Characterization method of urban inner lake ecosystem resilience based on watershed load-lake ecological model

Abstract

The present invention discloses a method for characterizing the resilience of an urban inner lake ecosystem based on a watershed load-lake ecological model. The method comprises: constructing a WET model and coupling it with a watershed load model; selecting a reference year based on a long-term series of abiotic and biotic state variables of an urban inner lake obtained by the WET model; calculating an index characterizing the resilience of an urban inner lake ecosystem using the biomass and energy transfer efficiency output by the WET model; establishing a response relationship between the watershed environment and the resilience of an urban inner lake ecosystem, and between the urban inner lake habitat and the resilience of an urban inner lake ecosystem, and analyzing and judging the effectiveness of the index; taking the 25% quantile of the total score of the resilience index of the reference year as the standard, and then dividing the distribution range less than the 25% quantile into four equal parts, and defining them in order from high to low as good resilience, relatively good resilience, average resilience, relatively poor resilience, and poor resilience; the present invention can use numerical simulation to achieve a quantitative assessment of the resilience level of an urban inner lake ecosystem.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological river and lake assessment, and in particular to a method for characterizing the resilience of an urban inner lake ecosystem based on a watershed load-lake ecological model. Background Art

[0002] Urban lakes are an integral part of the urban water system, and they undertake multiple functions such as flood storage, water supply, ecology, and landscape. However, due to global climate change and intensified human activities, the habitats of urban lakes have been severely degraded and their ecological functions have been damaged, leading to water pollution and reduced biodiversity. The resilience of urban lake ecosystems will largely affect the health of lake ecosystems. Resilience is defined as "the ability of an ecosystem to return to a stable state after being disturbed", emphasizing that social-ecological systems do not just return to their original state after being impacted, but become more flexible and adaptable. Current research on lake ecosystem resilience mainly focuses on concepts, measurements, and threshold identification. Due to the different evolutionary processes, nutrient status, and physical states of different lakes, the resilience of lake ecosystems is characterized by complexity and unpredictability.

[0003] Traditional indicators for characterizing the resilience of lake ecosystems focus on nutrients, chlorophyll a, submerged plant biomass, phytoplankton or benthic animal diversity, ecosystem integrity, etc. Single indicators are often limited by regional characteristics and have low applicability, while overly complex indicators are not sensitive enough to disturbances. The application of such characterization indicators will not be able to effectively reflect the health of lake ecosystems, which will lead to the failure of governance and restoration to restore lakes to an ideal state, and the phenomenon of "repeated governance, repeated governance" is prominent. Food webs provide a basis for analyzing dynamic feedback at the ecosystem level by simplifying real ecosystems and characterizing the interaction relationship between species in ecosystems. When external abiotic conditions change, food webs will change through food web structure and interactions between different trophic levels. Studies in recent years have shown that mastering the characteristics of lake food webs and identifying their change thresholds are of great significance to the protection and restoration of lake ecosystems. Therefore, focusing on the problems of habitat destruction and unclear health status of urban inner lakes, establishing indicators that effectively characterize the resilience of urban inner lake ecosystems is a key means to build a green, safe and resilient sustainable urban water system. Therefore, it is necessary to propose a method for characterizing the resilience of urban inner lake ecosystems based on the watershed load-lake ecological model to solve the above problems. Summary of the invention

[0004] The present invention provides a method for characterizing the resilience of urban inner lake ecosystems based on a watershed load-lake ecological model, so as to solve the problem that traditional indicators for characterizing the resilience of lake ecosystems are limited by regional characteristics and have low applicability, while overly complex indicators are not sensitive enough to interference.

[0005] The present invention provides a method for characterizing the resilience of an urban inner lake ecosystem based on a watershed load-lake ecological model, comprising the following steps:

[0006] Step 1: Construct a WET model based on the two-level steady-state transformation theory of lake eutrophication and ecological restoration process, and couple it with the watershed load model to simulate the nutrient cycle and food web dynamics of the urban inner lake under the external load of the lake;

[0007] Step 2: Based on the long-term series of abiotic and biotic state variables of urban lakes obtained by the WET model, the reference year was selected according to the human activity disturbance status, habitat status, and food web structure stability;

[0008] Step 3: Using the biomass and energy transfer efficiency output by the WET model, calculate the interaction strengths within and between species. The geometric mean of the absolute values ​​of each interaction strength is the maximum weight of the ring in the food web, which is used as an indicator to characterize the resilience of the urban lake ecosystem.

[0009] Step 4: Through geographic detector analysis and global sensitivity analysis methods, the response relationships between the watershed environment and the resilience of the urban inner lake ecosystem, and between the urban inner lake habitat and the resilience of the urban inner lake ecosystem are established respectively, and the sensitivity of the resilience indicators to environmental disturbances is analyzed; and the resilience indicators are analyzed by drawing box plots according to the reference year and the damaged year, and the overlap of the box plots of the reference year and the damaged year is checked to determine the validity of the indicators;

[0010] Step five, taking the 25% quantile of the total score of the resilience index in the reference year as the standard, divide the distribution range below the 25% quantile into four equal parts, and define them in order from high to low as good resilience, relatively good resilience, average resilience, relatively poor resilience, and poor resilience.

[0011] Furthermore, step one includes the following steps:

[0012] The watershed load model is used to estimate the exogenous nutrient load input to the WET model; the upstream catchment area of ​​the urban inner lake is obtained based on DEM and water system; the input data of the watershed load model include meteorological, land use, industrial point source data, and agricultural farming data; the flow and nutrient load output by the watershed load model are used as the boundary conditions of the WET model;

[0013] The WET model is used to simulate the nutrient cycle and food web dynamics of urban lakes. The input data of the WET model include lake characteristic data, sediment characteristic data and external driving data. The lake characteristic data are obtained through field surveys and data collection, and the lake characteristic data include lake water depth, area, and vegetation coverage. The lake sediment characteristic data are derived from the default sediment parameters set in the WET model. The external driving data include time series data of hydrology, meteorology, and exogenous nutrient loads. The WET model is calibrated using measured water quality and water ecological indicators, and the measured water quality and water ecological indicators include total nitrogen, total phosphorus, transparency, dissolved oxygen, chlorophyll a, and plankton biomass.

[0014] Furthermore, step 2 includes the following steps:

[0015] The Mann-Kendall mutation analysis method is used to determine the time point when the habitat quality of urban lakes drops sharply, including: sorting long time series data to obtain the sorted data sequence; calculating the statistic sequence UF and the reverse statistic sequence UB according to the sorted data sequence; given the significance level α, the two statistic curves UF and UB and the significance level line are plotted on the same graph; if the values ​​of UF and UB are greater than 0, it indicates that the sequence is on an upward trend, and if they are less than 0, it indicates that the sequence is on a downward trend; when it exceeds the critical straight line, it indicates that the upward or downward trend is significant, and the range exceeding the critical line is determined as the time area of ​​the mutation; if the two curves UF and UB have an intersection, and the intersection is between the critical lines, then the moment corresponding to the intersection is the time when the mutation begins.

[0016] Furthermore, step three includes the following steps:

[0017] The biomass and energy transfer efficiency were obtained from the WET model; the resilience index of the urban inner lake ecosystem was derived from the eigenvalues ​​or diagonal values ​​of the Jacobian matrix; the off-diagonal values ​​were the intensity of interspecific interactions, indicating the influence of trophic level j on trophic level i, and the diagonal values ​​were the intensity of intraspecific interactions; the interaction intensity value was obtained by calculating the partial derivatives of the LV equation of each trophic level at equilibrium;

[0018] The LV equation describes the biomass changes at each trophic level, and its equation is as follows:

[0019]

[0020] In the formula, and are the biomass of species i and j, respectively; is positive or negative, representing the growth rate or specific mortality rate, respectively; is the interaction strength between species i and j;

[0021] At equilibrium, the energy transfer efficiency from trophic level i to trophic level j is F ij It is expressed as:

[0022]

[0023] In the formula, and are the biomass at trophic level i and j in equilibrium, and the average biomass is and to replace;

[0024] Taking trophic level j as the predator and trophic level i as the prey, the interaction intensity between the predator and the prey is The calculation formula is:

[0025]

[0026] Interaction strength between prey and predator The calculation formula is:

[0027]

[0028] Intraspecific interaction strength The calculation formula is:

[0029]

[0030] In the formula, is the energy transfer efficiency at trophic level j; s is the specific mortality rate caused by intraspecific competition When the value of s is between 0 and 1, the ecosystem is stable; if s is greater than 1, the ecosystem is considered unstable; when comparing the stability of multiple ecosystems, the smaller s is, the more stable the system is;

[0031] The maximum weight of the ring in the food web is used as an indicator of the resilience of the urban lake ecosystem. The lower the value, the better the resilience. The length of the ring K is the number of trophic groups in the ring, and the weight of the ring is the geometric mean of the absolute values ​​of the action intensity in the ring, and the calculation formula is as follows:

[0032]

[0033] Rings with K ≥ 3 are stable, so the weight of all rings with a length of 3 is calculated; the maximum ring weight is approximately the maximum s that satisfies the stability of the community;

[0034] By comparing the difference in food web stability before and after removing a certain species, the impact of the species on the stability of the food web is analyzed; if the difference is negative, it means that removing the corresponding species from the Jacobian matrix makes the ecosystem more stable; if the difference is positive, the ecosystem becomes unstable; this species that is at the key link of the food web and supplements the missing key link in the food chain is defined as an indicator species.

[0035] Furthermore, step 4 includes the following steps:

[0036] The geographic detector analysis method is used to determine whether the constructed urban lake ecosystem resilience index is sensitive to the response of various environmental factors in the basin; based on the following assumptions: if an independent variable has an important impact on the dependent variable, then the spatial distribution of the two should have a certain similarity. The calculation formula is as follows:

[0037]

[0038] In the formula, SST is the total variance of the whole region, SSW is the intra-layer variance and q is the degree of explanation of the independent factor on the dependent variable, q The value range is 0~1. q The larger the value, the more obvious the spatial heterogeneity;

[0039] A global sensitivity analysis of the WET model was conducted to determine whether the constructed urban lake ecosystem resilience index was sensitive to the response of the urban lake habitat factors; the parallel sensitivity and automatic calibration tools included the density-based Borgonovo method for screening and the variance-based Sobol method for in-depth analysis; the importance of the sensitivity parameter ranking δ ranged from 0 to 1, and the larger the value, the more sensitive the parameter;

[0040] The resilience index of urban inner lake ecosystem was analyzed by drawing box plots according to the reference year and the damaged year, and the overlap of the box plots of the reference year and the damaged year was checked; if the box plots of the reference year and the damaged year have no overlap or partially overlap but the second quartile of both are not within the other box, it means that the construction of the resilience index of urban inner lake ecosystem is effective.

[0041] The present invention has the following beneficial effects: the method for characterizing the resilience of an urban inner lake ecosystem based on a watershed load-lake ecological model of the present invention utilizes the watershed load-lake ecological model to simulate the nutrient cycle and food web dynamics of an urban inner lake under an external lake load, and utilizes the model output parameters to calculate the maximum weight value of a ring in the food web, which is used as an indicator for characterizing the resilience of a lake ecosystem, thereby providing theoretical and technical support for the protection of an urban inner lake ecosystem and habitat restoration, and effectively solving the problem that traditional indicators for characterizing the resilience of a lake ecosystem are limited by regional characteristics and have low applicability, while overly complex indicators are insufficiently sensitive to interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 It is a flow chart of the method for characterizing the resilience of the urban inner lake ecosystem based on the watershed load-lake ecological model of the present invention;

[0044] Figure 2 is a schematic diagram of a mutation point test based on a long-term series of total phosphorus inflow into a lake in an embodiment of the present invention;

[0045] Figure 3 is a schematic diagram of box plot overlap between the reference year and the damaged year in an embodiment of the present invention;

[0046] Figure 4 It is a schematic diagram of the resilience evaluation results of the urban inner lake ecosystem in the current year in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0048] See also Figure 1 The present invention provides a method for characterizing the resilience of an urban inner lake ecosystem based on a watershed load-lake ecological model, comprising the following steps:

[0049] Step 1: Construct a WET model based on the two-state steady-state transformation theory of lake eutrophication and ecological restoration process, and couple it with the watershed load model to simulate the nutrient cycle and food web dynamics of urban lakes under external inflow load.

[0050] This embodiment selects a lake in a city as the research object. The watershed load model is used to estimate the exogenous nutrient load input into the lake ecology (WET) model. The upstream catchment area of ​​the urban lake is obtained based on the digital elevation model (DEM) and the water system. The input data of the watershed load model consists of meteorological, land use, industrial point source data, and agricultural farming data. The flow and nutrient load output by the watershed load model are used as the boundary conditions of the WET model.

[0051] The WET model is used to simulate the nutrient cycle and food web dynamics of urban lakes. The input data of the WET model consists of lake characteristics, sediment characteristics and external driving data. Lake characteristic data such as lake depth, area, and vegetation coverage are obtained through field surveys and data collection; lake sediment characteristic data are derived from the default sediment parameters set in the WET model; external driving data mainly include time series data of hydrology, meteorology, and exogenous nutrient load. The WET model is calibrated using the measured water quality and water ecological indicators - total nitrogen, total phosphorus, transparency, dissolved oxygen, chlorophyll a and plankton biomass. Plankton includes diatoms, cyanobacteria and zooplankton.

[0052] The calibration order of the WET model is generally abiotic processes first and then biotic processes, and the one-in-one method is used to determine the value of the sensitivity parameter. R 2 )、Relative error( RE ) and the normalized root mean square error ( RMSE ) to evaluate the simulation results. The evaluation results show that: R 2 =0.34, RE =0.25, RMSE =0.05; total nitrogen R 2 =0.52, RE =0.41, RMSE =0.03; total nitrogen R 2 =0.52, RE =0.41, RMSE =0.03; diatom biomass R 2 =0.02, RE =0.66, RMSE =0.44; zooplankton biomass R2 =0.83, RE =1.61, RMSE =0.18, the simulation results are within the acceptable range.

[0053] Step 2: Based on the long-term series of abiotic and biotic state variables of urban lakes obtained by the WET model, the reference year is selected according to the human activity disturbance status, habitat status and food web structure stability.

[0054] Based on the long-term series of abiotic and biotic state variables of urban inner lakes obtained by the WET model, the years with less human interference, habitats close to natural conditions, and stable food web structures are selected as reference years. The Mann-Kendall mutation analysis method is used to determine the time point when the habitat quality of urban inner lakes drops sharply, which is often the time point when human activities are strong, such as reclamation of lakes, aquaculture, and direct discharge of sewage. This step specifically includes: sorting the long-term series data to obtain the sorted data sequence; calculating the statistic sequence UF and the reverse statistic sequence UB based on the sorted data sequence; given the significance level α, the two statistic curves UF and UB and the significance level line are plotted on the same graph; if the values ​​of UF and UB are greater than 0, it indicates that the sequence is on an upward trend, and if they are less than 0, it indicates that the sequence is on a downward trend; when it exceeds the critical straight line, it indicates that the upward or downward trend is significant, and the range exceeding the critical line is determined as the time area of ​​the mutation; if the two curves of UF and UB have an intersection, and the intersection is between the critical lines, then the moment corresponding to the intersection is the time when the mutation begins.

[0055] In this step, the driving factors of the sudden change in the habitat quality of shallow lakes are relatively complex. Usually, nitrogen and phosphorus nutrients are the determining factors of water quality mutations, and there is an obvious threshold phenomenon between nitrogen and phosphorus loads and aquatic plants and algae. Therefore, a Mann-Kendall mutation analysis was performed based on a long-term series of total phosphorus loads entering the lake, with 2001 to 2003 as the reference years.

[0056] Step three, using the biomass and energy transfer efficiency output by the WET model, calculate the interaction strength within and between species. The geometric mean of the absolute values ​​of each interaction strength is the maximum weight value of the ring in the food web, which is used as an indicator to characterize the resilience of the urban lake ecosystem.

[0057] Specifically, parameters such as biomass and energy transfer efficiency are obtained from the WET model. The resilience index of the urban lake ecosystem is derived from the eigenvalues ​​or diagonal values ​​of the Jacobian matrix. Among them, the off-diagonal values ​​are the intensity of interspecific interactions, which represent the influence of trophic level j on trophic level i, and the diagonal values ​​are the intensity of intraspecific interactions. The interaction intensity value is obtained by calculating the partial derivatives of the Lotka-Volterra (LV) equation for each trophic level at equilibrium.

[0058] The LV equation describes the biomass changes at each trophic level, and its equation is as follows:

[0059]

[0060] In the formula, and Species i and j Biomass (gm -2 ); Can be positive or negative, representing the growth rate or specific mortality rate (yr -1 ); For species i and j The interaction strength between -2 ) -1 yr -1 ).

[0061] In equilibrium, from the trophic level i To trophic level j Energy transfer efficiency F ij It can be expressed as:

[0062]

[0063] In the formula, and Trophic level i and j Biomass at equilibrium (gm -2 ), usually expressed as average biomass and to replace.

[0064] Taking trophic level j as the predator and trophic level i as the prey, the interaction intensity between the predator and the prey is The calculation formula is:

[0065]

[0066] Interaction strength between prey and predator The calculation formula is:

[0067]

[0068] Intraspecific interaction strength The calculation formula is:

[0069]

[0070] In the formula, Trophic level jEnergy transfer efficiency; s Specific mortality due to intraspecific competition The proportion of s When the value range of is between 0 and 1, the ecosystem is stable; if s If it is greater than 1, the ecosystem is considered unstable. When comparing the stability of multiple ecosystems, s The smaller it is, the more stable the system is.

[0071] The maximum weight of the ring in the food web can be used as an indicator of the resilience of the urban lake ecosystem, and the lower the value, the better the resilience. K is the number of trophic groups in the ring, and the weight of the ring is the geometric mean of the absolute values ​​of the action intensity in the ring, and the calculation formula is as follows:

[0072]

[0073] K Rings ≥3 are stable, so the weight of all rings of length 3 is calculated; the maximum ring weight can be approximated as the maximum s that satisfies community stability.

[0074] By comparing the difference in food web stability before and after removing a certain species, the impact of the species on the stability of the food web is analyzed. If the difference is negative, it means that removing the corresponding species from the Jacobian matrix makes the ecosystem more stable; if the difference is positive, the ecosystem becomes unstable. Therefore, the present invention defines such species that are in the key link of the food web and supplement the missing key link in the food chain as indicator species.

[0075] Through calculation, it was found that the relative changes in food web stability were 11.9% and -48.7% respectively when diatoms and zooplankton were removed, which far exceeded the change rate of removing other species. Therefore, diatoms and zooplankton are indicator species of a certain lake. When the nutrient load entering the lake is close to the above mutation point, the weight of the loop of zooplankton→cyanobacteria, green algae and diatom→debris is also the largest, indicating that the system is in an unstable state. Figure 2 It is a schematic diagram of the mutation point test based on the long-term series of total phosphorus inflow into the lake in an embodiment of the present invention.

[0076] Step 4: Through geographic detector analysis and global sensitivity analysis methods, the response relationships between the watershed environment and the resilience of the urban lake ecosystem, and between the urban lake habitat and the resilience of the urban lake ecosystem are established respectively, and whether the resilience indicators are sensitive to environmental disturbances is analyzed; and the resilience indicators are analyzed by drawing box plots according to the reference year and the damaged year, checking the overlap of the box plots of the reference year and the damaged year, and judging the validity of the indicators.

[0077] Specifically, the geographic detector analysis method is used to determine whether the constructed urban lake ecosystem resilience index is sensitive to the response of various environmental factors in the basin. The core idea is based on the following assumption: if an independent variable has an important impact on the dependent variable, then the spatial distribution of the two should have a certain similarity. The calculation formula is as follows:

[0078]

[0079] Where SST is the total variance of the whole region, SSW is the sum of variance within the layer, and q is the degree of explanation of the independent factor on the dependent variable, ranging from 0 to 1. The larger the value, the more obvious the spatial heterogeneity.

[0080] A global sensitivity analysis of the WET model was performed to determine whether the constructed urban lake ecosystem resilience index was sensitive to the response of urban lake habitat factors. The parallel sensitivity and automatic calibration tool (parsac) includes screening using the density-based Borgonovo method and in-depth analysis using the variance-based Sobol method. The Borgonovo method has been proven to be effective in ranking sensitivity parameters, and its output results will be re-tested using the Sobol method. The importance of the sensitivity parameter ranking δ ranges from 0 to 1, and the larger the value, the more sensitive the parameter.

[0081] The results of factor analysis show that the influence intensity of the driving factors of ecological resilience of a lake varies in different periods. The changing trends of exogenous nutrient load, chlorophyll a concentration in water bodies, sediment phosphorus release flux and submerged plant coverage have become the dominant driving forces for the changes in the maximum weight of the food web, diatom and zooplankton community structure in the study area. The interaction between factors is mainly nonlinear enhancement, and the degree of influence shows an increasing trend year by year, as shown in Table 1.

[0082] Table 1 Detection results of ecological resilience change factors in the study area

[0083]

[0084] In addition, the urban inner lake ecosystem resilience index was analyzed by plotting box plots according to the reference year and the damaged year, and the overlap of the box plots of the reference year and the damaged year was checked. If the box plots of the reference year and the damaged year have no overlap at all (IQ = 3) or partially overlap but the second quartiles of both are not within the box of the other party (IQ = 2), it means that the construction of the urban inner lake ecosystem resilience index is effective. In this embodiment, if Figure 3 As shown, the box plots of the reference year and the damaged year partially overlap, but the second quartiles of both are not within the other box (IQ = 2), so the construction of the urban lake ecosystem resilience index is effective.

[0085] Step five, taking the 25% quantile of the total score of the resilience index in the reference year as the standard, divide the distribution range below the 25% quantile into four equal parts, and define them in order from high to low as good resilience, relatively good resilience, average resilience, relatively poor resilience, and poor resilience.

[0086] Specifically, taking the 25% quantile of the total score of the toughness index in the reference year as the standard, the distribution range below the 25% quantile was divided into four equal parts and ranked in order from high to low as good toughness (<10), relatively good toughness (10-15), average toughness (15-20), relatively poor toughness (20-25), and poor toughness (>25).

[0087] In this step, the results of the current year ecosystem resilience assessment for a lake are shown in Figure 4 The areas with better resilience of the ecosystem of a certain lake are distributed in the middle of West Lake and the north-central area of ​​East Lake. The sensitive areas with poor resilience are the southwest corner of East Lake, the southwest of the west of West Lake and the northwest bay. Among them, the southwest area of ​​East Lake has the lowest evaluation score and needs special attention.

[0088] Technologies not described in detail in the present invention are existing technologies, such as model construction, model calibration, model sensitivity analysis, aquatic organism identification methods, etc., and will not be described in detail.

[0089] The above-described embodiments of the present invention do not limit the protection scope of the present invention.

Claims

1. A method for characterizing the resilience of urban inner lake ecosystems based on a watershed load-lake ecological model, characterized in that: The steps include: Step 1: Construct a WET model based on the two-level steady-state transformation theory of lake eutrophication and ecological restoration process, and couple it with the watershed load model to simulate the nutrient cycle and food web dynamics of the urban inner lake under the external load of the lake; The specific step one includes the following steps: using the watershed load model to estimate the exogenous nutrient load input to the WET model; obtaining the upstream catchment area of ​​the urban inner lake based on DEM and water system; the input data of the watershed load model include meteorological, land use, industrial point source data, and agricultural farming data; The flow and nutrient load output by the watershed load model are used as boundary conditions of the WET model; The WET model is used to simulate the nutrient cycle and food web dynamics of urban lakes. The input data of the WET model include lake characteristic data, sediment characteristic data and external driving data. The lake characteristic data are obtained through field surveys and data collection. The lake characteristic data include lake water depth, area, and vegetation coverage. The lake sediment characteristic data are derived from the default sediment parameters set in the WET model. The external driving data include time series data of hydrology, meteorology, and exogenous nutrient load. The WET model is calibrated using measured water quality and water ecological indicators, including total nitrogen, total phosphorus, transparency, dissolved oxygen, chlorophyll a, and plankton biomass. Step 2: Based on the long-term series of abiotic and biotic state variables of urban lakes obtained by the WET model, the reference year was selected according to the human activity disturbance status, habitat status, and food web structure stability; The specific step 2 includes the following steps: using the Mann-Kendall mutation analysis method to determine the time point when the habitat quality of the urban inner lake drops sharply, including: sorting the long time series data to obtain the sorted data sequence; calculating the statistic sequence UF and the reverse statistic sequence UB according to the sorted data sequence; given the significance level α, the two statistic curves UF and UB and the significance level line are plotted on the same graph; if the values ​​of UF and UB are greater than 0, it indicates that the sequence is on an upward trend, and if they are less than 0, it indicates that the sequence is on a downward trend; when it exceeds the critical straight line, it indicates that the upward or downward trend is significant, and the range exceeding the critical line is determined as the time area of ​​the mutation; if the two curves UF and UB have an intersection, and the intersection is between the critical lines, then the moment corresponding to the intersection is the time when the mutation starts; Step 3: Using the biomass and energy transfer efficiency output by the WET model, calculate the interaction strengths within and between species. The geometric mean of the absolute values ​​of each interaction strength is the maximum weight of the ring in the food web, which is used as an indicator to characterize the resilience of the urban lake ecosystem. The specific step three includes the following steps: obtaining biomass and energy transfer efficiency from the WET model; deriving the urban inner lake ecosystem resilience index from the eigenvalues ​​or diagonal values ​​of the Jacobian matrix; where the off-diagonal values ​​are the intensity of interspecific interactions, representing the influence of trophic level j on trophic level i, and the diagonal values ​​are the intensity of intraspecific interactions; the interaction intensity value is obtained by calculating the partial derivatives of the LV equation of each trophic level at equilibrium; The LV equation describes the biomass changes at each trophic level, and its equation is as follows: In the formula, and are the biomass of species i and j, respectively; is positive or negative, representing the growth rate or specific mortality rate, respectively; is the interaction strength between species i and j; At equilibrium, the energy transfer efficiency from trophic level i to trophic level j is F ij It is expressed as: In the formula, and are the biomass at trophic level i and j in equilibrium, and the average biomass is and to replace; Taking trophic level j as the predator and trophic level i as the prey, the interaction intensity between the predator and the prey is The calculation formula is: Interaction strength between prey and predator The calculation formula is: Intraspecific interaction strength The calculation formula is: In the formula, is the energy transfer efficiency at trophic level j; s is the specific mortality rate caused by intraspecific competition When the value of s is between 0 and 1, the ecosystem is stable; if s is greater than 1, the ecosystem is considered unstable; when comparing the stability of multiple ecosystems, the smaller s is, the more stable the system is; The maximum weight of the ring in the food web is used as an indicator of the resilience of the urban lake ecosystem. The lower the value, the better the resilience. The length of the ring K is the number of trophic groups in the ring, and the weight of the ring is the geometric mean of the absolute values ​​of the action intensity in the ring, and the calculation formula is as follows: Rings with K ≥ 3 are stable, so the weight of all rings with a length of 3 is calculated; the maximum ring weight is approximately the maximum s that satisfies the stability of the community; By comparing the difference in food web stability before and after removing a certain species, the impact of the species on the stability of the food web is analyzed; if the difference is negative, it means that removing the corresponding species from the Jacobian matrix makes the ecosystem more stable; if the difference is positive, the ecosystem becomes unstable; this species that is at the key link of the food web and supplements the missing key link in the food chain is defined as an indicator species; Step 4: Through geographic detector analysis and global sensitivity analysis methods, the response relationships between the watershed environment and the resilience of the urban inner lake ecosystem, and between the urban inner lake habitat and the resilience of the urban inner lake ecosystem are established respectively, and the sensitivity of the resilience indicators to environmental disturbances is analyzed; and the resilience indicators are analyzed by drawing box plots according to the reference year and the damaged year, and the overlap of the box plots of the reference year and the damaged year is checked to determine the validity of the indicators; The specific step four includes the following steps: using the geographic detector analysis method to determine whether the constructed urban lake ecosystem resilience index is sensitive to the response of various environmental factors in the basin; based on the following assumptions: if an independent variable has an important impact on the dependent variable, then the spatial distribution of the two should have a certain similarity, and the calculation formula is as follows: In the formula, SST is the total variance of the whole region, SSW is the intra-layer variance and q is the degree of explanation of the independent factor on the dependent variable, q The value range is 0~1. q The larger the value, the more obvious the spatial heterogeneity; A global sensitivity analysis of the WET model was conducted to determine whether the constructed urban lake ecosystem resilience index was sensitive to the response of the urban lake habitat factors; the parallel sensitivity and automatic calibration tools included the density-based Borgonovo method for screening and the variance-based Sobol method for in-depth analysis; the importance of the sensitivity parameter ranking δ ranged from 0 to 1, and the larger the value, the more sensitive the parameter; The urban inner lake ecosystem resilience index was analyzed by drawing box plots according to the reference year and the damaged year, and the overlap of the box plots of the reference year and the damaged year was checked; if the box plots of the reference year and the damaged year had no overlap or partially overlapped but the second quartiles of both were not within the box of the other party, it indicated that the construction of the urban inner lake ecosystem resilience index was effective; Step five, taking the 25% quantile of the total score of the resilience index in the reference year as the standard, divide the distribution range below the 25% quantile into four equal parts, and define them in order from high to low as good resilience, relatively good resilience, average resilience, relatively poor resilience, and poor resilience.

Images