Ecological risk monitoring method based on quantile regression static CoVaR model

Abstract

The invention relates to the technical field of ecological risk monitoring, and discloses an ecological risk monitoring method based on a quantile regression static CoVaR model, and the method comprises the following steps: firstly, estimating a correlation coefficient matrix M0 and a conditional probability matrix C in the static CoVaR model according to historical ecological data; and then solving an optimal correlation coefficient matrix Mh * and a conditional probability matrix Ch * corresponding to M0 and C by using a standard linear programming method, and finally substituting a new sample into the model to calculate an ecological risk value. Compared with a traditional CoVaR model, the model has the advantages that a correlation coefficient matrix and a conditional probability matrix are estimated by using a quantile regression method, conditional correlation among variables is considered, an effective model is provided for solving the multivariable problem including ecological risk variables and disaster risk variables, and the accuracy of ecological risk measurement and calculation is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of ecological risk monitoring, and in particular, relates to an ecological risk monitoring method based on a quantile regression static CoVaR model. Background Art

[0002] As global ecological and environmental problems become increasingly severe, the monitoring and assessment of ecological risks has become a focus of attention for governments, scientific research institutions and enterprises. Ecological risks not only affect the stability of ecosystems, but may also have a profound impact on the sustainable development of human society. In this context, how to effectively identify and assess ecological risks, especially to respond to the dynamic changes of ecological risks, has become a technical problem that needs to be solved urgently.

[0003] At present, ecological risk monitoring methods mainly rely on traditional statistical models or expert evaluation methods, but these methods often have certain limitations. For example, most traditional ecological risk assessment models ignore the complex interactions and dynamic changes between variables and cannot accurately capture the nonlinear relationships and sudden risks in the ecosystem. In addition, the determination of ecological risk thresholds usually adopts static standards, ignoring the dynamic impact of environmental changes on risk levels, resulting in inaccurate prediction results and difficulty in achieving real-time monitoring and early warning.

[0004] In order to solve the above problems, in recent years, the quantile regression method has been widely used in risk analysis in the financial field because it can better deal with the tail risk problem in the conditional distribution. The static CoVaR model based on quantile regression can take into account the variable dependency under different risk levels when estimating systemic risk, especially the variable relationship under extreme risk situations. However, the traditional static CoVaR model is mostly used in risk assessment of the financial system, and its application research in the field of ecological risk is relatively small.

[0005] In addition, as the complexity and diversity of ecosystems increase, how to accurately assess and divide ecological risk thresholds is the key to achieving real-time dynamic monitoring. In the existing technology, although there are many methods involving the setting of ecological risk thresholds, they often lack systematicity and adaptability, and fail to effectively divide them according to actual monitoring needs, resulting in low response speed and accuracy of ecological risk monitoring.

[0006] Therefore, a new ecological risk monitoring method is urgently needed. By introducing the quantile regression static CoVaR model, it can effectively estimate and reflect the conditional correlation between variables, overcome the limitations of traditional methods in complex environments, and combine with the gridded threshold division method to realize real-time dynamic monitoring of ecological risks and provide more accurate early warning and decision-making support for ecological environmental protection. Summary of the invention

[0007] In order to solve the above technical problems, the basic concept of the technical solution adopted by the present invention is:

[0008] The ecological risk monitoring method based on the quantile regression static CoVaR model includes the following steps:

[0009] Step S1: Collect ecological data of the area to be monitored, including water quality monitoring, climate change, land use change and other related factors affecting ecological risks;

[0010] Step S2: preprocessing the ecological data, including data cleaning, missing value filling and outlier detection;

[0011] Step S3: Select appropriate quantile values, use the quantile regression model to perform regression modeling on the ecological data, and obtain regression coefficients under different quantiles;

[0012] Step S4: Based on the regression coefficient, the risk value VaR of each ecological unit is calculated to obtain the risk measure under each quantile;

[0013] Step S5: Calculate the transmission effect of each ecological unit on the overall ecosystem risk, obtain the CoVaR value, and evaluate the risk spillover effect of each unit through the CoVaR value;

[0014] Step S6: Rank the risk of each ecological unit according to the CoVaR value and identify the unit with the greatest impact on ecosystem risk;

[0015] Step S7: Formulate corresponding ecological protection and risk warning measures based on the ranking results.

[0016] As a preferred embodiment of the present invention, the regression equation used in the quantile regression model is:

[0017] Define the system as s, the unit as i, and the loss in unit i as X i In case of:

[0018]

[0019] in, Represents the q-quantile loss estimate of the system at a certain confidence level.

[0020] As a preferred embodiment of the present invention, it can be known from the definition of VaR:

[0021]

[0022] When X i Take When , the CoVaR value of institution i can be calculated:

[0023]

[0024] As a preferred embodiment of the present invention, the q quantile of the yield sequence of institution i can be calculated. set up is the risk spillover value, and its expression is:

[0025]

[0026] The quantile q = 0.5 refers to the normal situation of institution i. The quantile q = 0.05 means that the unit i has

[0027] As a preferred embodiment of the present invention, It represents a marginal effect, reflecting the level of systematic risk spillover of unit i to system S. At the same time, because the values ​​of VaR, CoVaR and △CoVaR are usually negative, their absolute values ​​are generally used to represent the degree of risk. The larger the absolute value, the greater the risk.

[0028] As a preferred embodiment of the present invention, the ecological data further includes ecological indicators such as soil quality, vegetation coverage, species diversity, etc. in the region.

[0029] As a preferred embodiment of the present invention, the ecological unit includes ecological functional areas such as water bodies, wetlands, forests, and grasslands.

[0030] As a preferred implementation of the present invention, the data preprocessing includes outlier detection, standardization processing and smoothing processing of time series data.

[0031] As a preferred embodiment of the present invention, the risk warning measures include giving priority to protecting ecological units with higher CoVaR values ​​and formulating targeted risk mitigation strategies therefor.

[0032] As a preferred embodiment of the present invention, the ecological risk monitoring system further includes a data input module, a quantile regression model calculation module, a CoVaR calculation module, a risk assessment module and a result output module.

[0033] As a preferred embodiment of the present invention, the present invention also discloses an ecological risk monitoring system based on a quantile regression static CoVaR model, comprising: a data input module, used to collect ecological data of the area to be monitored; a quantile regression model calculation module, used to perform quantile regression modeling on the collected ecological data to obtain regression coefficients under each quantile; a CoVaR calculation module, used to calculate the CoVaR value of each ecological unit according to the regression results, and evaluate the risk spillover effect between ecological units; a risk assessment module, used to rank the risks of each ecological unit according to the CoVaR value, and formulate ecological protection and risk warning measures; a result output module, used to display the risk assessment results of each ecological unit and corresponding risk management suggestions.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] Compared with the traditional CoVaR model, the model of the present invention uses the quantile regression method to estimate the correlation coefficient matrix and the conditional probability matrix, takes into account the conditional correlation between variables, provides an effective model for solving multivariate problems involving ecological risk variables and disaster risk variables, and improves the accuracy of ecological risk measurement. In order to achieve real-time dynamic monitoring of ecological risks, the present invention proposes a grid-based ecological risk threshold division method. When the ecological risk value exceeds the risk threshold, an early warning is issued.

[0036] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In the attached picture:

[0038] Figure 1 Flow chart of the ecological risk monitoring method based on the quantile regression static CoVaR model. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The following embodiments are used to illustrate the present invention.

[0040] like Figure 1 As shown in the figure, the ecological risk monitoring method based on the quantile regression static CoVaR model includes the following steps:

[0041] Step S1: Collect ecological data of the area to be monitored, including water quality monitoring, climate change, land use change and other related factors affecting ecological risks;

[0042] Step S2: preprocessing the ecological data, including data cleaning, missing value filling and outlier detection;

[0043] Step S3: Select appropriate quantile values, use the quantile regression model to perform regression modeling on the ecological data, and obtain regression coefficients under different quantiles;

[0044] Step S4: Based on the regression coefficient, the risk value VaR of each ecological unit is calculated to obtain the risk measure under each quantile;

[0045] Step S5: Calculate the transmission effect of each ecological unit on the overall ecosystem risk, obtain the CoVaR value, and evaluate the risk spillover effect of each unit through the CoVaR value;

[0046] Step S6: Rank the risk of each ecological unit according to the CoVaR value and identify the unit with the greatest impact on ecosystem risk;

[0047] Step S7: Formulate corresponding ecological protection and risk warning measures based on the ranking results.

[0048] Among them, the regression equation used in the quantile regression model is:

[0049] Define the system as s, the unit as i, and the loss in unit i as X i In case of:

[0050]

[0051] in, Represents the q-quantile loss estimate of the system at a certain confidence level.

[0052] From the definition of VaR, we can know that:

[0053]

[0054] When X i Take When , the CoVaR value of institution i can be calculated:

[0055]

[0056] Through the q quantile of the return series of institution i, we can calculate set up is the risk spillover value, and its expression is:

[0057]

[0058] The quantile q = 0.5 refers to the normal situation of institution i. The quantile q = 0.05 means that the unit i has It represents a marginal effect, reflecting the level of systemic risk spillover of unit i to system S. At the same time, because the values ​​of VaR, CoVaR and △CoVaR are usually negative, their absolute values ​​are generally used to represent the degree of risk. The larger the absolute value, the greater the risk. The ecological data further includes ecological indicators such as soil quality, vegetation coverage, and species diversity in the region. The ecological units include ecological functional areas such as water bodies, wetlands, forests, and grasslands. The data preprocessing includes outlier detection, standardization, and smoothing of time series data. The risk warning measures include giving priority to protecting ecological units with higher CoVaR values ​​and formulating targeted risk mitigation strategies for them. The ecological risk monitoring system further includes a data input module, a quantile regression model calculation module, a CoVaR calculation module, a risk assessment module, and a result output module.

[0059] The present invention also discloses an ecological risk monitoring system based on a quantile regression static CoVaR model, comprising: a data input module, used to collect ecological data of a to-be-monitored area; a quantile regression model calculation module, used to perform quantile regression modeling on the collected ecological data to obtain regression coefficients under each quantile; a CoVaR calculation module, used to calculate the CoVaR value of each ecological unit according to the regression result, and to evaluate the risk spillover effect between ecological units; a risk assessment module, used to rank the risks of each ecological unit according to the CoVaR value, and to formulate ecological protection and risk early warning measures; and a result output module, used to display the risk assessment results of each ecological unit and corresponding risk management suggestions.

Claims

1. The ecological risk monitoring method based on the quantile regression static CoVaR model is characterized by: The steps include: Step S1: Collect ecological data of the area to be monitored, including water quality monitoring, climate change, land use change and other related factors affecting ecological risks; Step S2: preprocessing the ecological data, including data cleaning, missing value filling and outlier detection; Step S3: Select appropriate quantile values, use the quantile regression model to perform regression modeling on the ecological data, and obtain regression coefficients under different quantiles; Step S4: Based on the regression coefficient, the risk value VaR of each ecological unit is calculated to obtain the risk measure under each quantile; Step S5: Calculate the transmission effect of each ecological unit on the overall ecosystem risk, obtain the CoVaR value, and evaluate the risk spillover effect of each unit through the CoVaR value; Step S6: Rank the risk of each ecological unit according to the CoVaR value and identify the unit with the greatest impact on ecosystem risk; Step S7: Formulate corresponding ecological protection and risk warning measures based on the ranking results.

2. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: in, The regression equation used in the quantile regression model is: Define the system as s, the unit as i, and the loss in unit i as X i In case of: in, Represents the q-quantile loss estimate of the system at a certain confidence level.

3. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 2 is characterized in that: From the definition of VaR, we can know that: When X i Take When , the CoVaR value of institution i can be calculated:

4. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 3 is characterized in that: Through the q quantile of the return series of institution i, we can calculate set up is the risk spillover value, and its expression is: The quantile q = 0.5 refers to the normal situation of institution i. The quantile q = 0.05 means that the unit i has 5. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: It represents a marginal effect, reflecting the level of systematic risk spillover of unit i to system S. At the same time, because the values ​​of VaR, CoVaR and △CoVaR are usually negative, their absolute values ​​are generally used to represent the degree of risk. The larger the absolute value, the greater the risk.

6. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: The ecological data further include ecological indicators such as soil quality, vegetation coverage, species diversity, etc. in the region.

7. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: The ecological units include ecological functional areas such as water bodies, wetlands, forests, and grasslands.

8. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: The data preprocessing includes outlier detection, standardization and smoothing of time series data.

9. The ecological risk monitoring method based on the quantile regression static CoVaR model according to claim 1 is characterized in that: The risk warning measures include giving priority to protecting ecological units with higher CoVaR values ​​and formulating targeted risk mitigation strategies for them. The ecological risk monitoring system further includes a data input module, a quantile regression model calculation module, a CoVaR calculation module, a risk assessment module and a result output module.

10. An ecological risk monitoring system based on a quantile regression static CoVaR model, applied to the ecological risk monitoring method based on a quantile regression static CoVaR model according to any one of claims 1 to 9, characterized in that: include: The data input module is used to collect ecological data of the area to be monitored; the quantile regression model calculation module is used to perform quantile regression modeling on the collected ecological data to obtain the regression coefficients under each quantile; the CoVaR calculation module is used to calculate the CoVaR value of each ecological unit based on the regression results and evaluate the risk spillover effect between ecological units; The risk assessment module is used to rank the risks of each ecological unit according to the CoVaR value and formulate ecological protection and risk warning measures; The result output module is used to display the risk assessment results of each ecological unit and the corresponding risk management recommendations.

Images