Method for predicting enteromorpha green tide outbreak scale based on meteorological and hydrological conditions

Abstract

The invention discloses an enteromorpha green tide outbreak scale prediction method based on meteorological and hydrological conditions, and belongs to the technical field of enteromorpha green tide prediction. Comprising the steps of collecting multi-source multi-element data of a target sea area, and performing quality control processing on the multi-source multi-element data; based on multi-source multi-element data analysis, obtaining a statistical relationship between the meteorological and hydrological factors and the enteromorpha green tide outbreak scale; establishing an enteromorpha green tide scale long-term prediction model based on the statistical relationship between the meteorological hydrological factors and the enteromorpha green tide outbreak scale; and predicting the outbreak scale of the enteromorpha green tide through the enteromorpha green tide scale long-term prediction model. According to the method, a solution can be provided for the problems that the number of enteromorpha green tide outbreak scale prediction methods is small and the accuracy is poor in the prior art.

Description

Technical Field

[0001] The invention relates to the technical field of forecasting the scale of Enteromorpha green tide, and more particularly to a method for forecasting the outbreak scale of Enteromorpha green tide based on meteorological and hydrological conditions. Background Art

[0002] Enteromorpha prolifera is a common marine ecological disaster that has frequently occurred in recent years and has attracted widespread attention. This phenomenon not only has a serious impact on the marine ecological environment, but also has a significant negative impact on marine aquaculture and coastal landscapes. The outbreak of Enteromorpha prolifera is usually manifested as a large-scale green tide phenomenon, which leads to hypoxia, insufficient light and imbalance of the ecosystem, thus affecting the survival and reproduction of marine organisms.

[0003] The outbreak of Enteromorpha green tide is closely related to a variety of environmental factors. The enrichment of nutrients is the core factor that triggers the large-scale outbreak of Enteromorpha green tide, and meteorological and hydrological conditions are important factors in the outbreak of Enteromorpha green tide. With the acceleration of industrialization and urbanization, agricultural non-point source pollution and urban sewage discharge have significantly increased the concentration of nutrients such as nitrogen and phosphorus in water bodies, thereby providing rich nutrients for the growth of Enteromorpha. Studies have shown that sea surface temperature is an important hydrological condition for the growth of Enteromorpha organisms, and has a significant impact on the growth and reproduction of Enteromorpha. Under suitable water temperature conditions, Enteromorpha can grow rapidly and reproduce in large numbers. Too high or too low temperatures will affect its growth rate and reproductive capacity. For example, the water temperature in the Yellow Sea is suitable in spring and summer, providing good natural conditions for the growth of Enteromorpha. However, in some years, the seawater temperature continues to rise in summer, even reaching more than 30 degrees Celsius, which is much higher than the suitable growth temperature of Enteromorpha. This high temperature environment has an inhibitory effect on the reproduction of Enteromorpha and can accelerate the decline of the scale of Enteromorpha green tide. The wind field on the sea surface is an important meteorological condition that affects the distribution and drift of Enteromorpha. The drift direction of Enteromorpha is usually basically consistent with the wind direction on the sea surface. The size of the wind and the change of the wind direction will directly affect the aggregation and diffusion of Enteromorpha. Continuous strong winds can also enhance the exchange between the upper and lower layers of the water body, bring the bottom nutrients to the surface, and promote the growth and reproduction of Enteromorpha. In addition, precipitation is also an important meteorological condition that affects the growth and reproduction of Enteromorpha. It is an important source of N, P nutrients and trace elements in seawater, and has an important role in the growth of marine phytoplankton. Studies have confirmed that whether it is a one-time or intermittent addition of high-nitrogen and low-phosphorus rainwater, the growth of phytoplankton has a positive response; the on-site enclosure experiment of Enteromorpha has confirmed that sufficient and continuous nutrient supplementation is the material basis for the outbreak of Enteromorpha green tide, and the addition of trace elements such as Fe and Mn can significantly promote the growth and reproduction of Enteromorpha; especially when the temperature reaches 15°C, effective precipitation can stimulate the triggering of green tide. In the rapid development stage, precipitation is an important factor affecting the scale of green tide growth.

[0004] However, there are few existing solutions for predicting the scale of Ulva green tides, with poor accuracy, and the many factors mentioned above are not taken into account. This is not conducive to a deep understanding of the mechanism of Ulva outbreaks and cannot provide a reference for effective prevention and control measures.

[0005] Therefore, how to provide a method for predicting the scale of Ulva green tide outbreaks based on meteorological and hydrological conditions that can solve the above problems is an issue that technical personnel in this field urgently need to solve. Summary of the invention

[0006] In view of this, the present invention provides a method for predicting the outbreak scale of Ulva green tide based on meteorological and hydrological conditions. By monitoring and analyzing the above meteorological and hydrological conditions, a relatively new prediction basis can be provided for predicting the outbreak scale of Ulva green tide.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] A method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions, comprising:

[0009] Collect multi-source and multi-factor data of the target sea area and perform quality control on the multi-source and multi-factor data;

[0010] Based on multi-source and multi-factor data, the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreak was obtained;

[0011] A long-term prediction model for Enteromorpha green tide is established based on the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks.

[0012] The outbreak scale of Enteromorpha green tide is predicted by the long-term prediction model of Enteromorpha green tide scale.

[0013] Furthermore, the collection of multi-source and multi-factor data of the target sea area includes: collecting wind, precipitation, air temperature, water temperature, and the area and location data of the Ulva green tide outbreak obtained through remote sensing, ocean stations, buoys, surface drifts, drones, on-site monitoring, and tracking monitoring.

[0014] Furthermore, the quality control processing of multi-source and multi-factor data includes:

[0015]

[0016] In the formula, x' is the standardized data, x is the original multi-source multi-factor data, and x max and x min They are the maximum and minimum values ​​of multi-source and multi-factor data respectively.

[0017] Furthermore, the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks is obtained based on multi-source and multi-factor data, including:

[0018] The linear correlation analysis method was used to quantitatively analyze the correlation between meteorological and hydrological factors and the maximum distribution area or maximum coverage area of ​​the green tide of Enteromorpha in the target area over the years. The correlation coefficient r formula is:

[0019]

[0020] Among them, z is the maximum distribution area or maximum coverage area of ​​Enteromorpha green tide in previous years; y is the meteorological and hydrological factors extracted by analysis; is the average value of z; is the mean value of y.

[0021] Furthermore, the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks obtained based on multi-source and multi-factor data also includes:

[0022] The correlation between sea surface wind factors and the scale of Enteromorpha green tide outbreak;

[0023] The correlation between precipitation factor and the outbreak scale of Enteromorpha green tide;

[0024] Correlation between sea surface temperature factors and the scale of Enteromorpha green tide outbreaks.

[0025] Furthermore, the long-term prediction model for the Enteromorpha green tide includes: a single-factor long-term trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area and a multi-factor trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area.

[0026] Furthermore, based on the correlation between the sea surface wind factor and the scale of the green tide outbreak, a single factor long-term trend prediction model for the scale of the green tide outbreak in the target area was established, including:

[0027] Take the mean meridional wind value of the target sea area at different times, analyze the statistical relationship between it and the annual maximum scale of the Enteromorpha green tide, and select the main control index of sea surface wind for the annual maximum scale prediction of the Enteromorpha green tide;

[0028] A single-factor long-term trend prediction model for the outbreak scale of Ulva green tide in the target area is established based on the main control index of sea surface wind.

[0029] Furthermore, based on the correlation between precipitation factors and the scale of the green tide outbreak, a single factor long-term trend prediction model for the scale of the green tide outbreak in the target area was established, including:

[0030] The precipitation values ​​in different periods and different sea areas were taken to analyze their correlation with the annual maximum scale of Enteromorpha green tide, and the main control indicators of precipitation for the annual maximum scale prediction of Enteromorpha green tide were selected;

[0031] A single-factor long-term trend prediction model for the outbreak scale of Enteromorpha green tide in the target area is established based on the precipitation main control index.

[0032] Furthermore, the PSO-BP neural network model architecture was used to establish a multi-factor trend prediction model for the outbreak scale of Enteromorpha prolifera in the target area. The construction process was as follows:

[0033] (1) Normalize the data, establish a BP neural network, determine the topology and initialize the network weights and thresholds;

[0034] (2) Initialize PSO parameters, including maximum number of iterations, population size, individual learning factor, social learning factor, and inertia weight parameter;

[0035] (3) Initialize the population position of PSO and calculate the number of variable elements that need to be optimized based on the BP neural network structure;

[0036] (4) PSO optimization, the fitness function is set to the mean square error predicted by the BP neural network, the PSO optimization process is cyclic, the position of the optimal particle is continuously updated until the maximum number of iterations is reached, and the PSO algorithm is terminated;

[0037] The optimal weight threshold parameters optimized by the PSO algorithm were assigned to the BP neural network, and the optimal PSO-BP neural network model was output as a multi-factor trend prediction model for the outbreak scale of Enteromorpha green tide in the target area.

[0038] Furthermore, the tansig function is selected as the activation function of the BP neural network.

[0039] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method for predicting the scale of the outbreak of Enteromorpha green tide based on meteorological and hydrological conditions, and the specific beneficial effects are:

[0040] (1) A long-term prediction model for Enteromorpha green tide is established by using the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks, which can provide a solution for the long-term prediction of Enteromorpha green tide outbreaks;

[0041] (2) The prediction results of the scale of Enteromorpha can provide a scientific basis for formulating prevention and control measures for Enteromorpha ecological disasters and the deployment of salvage scale, which is an urgent need for emergency response to Enteromorpha ecological disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0043] Figure 1A schematic diagram of the method provided by the present invention;

[0044] Figure 2 A schematic diagram of the PSO-BP neural network model construction process provided by the present invention;

[0045] Figure 3 The relative error curve of the maximum distribution area of ​​the Enteromorpha green tide provided by the present invention;

[0046] Figure 4 A comparison chart of the predicted values ​​and actual values ​​of the training data provided by the present invention;

[0047] Figure 5 The BP data fitting diagram provided by the present invention;

[0048] Figure 6 This is a data fitting diagram of the PSO-BP neural network model provided by the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.

[0050] The purpose of the present invention is to provide a method for predicting the outbreak scale of Enteromorpha green tide based on meteorological and hydrological conditions, the method comprising: collecting multi-source and multi-factor data of the target sea area, and performing quality control processing on the multi-source and multi-factor data; obtaining the statistical relationship between meteorological and hydrological factors and the outbreak scale of Enteromorpha green tide based on the multi-source and multi-factor data; establishing a long-term prediction model for Enteromorpha green tide based on the statistical relationship between meteorological and hydrological factors and the outbreak scale of Enteromorpha green tide; predicting the outbreak scale of Enteromorpha green tide by using the long-term prediction model for Enteromorpha green tide. The present invention provides a solution to the fact that there are few methods for predicting the outbreak scale of Enteromorpha green tide in the prior art and the accuracy is poor.

[0051] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] join Figure 1 The embodiment of the present invention discloses a method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions, comprising:

[0053] Collect multi-source and multi-factor data of the target sea area and perform quality control on the multi-source and multi-factor data;

[0054] Based on multi-source and multi-factor data, the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreak was obtained;

[0055] A long-term prediction model for Enteromorpha green tide is established based on the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks.

[0056] The outbreak scale of Enteromorpha green tide is predicted by the Enteromorpha green tide long-term prediction model.

[0057] In a specific embodiment, collecting multi-source and multi-factor data of the target sea area includes: collecting data such as wind, precipitation, air temperature, water temperature, scale and location of Ulva green tide outbreaks obtained through remote sensing, ocean stations, buoys, drifts, drones, on-site monitoring, and tracking monitoring.

[0058] In a specific embodiment, quality control processing is performed on multi-source and multi-factor data, including:

[0059]

[0060] In the formula, x' is the standardized data, x is the original multi-source multi-factor data, and x max and x min They are the maximum and minimum values ​​of multi-source and multi-factor data respectively.

[0061] Based on multi-source and multi-factor data, the statistical relationship between meteorological and hydrological factors and the scale factor of the green tide outbreak was obtained, including:

[0062] In a specific embodiment, a linear correlation analysis method is used to quantitatively analyze the correlation between meteorological and hydrological factors and the maximum distribution area or maximum coverage area of ​​the green tide of Enteromorpha in the target area over the years. The correlation coefficient r formula is:

[0063]

[0064] Among them, z is the maximum distribution area or maximum coverage area of ​​Enteromorpha green tide in previous years; y is the meteorological and hydrological factors extracted by analysis; is the average value of z; is the mean value of y.

[0065] Specifically, the correlation coefficient r indicates the closeness of the correlation between the two, and the value range is [-1, 1]; r greater than 0 indicates positive correlation, and r less than 0 indicates negative correlation; the closer the correlation coefficient r is to 1 or -1, the stronger the correlation, and the closer the correlation coefficient r is to 0, the weaker the correlation; the correlation coefficient r equals 1, indicating that the two are completely positively correlated, and the correlation coefficient r = -1, indicating that the two are completely negatively correlated. Refer to Table 1, and the correlation degree of the variables is usually determined by the following value range.

[0066] Table 1 Correlation coefficient description

[0067] Serial number Correlation coefficient r value range Relevance 1 |r|＝1.0 Perfectly positive or negative correlation 2 0.8≤|r|<1.0 Very strong correlation 3 0.6≤|r|<0.8 Strong correlation 4 0.4≤|r|<0.6 Moderately related 5 0.2≤|r|<0.4 Weak correlation 6 0.0≤|r|<0.2 Very weak or no correlation

[0068] In a specific embodiment, the statistical relationship between meteorological and hydrological factors and the scale of the outbreak of Enteromorpha green tide is obtained based on multi-source and multi-factor data, and further includes:

[0069] The correlation between sea surface wind factors and the scale of Enteromorpha green tide outbreak;

[0070] The correlation between precipitation factor and the outbreak scale of Enteromorpha green tide;

[0071] Correlation between sea surface temperature factors and the scale of Enteromorpha green tide outbreaks.

[0072] Specifically, the correlation between sea surface wind factors and the scale of Enteromorpha green tide outbreaks includes: the relationship between the starting time of the East Asian summer monsoon and the scale of the Enteromorpha green tide, the relationship between the time when the summer monsoon affects northern Jiangsu and the scale of the Enteromorpha green tide, the relationship between the meridional wind and the scale of the Enteromorpha green tide, and the relationship between the number of strong wind days and the scale of the Enteromorpha green tide.

[0073] Specifically, the target area of ​​this embodiment is the waters of northern Jiangsu. Based on the previous research conclusion that "the sea surface wind during the growth period of Enteromorpha is closely related to the drift, aggregation and maximum scale of Enteromorpha", and based on the Enteromorpha results interpreted from satellite remote sensing images of the waters of northern Jiangsu, the relationship between the start time of the East Asian summer monsoon, the time of impact on the waters of northern Jiangsu, the meridional wind component of the waters of northern Jiangsu, and the number of strong wind days in the waters of northern Jiangsu and the annual maximum scale of the Enteromorpha green tide are analyzed. The following conclusions are drawn through statistical analysis of data over many years:

[0074] (1) According to the statistical analysis results from 2008 to 2022, the starting time of the East Asian summer monsoon is consistent with the changing trend of the annual maximum distribution area of ​​the Enteromorpha green tide, and the two are positively correlated. The correlation coefficients of polynomial regression and linear regression are 0.48 and 0.47, respectively, which are moderately correlated.

[0075] (2) According to the statistical analysis results from 2008 to 2022, the time when the East Asian summer monsoon affects the northern Jiangsu waters is relatively consistent with the change trend of the annual maximum distribution area of ​​the Enteromorpha green tide, and the two are positively correlated. The correlation coefficients of polynomial regression and linear regression are 0.68 and 0.61, respectively, which are strongly correlated.

[0076] (3) According to the statistical analysis results from 2008 to 2022, the meridional wind in mid-April is positively correlated with the annual maximum distribution area of ​​Enteromorpha green tide, with a polynomial regression correlation coefficient of 0.62, which is a strong correlation, and a linear regression correlation coefficient of 0.54, which is a moderate correlation. The meridional wind in mid-May is negatively correlated with the annual maximum distribution area of ​​Enteromorpha green tide, with a polynomial regression correlation coefficient of -0.6, which is a strong correlation, and a linear regression correlation coefficient of -0.58, which is a moderate correlation.

[0077] (4) According to the statistical analysis results from 2008 to 2023, the cumulative number of strong wind days in the 30 days before the annual maximum distribution area of ​​Ulva green tide is consistent with the changing trend of the annual maximum distribution area of ​​Ulva green tide. The polynomial regression correlation coefficient is as high as 0.72, which is a strong correlation, and the linear regression correlation coefficient is 0.5, which is a moderate correlation.

[0078] Based on the above conclusions, the time when the East Asian summer monsoon affects the waters of northern Jiangsu has the strongest correlation with the annual maximum distribution area of ​​the Enteromorpha green tide, but it is not predictable, so the meridional wind is considered as the main control indicator for the annual maximum scale prediction of the Enteromorpha green tide. Since the correlation between the meridional wind in mid-April and the annual maximum distribution area of ​​the Enteromorpha green tide is slightly stronger than that in mid-May, the average meridional wind in the northern Jiangsu waters in mid-April is used as one of the main control indicators for the annual maximum scale prediction of the Enteromorpha green tide.

[0079] Specifically, the correlation between precipitation factors and the scale of Enteromorpha green tide outbreaks includes using precipitation data in the South Yellow Sea and Enteromorpha green tide scale data (from satellite remote sensing interpretation results) to analyze and discuss the correlation between precipitation and the annual maximum scale of Enteromorpha green tide. Specifically, there are: the relationship between precipitation and the Enteromorpha outbreak process, the selection of statistical analysis indicators for precipitation data, the relationship between precipitation at ocean stations and the annual scale of Enteromorpha, and the relationship between precipitation in the sea area and the annual scale of Enteromorpha.

[0080] Specifically, taking the northern Jiangsu waters as an example, based on the statistics of the multi-year average daily precipitation in 18 different time periods (April, May, June, July, April-May, May-June, June-July, April-June, May-July, April-July, and 20 days, 30 days, 40 days, 50 days, 60 days, 70 days, 37 days (the average interval from scale to maximum distribution area over many years), and 77 days (the average interval from ship discovery to maximum distribution area over many years)) analyzed by ERA5 in 4 marine stations and 7 sea areas along the coast of Jiangsu Province, the correlation between the daily average precipitation and the annual maximum distribution area was analyzed, and it was found that:

[0081] (1) For the precipitation statistics of the ocean stations, the average daily precipitation in July and June-July was negatively correlated with the maximum distribution area of ​​the Enteromorpha green tide, showing strong correlation and weak correlation, respectively; the average daily precipitation in April and April-May was weakly positively correlated with the maximum distribution area of ​​the Enteromorpha green tide; the correlation between the precipitation statistics of the other ocean stations and the maximum distribution area of ​​the Enteromorpha green tide was even weaker, showing weak, very weak or no correlation.

[0082] (2) For the precipitation statistics of the sea area, the average daily precipitation in April-May and the average daily precipitation in April in the South 1 area and the radiating sandbar area were positively correlated with the maximum distribution area of ​​the Enteromorpha green tide, which were strongly correlated and moderately correlated, respectively. The average daily precipitation in July and June-July in the North area were negatively correlated with the maximum distribution area of ​​the Enteromorpha green tide, which were strongly correlated and moderately correlated, respectively. The correlation between the precipitation statistics of the remaining sea areas and the maximum distribution area of ​​the Enteromorpha green tide was weaker, showing moderate, weak, very weak or no correlation.

[0083] Based on the above conclusions, taking into account the predictability and related intensity, this embodiment considers using the average daily precipitation in April in the South 1 area as one of the main control indicators for the annual scale prediction of the Ulva green tide, and the average daily precipitation in April-May in the South 1 area as one of the main control indicators for the annual scale prediction and correction of the Ulva green tide.

[0084] Specifically, the correlation between the sea surface temperature factor and the scale of the Enteromorpha green tide outbreak includes: analysis of the annual relationship between the sea surface temperature factor and the distribution area of ​​the Enteromorpha green tide, and analysis of the interannual relationship between the sea surface temperature factor and the maximum distribution area of ​​the Enteromorpha green tide; among them, analysis of the annual relationship between the sea surface temperature factor and the distribution area of ​​the Enteromorpha green tide includes: the relationship between the SST of the ocean station and the distribution area of ​​the Enteromorpha green tide, and the relationship between the large-scale OISST and the distribution area of ​​the Enteromorpha green tide; analysis of the interannual relationship between the sea surface temperature factor and the maximum distribution area of ​​the Enteromorpha green tide includes: the relationship between the SST of the ocean station and the maximum distribution area of ​​the Enteromorpha green tide, and the relationship between the large-scale OISST and the maximum distribution area of ​​the Enteromorpha green tide.

[0085] Specifically, taking the northern Jiangsu waters as an example, based on the measured data from multiple oceanographic stations along the Jiangsu coast over the years and the analysis of large-scale OISST and the scale of green tides in the past 15 years, the following conclusions were drawn:

[0086] (1) In terms of intra-annual correlation analysis, the correlation analysis of the scale (distribution area) of the Enteromorpha green tide over the years and the measured SST of each ocean station showed that the correlation between the Binhai and Dafeng stations was high, with correlation coefficients of 0.60 and 0.65, respectively, indicating a strong correlation. The correlation analysis of the mean SST of the South Yellow Sea over the years and the distribution area of ​​the Enteromorpha green tide over the years showed a correlation coefficient of 0.64, indicating a strong correlation.

[0087] (2) In terms of interannual correlation analysis, the maximum distribution area of ​​the Enteromorpha green tide over the years has a high correlation with the SST anomalies in May and June at Dafeng Station and Binhai Heping Island, showing a negative correlation. In addition, the maximum distribution area of ​​the Enteromorpha green tide has a moderate positive correlation with the high temperature anomaly of the large-scale OISST in the South Yellow Sea in April, which can be used as one of the main control indicators for predicting the annual maximum distribution area of ​​the Enteromorpha green tide.

[0088] (3) In terms of the development process of the Enteromorpha green tide, the correlation coefficient between the time of the maximum distribution area of ​​the Enteromorpha green tide and the average sea surface temperature in June in the South Yellow Sea was -0.73, showing a strong correlation. The time of the maximum distribution area of ​​the Enteromorpha green tide and the average sea surface temperature in May and the time when the sea surface temperature reached 25°C showed a moderate correlation, which can be used as one of the key indicators for predicting the key time nodes of the development of the Enteromorpha green tide.

[0089] In a specific embodiment, the long-term prediction model for the Enteromorpha green tide includes: a single-factor long-term trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area and a multi-factor trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area.

[0090] In a specific embodiment, a single factor long-term trend prediction model for the outbreak scale of Enteromorpha green tide in the target area is established based on the correlation between the sea surface wind factor and the outbreak scale of Enteromorpha green tide, including:

[0091] The mean meridional wind value of the target sea area at different times is taken as one of the main control indicators of sea surface wind for the annual largest-scale prediction of Enteromorpha green tide;

[0092] A single-factor long-term trend prediction model for the outbreak scale of Ulva green tide in the target area is established based on the main control index of sea surface wind.

[0093] Specifically, based on the correlation analysis results between the sea surface wind factor and the scale of the Ulva green tide, the starting time of the East Asian summer monsoon and the average meridional wind in the northern Jiangsu waters in mid-April and mid-May were taken as one of the main control indicators of sea surface wind for the annual largest-scale prediction of the Ulva green tide.

[0094] (1) The polynomial regression model between the onset time of the East Asian summer monsoon and the scale of the Enteromorpha green tide is established as follows:

[0095] y=383.52x 2 +3682.3x-1363.6

[0096] Among them, x is the East Asian summer monsoon starting time index, y is the annual maximum distribution area of ​​Ulva green tide, and the correlation coefficient r is 0.48.

[0097] (2) The polynomial regression model between the mean meridional wind value and the scale of the lake in the northern Jiangsu waters in mid-April is as follows:

[0098] y=4392.2x 2 -991.59x+29807

[0099] Among them, x is the mean meridional wind value in the northern Jiangsu waters in mid-April (unit: m / s), y is the annual maximum distribution area of ​​the Enteromorpha green tide, and the correlation coefficient r is 0.62.

[0100] (3) The polynomial regression model between the mean meridional wind value and the scale of the Enteromorpha green tide in the northern Jiangsu waters in mid-May is as follows:

[0101] y=1848.3x 2 -9938.8x-1162.5

[0102] Among them, x is the mean meridional wind value in the northern Jiangsu waters in mid-May (unit: m / s), y is the annual maximum distribution area of ​​the Ulva green tide, and the correlation coefficient r is 0.6.

[0103] In a specific embodiment, a statistical model is established based on the correlation between the precipitation factor and the scale of the Enteromorpha green tide outbreak, and a power function regression equation is obtained, which is as follows:

[0104] Based on the correlation analysis between precipitation statistical indicators and the scale of Enteromorpha green tide, it can be seen that the average daily precipitation in April and April-May in South Zone 1 and Radiation Sandbar Area is positively correlated with the maximum distribution area of ​​Enteromorpha green tide, which is strongly correlated and moderately correlated, respectively. Therefore, from the perspective of predictability, the average daily precipitation in April in South Zone 1, Radiation Sandbar Area, and Radiation Sandbar Area + South Zone 1 is selected as one of the main precipitation control indicators for the annual maximum scale prediction of Enteromorpha green tide; from the perspective of predictability and high accuracy, the average daily precipitation in April-May in South Zone 1, Radiation Sandbar Area, and Radiation Sandbar Area + South Zone 1 is selected as one of the main precipitation control indicators for the annual maximum scale correction prediction of Enteromorpha green tide. They are respectively established with the annual maximum distribution area of ​​Enteromorpha green tide in pairs, and the power function regression equation is obtained as follows.

[0105] (1) Fitting formula for the scale of the South 1 area and Enteromorpha green tide:

[0106] y=8145.3x+21401

[0107] Where x is the average daily precipitation in April in South Zone 1, in mm; y is the annual maximum distribution area of ​​Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.47.

[0108] y=20385x 0.7114

[0109] Where x is the average daily precipitation in South Zone 1 from April to May, in mm; y is the annual maximum distribution area of ​​Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.69.

[0110] (2) Fitting formula for the scale of the radiation sandbar area + South 1 area and Enteromorpha green tide:

[0111] y=7843.8x+19654

[0112] Where x is the average daily precipitation in April in the radiation sandbar area + South 1 area, in mm; y is the annual maximum distribution area of ​​the Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.50.

[0113] y=17961x 0.7455

[0114] Where x is the average daily precipitation in the radiation sandbar area + South 1 area from April to May, in mm; y is the annual maximum distribution area of ​​the Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.66.

[0115] (3) Fitting formula for the scale of radiation sandbar area and Enteromorpha green tide:

[0116] y=6176.7x+21074

[0117] Where x is the average daily precipitation in the radiation sandbar area in April, in mm; y is the annual maximum distribution area of ​​the Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.48.

[0118] y=17493x 0.6679

[0119] Where x is the average daily precipitation in the radiation sandbar area from April to May, in mm; y is the annual maximum distribution area of ​​the Enteromorpha green tide, in km 2 ; The correlation coefficient between the two is R=0.60.

[0120] In a specific embodiment, based on the analysis of the annual largest-scale main control indicators of the Enteromorpha green tide, a PSO-BP neural network model is used to construct an annual-scale long-term trend prediction model.

[0121] Specifically, BP neural network is a multi-layer feedforward neural network based on the error back propagation theory, which is used to process nonlinear continuous functions. Its main features are signal forward transmission and error back propagation. It has the advantages of self-organization and self-learning nonlinear mapping capabilities, and can be used in information processing, numerical approximation, image recognition, speech analysis and other fields. PSO (particle swarm optimization algorithm) is a population-based random optimization technology. The algorithm imitates the clustering behavior of insects, animal flocks, bird flocks and fish schools, and searches for food in a cooperative manner. Each member of the group constantly changes its search mode by learning from its own experience and the experience of other members to find the best foraging point.

[0122] In the process of establishing the BP neural network, the random setting of the connection weights will lead to errors in the prediction results, and the gradient descent training has the disadvantages of slow speed and local minimum value, which makes it difficult to achieve global optimal training of the neural network. The particle swarm optimization algorithm PSO is used to optimize it to improve the prediction accuracy and generalization ability. Figure 2 As shown, the construction process is:

[0123] (1) Normalize the data, establish the BP neural network, determine the topology and initialize the network weights and thresholds;

[0124] (2) Initialize PSO parameters, such as the maximum number of iterations, population size, individual learning factor, social learning factor, inertia weight, etc.

[0125] (3) Initialize the population position of PSO and calculate the number of variable elements that need to be optimized based on the BP neural network structure;

[0126] (4) PSO optimization, the fitness function is set to the mean square error predicted by the BP network, the PSO optimization process is cyclic, the position of the optimal particle is continuously updated until the maximum number of iterations is reached, and the PSO algorithm is terminated;

[0127] The optimal weight threshold parameters optimized by the PSO algorithm were assigned to the BP neural network, and the optimal PSO-BP neural network model was output as a multi-factor trend prediction model for the outbreak scale of Enteromorpha green tide in the target area.

[0128] Specifically, data sources and preprocessing include:

[0129] (1) Data source

[0130] The time series statistical variables of the mid-April meridional wind intensity in the South Yellow Sea, the April SST high temperature anomaly, the April average daily precipitation in the South 1 area, and the annual maximum distribution area of ​​the Ulva green tide were collected and organized from 2008 to 2022, forming a total of 15 sets of interannual sample data.

[0131] (2) Data preprocessing

[0132] There will be significant differences in the dimensional magnitudes of the data in each dimension of the multi-factor original data, and they cannot be directly used in the prediction model. Therefore, the data must be normalized.

[0133] This embodiment uses the tansig function as the activation function of the BP neural network, and the normalization interval of the original sample data is [-1, 1]. The mapminmax function in the MATLAB program is enabled for normalization processing, and the format is as follows:

[0134] [y,PS]=mapminmax(x)

[0135] x represents the original data, y represents the normalized data, PS represents the parameters used for normalization, and the basic idea of ​​the mapminmax function is:

[0136]

[0137] In the above formula, x max 、x min They represent the maximum and minimum values ​​of the original sample data, respectively. x represents the original sample data, and y max With y min The default values ​​are 1 and -1.

[0138] In a specific embodiment, the model parameter setting includes:

[0139] There are three main steps in constructing a BP neural network model: determining the topological structure of the neural network, initializing the neural network weight threshold, and setting parameters such as learning rate and error.

[0140] (1) This model is determined as a three-layer BP neural network topology, and the number of neurons in each layer needs to be determined. According to the original sample data, the number of input nodes is 3 and the number of output nodes is 1. At present, the number of neurons in the hidden layer can be determined based on the empirical formula, which is as follows:

[0141]

[0142] In the above formula, m is the number of input layer nodes, n is the number of output layer nodes, and a is a constant ranging from 1 to 10. According to the formula, the number of hidden layer nodes is between 3 and 12.

[0143] (2) BP neural network can use many different types of activation functions, and select different weight ranges according to different activation functions. This work selects the tansig function as the hidden layer activation function, with an initial weight range of (-1,1), and the purelin function as the output layer activation function, with the initial weights being any value.

[0144] (3) Since BP neural network is prone to local minima, the learning rate should not be too large to ensure the convergence of the training process. Considering the limitation of the number of learning samples, the learning rate is selected as 0.05 in this embodiment, the number of training times is set to 1000, and the error accuracy is 0.001.

[0145] In a specific embodiment, the parameter settings of the PSO algorithm mainly include: maximum number of iterations, population size, individual learning factor, social learning factor, inertia weight, flight speed and other parameters:

[0146] (1) Maximum number of iterations: The maximum number of optimization attempts for each particle, ranging from 50 to 200.

[0147] (2) Population size: The number of particles involved in the optimization. The more particles there are, the higher the accuracy and the greater the amount of calculation. The value is between 40 and 200.

[0148] (3) Individual learning factor and social learning factor: The weight coefficient of two adjacent optimal solutions. By adjusting this parameter, the weight of global search and local search can be balanced. The value is between 1 and 2.

[0149] (4) Inertia weight: determines the convergence speed. The faster the convergence, the wider the search, but it will reduce the local optimization accuracy. The value is between 0.5 and 1.5.

[0150] (5) Flight speed: Optimizing the speed and determining the appropriate speed will affect the convergence speed. The value is between 0.5 and 1.5.

[0151] The present application obtains a model with an accuracy of 90.77% through a MATLAB simulation experiment of the maximum distribution area of ​​Enteromorpha green tide based on a PSO-BP neural network. Figure 3 ), a comparison chart of the predicted and actual values ​​of the training data (such as Figure 4 ), BP data fitting diagram (such as Figure 5 ) and PSO-BP neural network model data fitting diagram (such as Figure 6 ). Based on this prediction model, the 2023 Enteromorpha test data was input for prediction, and the minimum relative error was 1.55%. Both the particle swarm optimization (PSO) and the BP neural network are algorithms based on randomness. The optimization results of the same set of parameters may be different, and multiple repeated experiments are required to verify the robustness and reliability of the model. After multiple repeated tests, the accuracy of the PSO-BP neural network model is in the range of 88.22%-94.32%.

[0152] Specifically, the present invention can provide a scientific basis for predicting the scale of the outbreak of Enteromorpha green tide by monitoring and analyzing meteorological and hydrological conditions. This not only helps to deeply understand the mechanism of Enteromorpha green tide outbreak, but also provides a reference for formulating effective prevention and control measures. In order to cope with the threat of Enteromorpha green tide, strengthen the monitoring of the marine ecological environment, establish and improve the early warning system, and promote the in-depth development of related research, so as to promote sustainable marine economic development while protecting the marine ecological environment.

[0153] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0154] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined in the present embodiments may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown in the present embodiment, but will conform to the widest range consistent with the principles and novel features disclosed in the present embodiment.

Claims

1. A method for predicting the scale of Enteromorpha green tide outbreak based on meteorological and hydrological conditions, characterized in that: include: Collect multi-source and multi-factor data of the target sea area and perform quality control on the multi-source and multi-factor data; Based on the analysis of multi-source and multi-factor data, the statistical relationship between meteorological and hydrological factors and the scale of the green tide outbreak of Enteromorpha was obtained; A long-term prediction model for Enteromorpha green tide is established based on the statistical relationship between meteorological and hydrological factors and the scale of Enteromorpha green tide outbreaks. The outbreak scale of Enteromorpha green tide is predicted by the long-term prediction model of Enteromorpha green tide scale.

2. The method for predicting the scale of the outbreak of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 1, characterized in that: The multi-source and multi-factor data collected in the target sea area include: collecting wind, precipitation, air temperature, water temperature, and the area and location data of the Ulva green tide outbreak obtained through remote sensing, ocean stations, buoys, surface drifts, drones, on-site monitoring, and tracking monitoring.

3. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 1, characterized in that: The quality control processing of multi-source and multi-factor data includes: In the formula, x' is the standardized data, x is the original multi-source multi-factor data, and x max and x min They are the maximum and minimum values ​​of multi-source and multi-factor data respectively.

4. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 1, characterized in that: The statistical relationship between meteorological and hydrological factors and the scale factor of the green tide outbreak obtained based on multi-source and multi-factor data includes: The linear correlation analysis method was used to quantitatively analyze the correlation between meteorological and hydrological conditions and the maximum distribution area or maximum coverage area of ​​the green tide of Enteromorpha in the target area over the years. The correlation coefficient r formula is: Among them, z is the maximum distribution area or maximum coverage area of ​​Enteromorpha green tide in previous years; y is the meteorological and hydrological factors extracted by analysis; is the average value of z; is the mean value of y.

5. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 1, characterized in that: The statistical relationship between meteorological and hydrological factors and the outbreak scale of Enteromorpha enteromorpha based on multi-source and multi-factor data also includes: The correlation between sea surface wind factors and the outbreak scale of Enteromorpha green tide; The correlation between precipitation factor and the outbreak scale of Enteromorpha green tide; Correlation between sea surface temperature factors and the scale of Enteromorpha green tide outbreaks.

6. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 5, characterized in that: The long-term prediction models for the Enteromorpha green tide include: a single-factor long-term trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area and a multi-factor long-term trend prediction model for the outbreak scale of the Enteromorpha green tide in the target area.

7. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 6, characterized in that: Based on the correlation between sea surface wind factors and the scale of Enteromorpha green tide outbreak, a single factor long-term trend prediction model for the scale of Enteromorpha green tide outbreak in the target area was established, including: The correlation between the mean meridional wind value of the target sea area at different times and the annual maximum scale of the Enteromorpha green tide was analyzed, and the main control index of sea surface wind for the annual maximum scale prediction of the Enteromorpha green tide was selected; A single-factor long-term trend prediction model for the outbreak scale of Ulva green tide in the target area is established based on the main control index of sea surface wind.

8. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 6, characterized in that: Based on the correlation between precipitation factors and the scale of the green tide outbreak, a single factor long-term trend prediction model for the scale of the green tide outbreak in the target area was established, including: The correlation between precipitation values ​​in different periods and different sea areas and the annual maximum scale of Enteromorpha green tide was analyzed, and the main control indicators of precipitation for the annual maximum scale prediction of Enteromorpha green tide were selected; A single-factor long-term trend prediction model for the outbreak scale of Enteromorpha green tide in the target area is established based on the precipitation main control index.

9. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 6, characterized in that: The PSO-BP neural network model architecture is used to establish a multi-factor long-term trend prediction model for the outbreak scale of Enteromorpha in the target area. The construction process is as follows: (1) Normalize the data, establish a BP neural network, determine the topology and initialize the network weights and thresholds; (2) Initialize PSO parameters, including maximum number of iterations, population size, individual learning factor, social learning factor, and inertia weight parameter; (3) Initialize the population position of PSO and calculate the number of variable elements that need to be optimized based on the BP neural network structure; (4) PSO optimization, the fitness function is set to the mean square error predicted by the BP neural network, the PSO optimization process is cyclic, the position of the optimal particle is continuously updated until the maximum number of iterations is reached, and the PSO algorithm is terminated; The optimal weight threshold parameters optimized by the PSO algorithm were assigned to the BP neural network, and the optimal PSO-BP neural network model was output as a multi-factor trend prediction model for the outbreak scale of Enteromorpha green tide in the target area.

10. The method for predicting the outbreak scale of Enteromorpha enteromorpha based on meteorological and hydrological conditions according to claim 9, characterized in that: The tansig function is selected as the activation function of the BP neural network.

Images