as rural sites.

[0057] (3) Analysis of the impact of urbanization on the long-term trend of extreme precipitation;

[0058] In this example, based on the data of rural, urbanized and urban sites in China obtained in Example (2), extreme precipitation is defined according to the maximum annual and seasonal precipitation of each site (spring (March to May), summer (June to August), autumn (September to November), winter (December to February)). The non-parametric Mann-Kendall test is used to quantify the long-term trends of annual and seasonal extreme precipitation at different types of sites in China. Three types of trends are considered here: positive trend (increase at the 0.05 significance level), negative trend (decrease at the 0.05 significance level) and no trend (not significant at the 0.05 level). Figure 2As shown, the blue line (red line) represents the linear regression fitting line before (after) 1987. The blue (red) numbers represent the linear regression coefficients corresponding to the linear regression. The trends in extreme precipitation before and after 1987 have similar changes in rural, urbanized and urban areas. However, there is no significant difference in the slope of increase between rural, urbanized and urban areas, indicating that urbanization has limited impact on the increasing trend of annual and seasonal extreme precipitation after 1987. More than 90% of the stations showed no trend in annual and seasonal extreme precipitation in both the periods before and after 1987. No obvious spatial pattern of annual and seasonal extreme precipitation was observed in the stations with significant trends before 1987. The spatial variation of stations with significant trends in China after 1987 is small, especially in autumn and winter. The northern part of China is dominated by stations with increasing trends in autumn, while the southwestern and central parts of China are dominated by stations with decreasing trends in winter. In addition, Pettit's breakpoint analysis was applied to identify non-stationary points in time to reveal the period of change in extreme precipitation time series. The results showed that there was no significant change in the trend direction of annual and seasonal extreme precipitation at urban stations.

[0059] In this embodiment, the stationarity of the extreme precipitation time series is defined as "no trend, change or periodicity". Therefore, the extreme precipitation time series is sampled using the bootstrap method to estimate the change in the significance level to detect the non-stationarity of extreme precipitation. From the extreme precipitation time series from 1961 to 1987, the time series is resampled using the bootstrap method and fitted with the generalized extreme value distribution, thereby obtaining a set of estimated parameters of the generalized extreme value distribution. The above process is repeated 1000 times, thereby obtaining 1000 sets of estimated parameters from 1961 to 1987. Similarly, 1000 sets of estimated parameters can also be obtained for the extreme precipitation time series from 1988 to 2014. The results show that about 10% of the stations for annual and seasonal extreme precipitation show obvious non-stationarity, while about 20% of the stations in winter show non-stationarity. The percentage of rural, urbanized and urban stations in non-stationary stations is basically consistent with the percentage of rural, urbanized and urban stations in total stations (60%, 18% and 22% respectively). However, in North China and East China, two regions with high urbanization levels, urbanized and urban stations are prone to non-stationary states, especially in annual, spring, and summer extreme precipitation. This indicates that the non-stationary state of extreme precipitation is prominent at the regional scale.

[0060] (4) Analysis of the non-steady-state frequency of extreme precipitation due to urbanization;

[0061] In order to conduct non-stationary frequency analysis of extreme precipitation, 66 frequency analysis models based on location, scale and shape of the generalized additive model were used, and only the stations that showed non-stationary behavior in terms of extreme precipitation were used. The generalized additive model assumes that the independent observed variable Y i (i=1,2,3,…,n)Follow F y (Y i / θ i ), where n is the total number of observed variables, θ i =(θ i1 ,θ i2 ,…,θ ip ) is a vector of p parameters containing the characteristics of the location, scale and shape random variables. k is the index of the distribution parameter. When k = 1, 2, ..., p, the distribution parameter is obtained through the monotonic link function g k (·) Associate the distribution parameters with the explanatory variables. The generalized additive model involves multiple models and uses the semiparametric additive formula as follows:

[0062]

[0063] In the formula, θ k is a parameter vector of length n; X k is an n×m-order explanatory variable matrix, β k is a parameter vector of length m, h jk Represents the distribution parameter for the explanatory variable x jk The functional dependency of x jk is the matrix X k A column.

[0064] The 50-year return period of the non-stationary stations was estimated using a frequency analysis model based on location, scale, and shape of a generalized additive model and compared with the return period estimated by the steady-state model. The results show that among these non-stationary stations, more than 50% of the stations showed positive changes in annual and seasonal extreme precipitation, except in autumn. Within the 50-year return period, the difference percentages of the annual and seasonal steady-state and non-stationary frequency analysis were 25%, -23.5%, 22.9%, 19.9%, and 12% higher than those of the rural and urbanized stations, respectively. At the regional scale, the return periods of the steady-state and non-stationary frequencies of the zonal average extreme precipitation were analyzed. As Figure 3 As shown in Figure 2, the steady-state (non-steady-state) return period is represented by the blue (red) line. The latitudinal mean extreme precipitation return period of urbanized stations changes significantly and is greater than that of rural and urban stations. This feature is most obvious in annual extreme precipitation (see Figure 3 ac). These results suggest that urbanization (i.e., transition from rural to urban areas) can lead to non-stationary states in extreme precipitation.

[0065] (5) Causes of the non-steady-state impact of urbanization on extreme precipitation;

[0066] In this example, eight meteorologically homogeneous regions in China are first considered, and the latitudinal average precipitation in each region is calculated. Secondly, six large-scale daily driving factors (i.e., temperature, relative humidity, air pressure, zonal wind, meridional wind, and vertical wind speed) are considered after standardization in each region, and the dimension is reduced without reducing the variability of the original data set through principal component analysis. Third, the calculated principal components are used as input for cluster analysis through a machine learning clustering algorithm. Finally, the Dunn index is used to optimize the total number of clusters in the eight meteorologically homogeneous regions. It is observed that there are differences between the cluster centers of all variables, indicating that classification is of great significance for determining different circulation patterns. In order to clarify the impact of urbanization on extreme precipitation, considering the period of ERA-Interim reanalysis data, the time series of latitudinal average daily precipitation in each region is divided into two equal 27 years: 1979-1997 and 1998-2014, and the quantiles from 75% to 99% of the first and second clusters in each period are calculated (see Figure 4 , 5 ). Figure 4 In the figure, the blue line (red line) shows the quantiles during the period 1979-1997 (1998-2014); Figure 5 In the figure, the blue line (red line) represents the quantiles during the period 1979-1997 (1998-2014). Figure 4 , 5 The results in suggest that the difference in quantiles between the two periods may be caused by local urbanization, as the large-scale circulation patterns are controlled. Figure 4 As shown in Figure 1, in cluster 1, changes in the high quantiles (i.e., values ​​above the 90th quantile of daily precipitation) of urban and urbanized stations are only found in regions I, III, IV, and VII, while changes in the high quantile values ​​of rural stations in these regions are not prominent. Figure 5 As shown in the figure, in cluster 2, region I is the only region with a significant difference in high quantiles between the two periods, and this difference does not appear in rural sites, but in urbanized sites. In most regions, the high quantiles of rural, urbanized, and urban sites between the two periods change very little. It can be seen that when controlling for large-scale circulation patterns, changes in precipitation quantiles at urban sites are mainly related to the influence of local changes (such as urbanization).

[0067] (6) Attribution analysis of the contribution of climate change and urbanization to precipitation trends;

[0068] In this embodiment, based on the climate change index, 9 precipitation indices are defined, including: the maximum precipitation in one day per month (Rx1), the maximum precipitation in five consecutive days per month (Rx5), the simple precipitation intensity index (SDII), the annual total precipitation when the daily precipitation is >95p (R95pTOT), the annual total precipitation when the daily precipitation is >99p (R99pTOT), the annual total precipitation on wet days (PRCPTOT), the maximum dry period length, i.e., the maximum number of consecutive days with daily precipitation <1mm (CDD), the maximum wet period length, i.e., the maximum number of consecutive days with daily precipitation >1mm (CWD), and the number of days with precipitation ≥95p per year (EPE95p). Among them, "95p" and "99p" represent the 95th and 99th percentiles of the corresponding time series, respectively. Wet days refer to the number of days with daily precipitation greater than 1mm. Table 2 is the detailed information of the precipitation index used.

[0069] Table 2 Detailed information of precipitation index

[0070]

[0071]

[0072] The individual contributions of climate change and urbanization to the trends of nine precipitation indices at national, regional, and local scales in China were quantified through attribution analysis. Changes in precipitation in rural areas represent the impact of climate and can be used as a reference to isolate precipitation changes caused by urbanization in urban areas. At the national scale, the contribution was calculated using a weighted aggregation method:

[0073] ConNclimate=(100-ConNurban)×100%

[0074]

[0075] Where, Con Nclimate and Con Nurban is the contribution of climate change and urbanization to precipitation trends at the national scale; S sub is the proportion of urban stations with positive or negative precipitation trends to all stations; P urban and P rural are precipitation trends in urban and rural areas, respectively; P N It is the average precipitation trend of all stations with positive or negative precipitation trends in China.

[0076] At the regional and local scales, climate change and urbanization jointly affect the precipitation trend at a station, and their contribution can be calculated as:

[0077]

[0078] Where, Con Rclimate and Con Rurbanis the contribution of climate change and urbanization to precipitation trends at the regional or local scale; P urban and P rural are precipitation trends in urban and rural areas, respectively; P R It is the precipitation trend of a certain city station in the region or the local area, which is positive or negative. N and P R The contributions at the national, regional and local scales are calculated separately. For the local scale, the average trend of the nine precipitation indices at rural stations within a 1°×1° resolution grid is taken as the contribution of local climate change. Afterwards, the contribution of urbanization is extracted in the grid. Figure 6 As shown, at the national scale, the four extreme precipitation indices of Rx1, Rx5, R95pTOT, and R99pTOT have no or only weak urban signals. For these four extreme precipitation indices, urbanization contributes 0.1%, 5.3%, 0.5%, and 4.7% to the positive trends, respectively, while urbanization contributes -0.2%, -5.8%, -0.8%, and -0.6% to the negative trends, respectively. This shows that although the contribution of urbanization to extreme precipitation is negligible, it still makes a slight contribution to the increase (decrease) of extreme precipitation. Figure 7 As shown in the figure, at the regional scale, urbanization contributes 0.1%, 5.3%, 0.5% and 4.7% to the positive trends of four extreme precipitations, Rx1, Rx5, R95pTOT and R99pTOT, respectively, while urbanization contributes -0.7%, -5.8%, -0.7% and -0.6% to the negative trends, respectively. Climate change is still the dominant factor affecting the trend of urban extreme precipitation. However, the impact of urbanization on CWD and PRCPTOT is significant at the regional scale. Among them, the contribution of urbanization to the positive and negative trends of PRCPTOT increases to 20.19% and -30.63%, respectively. At the local scale, urbanization can slow down the decreasing trend of extreme precipitation and total precipitation in northern China, especially in the Beijing-Tianjin-Hebei urban agglomeration area; urbanization can enhance the increasing trend of extreme precipitation and total precipitation in eastern China, especially in the Yangtze River Delta urban agglomeration area. In addition, although the urbanization level in northwest China is low compared with other parts of China, urban characteristics are also found in the precipitation trend.

[0079] See also Figure 8 , Figure 8 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 multi-temporal and spatial scale urban precipitation attribution and impact evaluation device 401, a processor 402 and a storage medium 403.

[0080] A multi-temporal and spatial scale urban precipitation attribution and impact assessment device 401: The multi-temporal and spatial scale urban precipitation attribution and impact assessment device 401 implements the multi-temporal and spatial scale urban precipitation attribution and impact assessment method.

[0081] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the multi-temporal and spatial scale urban precipitation attribution and impact assessment method.

[0082] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the multi-temporal and spatial scale urban precipitation attribution and impact assessment method.

[0083] The beneficial effects of the present invention are:

[0084] (1) The impact of urbanization on precipitation in the present invention is a multi-temporal and spatial scale analysis, which can comprehensively analyze the trend of extreme precipitation from different time scales and spatial scales (time: annual and seasonal; space: national, regional and local). The impact of urbanization on precipitation varies with the change of research area and time, and the results are extensive and generalizable.

[0085] (2) At present, there are many studies on the impact of urbanization on precipitation, but the research on the impact of urbanization on precipitation at multiple spatiotemporal scales in China is very limited. In addition, due to the complex interactions between climate change and urbanization, their respective contributions to extreme precipitation changes are still unclear. The proposed method helps to understand the impact and attribution of urbanization on precipitation patterns at multiple spatiotemporal scales, and provides some more valuable references and methods for urban planning and climate adaptation strategies in my country.

[0086] 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 multi-temporal and spatial scale urban precipitation attribution and impact assessment method, characterized by: The method comprises the following steps: S1: Data collection: collect daily precipitation observation data, land use coverage data, population density data and large-scale meteorological data; S2: Identification of urbanization degree: combining daily precipitation observation data, land use coverage data and population density data obtained in step S1 to identify rural, urbanized and urban sites in China; S3: Analysis of the impact of urbanization on the long-term trend of extreme precipitation; combining the rural, urbanized and urban sites obtained in step S2, quantifying the annual and seasonal trends of extreme precipitation, conducting mutation point analysis, and detecting the non-steady state of extreme precipitation; S4: Analysis of the non-steady frequency of extreme precipitation due to urbanization; Combined with the non-steady stations analyzed in step S3, non-steady frequency analysis is performed to evaluate the non-steady state of extreme precipitation; S5: Causes of the non-steady-state impact of urbanization on extreme precipitation; combining the daily precipitation observation data and large-scale meteorological data obtained in step S1, analyzing the extreme precipitation changes caused by large-scale circulation or local changes; S6: Attribution analysis of the contribution of climate change and urbanization to extreme precipitation trends; Combine the rural, urbanized, and urban sites obtained in step S2 to clarify the individual contributions of climate change and urbanization to extreme precipitation trends at the national, regional, and local scales.

2. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: The daily precipitation observation data is daily precipitation records covering multiple stations over many years; the land use coverage data includes surface type data representing the city; the population density data is data with a spatial resolution of 1km×1km; the large-scale meteorological data includes temperature, humidity, air pressure, zonal wind, meridional wind and vertical wind speed.

3. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: According to the daily precipitation observation data, land use data and population density data obtained in step S1, when the site is located in an urban area and the population density is greater than a preset value, it is defined as an urban site; the site within a preset radius centered on the urban site is defined as an urbanized site, and other unassigned sites are defined as rural sites.

4. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: In step S3, extreme precipitation is defined as the maximum annual and seasonal precipitation at each station; the nonparametric Mann-Kendall test is used to quantify the long-term trends of annual and seasonal extreme precipitation, and the Pettit mutation point analysis is applied to identify non-stationary points in the time series to reveal the period of change in the extreme precipitation time series; the bootstrap method is used to sample the time series of extreme precipitation to estimate the changes in the significance level, so as to detect the non-stationary state of extreme precipitation.

5. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: In step S4, the non-stationary frequency analysis uses 66 frequency analysis models based on the location, scale and shape of the generalized additive model; the generalized additive model assumes that the independent observed variable Y i (i=1,2,3,…,n)Follow F y (Y i / θ i ), where n is the total number of observed variables, θ i =(θ i1 ,θ i2 ,…,θ ip ) is a vector of p parameters containing the characteristics of the random variables of location, scale and shape; k is the index of the distribution parameter, when k = 1, 2, ..., p, the distribution parameter is obtained through the monotonic link function g k (·) Associate the distribution parameters with the explanatory variables; the generalized additive model involves multiple models and uses the semiparametric additive formula as follows: In the formula, θ k is a parameter vector of length n; X k is an n×m-order explanatory variable matrix, β k is a parameter vector of length m, h jk Represents the distribution parameter for the explanatory variable x jk The functional dependency of x jk is the matrix X k A column.

6. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: Step S5 is specifically as follows: S51. Considering the eight climate zones in China, calculate the latitudinal average precipitation in each zone; S52. Six driving factors were considered in each region after standardization, including air temperature, relative humidity, air pressure, zonal wind, meridional wind and vertical wind speed, and principal component analysis was performed to reduce the dimensionality without reducing the variability of the original dataset. S53, using the calculated principal components as input, and performing cluster analysis through a machine learning clustering algorithm; S54, use Dunn's index to optimize the total number of clusters; S55. Divide the time series of latitudinal mean precipitation in each region into two equal periods of 27 years and estimate the quantiles for each group of climate zones.

7. The multi-temporal and spatial scale urban precipitation attribution and impact assessment method according to claim 1, characterized in that: In step S6, the contribution is calculated at the national scale using the weighted aggregation method: CoNclimate=(100-ConNurban)×100% Where, Con Nclimate and Con Nurban is the contribution of climate change and urbanization to precipitation trends at the national scale; S sub is the proportion of urban stations with positive or negative precipitation trends to all stations; P urban and P rural They are precipitation trends in urban and rural areas, respectively; P N is the average precipitation trend of all stations with positive or negative precipitation trends in China; At the regional and local scales, climate change and urbanization jointly affect the precipitation trend at a station, and their contribution is calculated as: Where, Con Rclimate and Con Rurban is the contribution of climate change and urbanization to precipitation trends at regional or local scales; P urban and P rural are precipitation trends in urban and rural areas, respectively; P R It is the regional or local precipitation trend of a certain city station, which is positive or negative.

8. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a multi-temporal and spatial scale urban precipitation attribution and impact assessment method as described in any one of claims 1 to 7.

9. A multi-temporal and spatial scale urban precipitation attribution and impact assessment device, characterized by: include: Processor and storage medium; the processor loads and executes instructions and data in the storage medium to implement a multi-temporal and spatial scale urban precipitation attribution and impact assessment method as described in any one of claims 1 to 7.

Images