A method for evaluating water ecological health in wetland buffer zones based on trend judgment

Abstract

The present invention discloses a method for evaluating the water ecological health of a wetland buffer zone based on trend judgment, including: S1: determining influencing factors based on Meta analysis: setting search keywords to collect experimental observation data of potential influencing factors, obtaining key influencing factors of water ecological health through environmental factor association analysis, and calculating the weight of influencing factors; S2: collecting indicator data; S3: establishing a prediction model based on machine learning; S4: trend judgment: inputting the predicted values ​​of each key influencing factor and the influence weights of different influencing factor types into the water ecological health evaluation model, obtaining the water ecological health status in the future n years, and thus, as a prediction evaluation of the current water ecological health status. The method can obtain influencing factors with strong correlation and sensitive impact, eliminate the limitations caused by regional scale surveys to the maximum extent, evaluate the current water ecological health from a future perspective, and better reflect the development trend and current status of the current water ecological health status.

Description

Technical Field

[0001] The invention belongs to the field of water ecological health assessment, and mainly relates to a wetland buffer zone water ecological health assessment method based on trend judgment. Background Art

[0002] Wetland refers to an area where the surface is too wet or water often accumulates, and wetland organisms grow. The wetland ecosystem is a unified whole composed of wetland plants, animals living in the wetland, microorganisms and their environment. Wetlands have multiple functions: protecting biodiversity, regulating runoff, improving water quality, regulating microclimate, providing food and industrial raw materials, providing tourism resources, etc. Water quality evaluation refers to the selection of corresponding water quality parameters, water quality standards and evaluation methods according to the evaluation objectives, and the evaluation of the quality utilization value of water bodies and the treatment effect of water. Water quality evaluation is a basic work for the rational development, utilization and protection of water resources. According to different evaluation types, the corresponding water quality standards are adopted. To evaluate the quality of water environment, the surface water environment quality standards are adopted; to evaluate the quality of aquaculture water bodies, the water quality standards for fishery water are adopted; to evaluate the water quality of water sources at centralized drinking water intake points, the surface water hygiene standards are used; to evaluate farmland irrigation water, the farmland irrigation water quality standards are adopted. Generally, various water quality standards promulgated by the national or local governments are used as evaluation standards.

[0003] Most of the current evaluation methods are based on field monitoring sampling and evaluate the water ecological health status of the monitored time and space, which is not forward-looking. In terms of the selection of evaluation factors, most of them consider a single type of impact, which has a one-sided impact on the overall water ecological health evaluation. However, due to the temporal and spatial limitations of sampling, the selection of multiple factors cannot determine the corresponding weights according to their actual impact, resulting in the loss of parameter information expression. In addition, in terms of prediction and forecasting, machine learning is an advanced and efficient means, but a common problem for machine learning is that its predictions are often not based on physical meaning. In other words, the algorithm is only looking for the correlation between input and output, but sometimes this correlation may be "false" or may give wrong results. Direct use in water ecological health evaluation may result in the algorithm model being unable to handle the results when the relationship between input and output changes. Summary of the invention

[0004] The purpose of the invention is to address the problems existing in the above-mentioned prior art and to propose a wetland buffer zone water ecological health assessment method based on trend judgment.

[0005] The purpose of the invention is achieved by a method for evaluating the water ecological health of a wetland buffer zone based on trend judgment, which is characterized by comprising the following steps:

[0006] S1: Determine influencing factors based on Meta-analysis: set search keywords, conduct literature surveys in academic databases, collect experimental observation data of different types of potential influencing factors, and obtain key influencing factors of water ecological health through environmental factor association analysis; obtain the influence weights of different factor types and the influence weights of different influencing factors of the same type through weight analysis;

[0007] S2: Collect indicator data: Collect multi-year datasets of key influencing factors in the target area;

[0008] S3: Establish a prediction model based on machine learning: divide the multi-year data set of key influencing factors into training sets and prediction sets according to different time windows, and establish prediction models corresponding to each key influencing factor through machine learning methods;

[0009] S4: Trend judgment: Take the overall data of key influencing factors collected as the basis for prediction, input it into the corresponding prediction model, and obtain the predicted value of the key influencing factor n years in the future; after scoring the predicted value of each key influencing factor, combine it with the different influencing factor types and the corresponding influence weights of different influencing factors and input it into the water ecological health assessment model to obtain the water ecological health status n years in the future, thus serving as a predictive evaluation of the current water ecological health status.

[0010] Preferably, the keywords are "phytoplankton", "zooplankton", "diversity", and "water ecological health".

[0011] Preferably, the weight analysis in S1 is variance decomposition analysis.

[0012] Preferably, the environmental factor association analysis in S1 is to calculate the direct correlation between the influencing factors and the water ecological health evaluation indicators through the maximum information coefficient, and the calculation method of the maximum information coefficient MIC is as follows:

[0013]

[0014] In the formula, X and Y are environmental impact factors and water ecological health evaluation indicators respectively; B is a variable, which is generally selected as 0.6 power of the data volume.

[0015] Preferably, the water ecological health evaluation model in S4 is a water ecological health index model, wherein the water ecological health index M is calculated as follows:

[0016]

[0017] In the formula, K j is the classification weight of the j-th indicator, G ji is the indicator weight of the i-th indicator in the j-th indicator category, X jiAssign a score to the i-th indicator in the j-th category of indicators.

[0018] Preferably, the scoring system for indicator scoring can refer to the Technical Specifications for Water Ecological Health Assessment DB11 / T1722-2020.

[0019] The present invention discloses a method for evaluating the water ecological health of a wetland buffer zone based on trend judgment, which has the following advantages: (1) the potential influencing factors of water ecological health are collected by integrating existing research, and the key influencing factors are determined on a global scale based on Meta-analysis, thereby obtaining influencing factors with stronger correlation and more sensitive impact, and maximally eliminating the limitations caused by regional scale surveys; (2) based on the time series data set of existing key influencing factors, future key influencing factors are predicted through machine learning, thereby improving the prediction accuracy; (3) by evaluating the future prediction data of key influencing factors, the current water ecological health evaluation from a future perspective is obtained, and the evaluation can better reflect the development trend and health status of the current water ecological health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of the principle of a method for evaluating water ecological health in a wetland buffer zone based on trend judgment according to the present invention. DETAILED DESCRIPTION

[0021] A method for evaluating the water ecological health of a wetland buffer zone based on trend judgment, characterized in that it comprises the following steps:

[0022] S1: Determine influencing factors based on Meta-analysis: Set "phytoplankton", "zooplankton", "diversity" and "water ecological health" as search keywords, conduct literature surveys in academic databases, collect experimental observation data of different types of potential influencing factors, and obtain the influence weights of different factor types and the influence weights of different influencing factors of the same type through variance decomposition analysis; obtain the key influencing factors of water ecological health through environmental factor association analysis;

[0023] Environmental factor association analysis is to calculate the direct correlation between influencing factors and water ecological health evaluation indicators through the maximum information coefficient. The calculation method of the maximum information coefficient MIC is as follows:

[0024]

[0025] In the formula, X and Y are environmental impact factors and water ecological health evaluation indicators respectively; B is a variable, which is generally selected as 0.6 power of the data volume.

[0026] S2: Collect indicator data: Collect multi-year datasets of key influencing factors in the target area;

[0027] S3: Establish a prediction model based on machine learning: divide the multi-year data set of key influencing factors into training sets and prediction sets according to different time windows, and establish prediction models corresponding to each key influencing factor through machine learning methods;

[0028] S4: Trend judgment: Take the overall data of key influencing factors collected as the prediction basis, input the corresponding prediction model, and obtain the predicted value of the key influencing factor in the next n years; input the predicted value of each key influencing factor and the influence weight of different influencing factor types into the water ecological health evaluation model to obtain the water ecological health status in the next n years. Therefore, as a predictive evaluation of the current water ecological health status, the water ecological health evaluation model is a water ecological health index model, wherein the calculation method of the water ecological health index M is as follows:

[0029] M=∑K j (ΣG ji X ji )

[0030] In the formula, K j is the classification weight of the j-th indicator, G ji is the indicator weight of the i-th indicator in the j-th indicator category, X ji Assign points to the i-th indicator in the j-th category. The scoring system for indicator assignment can refer to the Technical Specification for Water Ecological Health Assessment DB11 / T1722-2020.

Claims

1. A method for evaluating the water ecological health of wetland buffer zones based on trend judgment, characterized in that: The following steps are included: S1: Determine influencing factors based on Meta-analysis: set search keywords, conduct literature surveys in academic databases, collect experimental observation data of different types of potential influencing factors, and obtain key influencing factors of water ecological health through environmental factor association analysis; obtain the influence weights of different factor types and the influence weights of different influencing factors of the same type through weight analysis; S2: Collect indicator data: Collect multi-year datasets of key influencing factors in the target area; S3: Establish a prediction model based on machine learning: divide the multi-year data set of key influencing factors into training sets and prediction sets according to different time windows, and establish prediction models corresponding to each key influencing factor through machine learning methods; S4: Trend judgment: Take the overall data of key influencing factors collected as the prediction basis, input the corresponding prediction model, and obtain the predicted value of the key influencing factor in the next n years; after scoring the predicted value of each key influencing factor, combine it with the different influencing factor types and the corresponding influence weights of different influencing factors and input it into the water ecological health evaluation model to obtain the water ecological health status in the next n years, which is used as the prediction evaluation of the current water ecological health status; The water ecological health evaluation model is a water ecological health index model, wherein the water ecological health index M is calculated as follows: M=∑K j (ΣG ji X ji ) In the formula, K j is the classification weight of the j-th indicator, G ji is the indicator weight of the i-th indicator in the j-th indicator category, X ji Assign a score to the i-th indicator in the j-th category of indicators; The scoring system for assigning points to the indicators is the Technical Specifications for Water Ecological Health Assessment DB11 / T1722-2020.

2. The method for evaluating the water ecological health of a wetland buffer zone based on trend judgment according to claim 1 is characterized by: The keywords are "phytoplankton", "zooplankton", "diversity" and "water ecological health".

3. The method for evaluating the water ecological health of a wetland buffer zone based on trend judgment according to claim 1 is characterized by: The weight analysis described in S1 is a variance decomposition analysis.

4. The method for evaluating the water ecological health of a wetland buffer zone based on trend judgment according to claim 1 is characterized by: The environmental factor association analysis described in S1 is to calculate the direct correlation between the influencing factors and the water ecological health evaluation indicators through the maximum information coefficient. The calculation method of the maximum information coefficient MIC is as follows: In the formula, X and Y are environmental impact factors and water ecological health assessment indicators respectively; B is a variable, which is generally selected as 0.6 power of the data volume.

Images