Marine fishery habitat assessment method and device and storage medium

Abstract

The invention provides a marine fishery habitat assessment method and device and a storage medium. The method comprises the steps that acoustic data collected by a light-induced fishing boat shipborne fish finder in voyage times, fishery yield data of fishing boats homologous with the acoustic data and marine environment element data in a research sea area are obtained; processing the data to obtain a single target number, a fish school NASC and a full voyage NASC; establishing three optimal generalized additive models by taking the single target number, the fish school NASC and the full-voyage NASC as response variables and the environmental elements as explanatory variables, selecting an optimal response variable and calculating the weight of each environmental element in the model; combining the suitability index of each environment element based on the calculated weight, constructing two weighted habitat suitability index models, and selecting an optimal model; and calculating a habitat suitability index HSI value by adopting the optimal model, and verifying by means of yield data. According to the invention, the accuracy of northwest Pacific Ocean fishery habitat evaluation can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fishery habitat assessment and prediction, and in particular to a method, device and storage medium for assessing marine fishery habitat. Background Art

[0002] Most fish are phototactic and can gather in large numbers on the ocean surface at night under the illumination of appropriate artificial light sources. Light-induced fisheries based on this characteristic have now become an important part of offshore fisheries at home and abroad. The Northwest Pacific Ocean is rich in fishery resources and has formed an excellent fishing ground for economic fish such as mackerel, anchovies, sardines, saury and squid. It is not only one of the most important fishing areas in my country, but also the area with the highest marine fishery production in the world, with a large number of light-induced fishing boats operating here.

[0003] Although my country's offshore fisheries have achieved remarkable results after more than 30 years of development, the understanding of the fishing grounds in the Northwest Pacific is still insufficient. Suitable habitat is the basis for the existence of fishing grounds, and it usually represents ecological and environmental conditions that are favorable for the survival of species. Habitats change with sea surface temperature (SST), sea surface temperature gradient (SSTG), sea surface height (SSH), chlorophyll a concentration (Chl) and other oceanographic variables, which will significantly affect the population dynamics and distribution of offshore fish. When suitable habitats decrease or disappear, the abundance of species may decrease accordingly. Therefore, assessing the suitability of fish habitats has become a key link in the high-quality development of light-induced fisheries in the Northwest Pacific.

[0004] At present, many species distribution models, including polynomial regression models, maximum entropy models and neural network models, are used to analyze the habitat preferences of fish. Among these models, the fishing effort or catch per unit fishing effort obtained by distant-water fishing vessels is selected as the response variable for habitat modeling by most studies due to its low acquisition cost, wide spatiotemporal coverage and long time series. However, on the one hand, due to the inherent limitations of the above models, the model performance and prediction performance are relatively limited in some scenarios. On the other hand, the fishing effort and catch per unit fishing effort of commercial fisheries are not reliable relative abundance indices, because the distribution of fishing effort is not random, and fishermen tend to invest more fishing effort in locations where they think they may obtain high yields; and when both the catch and effort are high or low, the calculated catch per unit fishing effort may be low; only when the catch is high and the effort is low, the catch per unit fishing effort will be at a high level. Summary of the invention

[0005] In view of the defects in the prior art, the present invention, based on the commercial fish finders currently widely carried on Northwest Pacific light-lured fishing vessels, uses the acoustic data collected during their navigation and fishing operations to propose a method, equipment and storage medium that can more accurately reflect the distribution characteristics of suitable habitats for fishery resources in the Northwest Pacific and thereby improve the accuracy of habitat assessment in Northwest Pacific fishing grounds.

[0006] To solve the above problems, the technical solution of the present invention is:

[0007] A method for assessing marine fishery habitats comprises the following steps:

[0008] Acquire the acoustic data collected by the fish finder on board the light-attracting fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the data of marine environmental elements in the research area;

[0009] After processing the data, the number of individual targets, the NASC of the fish school and the NASC of the entire voyage are obtained;

[0010] Three optimal generalized additive models were established with the number of single targets, fish NASC, and full-course NASC as response variables and various environmental factors as explanatory variables. The optimal response variable was selected and the weights of various environmental factors in the model were calculated.

[0011] The suitability index of each environmental factor was combined based on the calculated weights, and two weighted habitat suitability index models were constructed, and the optimal model was selected.

[0012] The optimal model was used to calculate the habitat suitability index (HSI) value and verified with the help of production data.

[0013] Preferably, the step of obtaining the acoustic data collected by the fish finder onboard the light-luring fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the data of marine environmental elements in the study sea area specifically includes: obtaining the acoustic backscattering intensity data within a certain water depth range collected by the fish finder onboard the light-luring fishing vessel during the voyage, obtaining the fishery production data of the fishing vessel with the same source as the acoustic data, including the catch, longitude, latitude and date, and obtaining the data of marine environmental elements in the study sea area, wherein the environmental elements include sea surface temperature, sea surface temperature gradient, sea surface height and chlorophyll a concentration.

[0014] Preferably, in the step of obtaining the acoustic data collected by the fish finder on the light-attracting fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the marine environmental element data in the research sea area, the route length and the area of ​​the research sea area contained in the acoustic data are used to calculate whether the acoustic data meets the survey coverage requirement through the acoustic survey route coverage formula, and the calculation formula is: Where Λ is the coverage rate of the acoustic survey route, D is the length of the route, and A is the area of ​​the studied sea area.

[0015] Preferably, the step of obtaining the number of single targets, the NASC of fish schools and the NASC of the entire voyage after processing the data specifically includes: performing background noise removal, impulse noise removal and transmission noise removal preprocessing on the acoustic backscatter intensity data, and then using a single detection algorithm, a fish school detection algorithm and an echo integration method respectively to obtain the number of single targets, the NASC of fish schools and the NASC of the entire voyage.

[0016] Preferably, the steps of establishing three optimal generalized additive models with the number of single targets, NASC of fish schools, and NASC of the entire voyage as response variables and various environmental factors as explanatory variables, selecting the optimal response variable and calculating the weight of each environmental factor in the model specifically include: establishing three optimal generalized additive models with the number of single targets, NASC of fish schools, and NASC of the entire voyage as response variables and various environmental factors as explanatory variables, analyzing the significance of the influence of each environmental factor on the three response variables, and calculating the weight of each environmental factor in the model according to the deviation explanation rate, R 2 The optimal response variable was determined based on the significance results, and the suitability index SI of each environmental factor under different classification numbers was calculated using the optimal response variable. The optimal number of classifications for each environmental factor was selected according to the determination coefficient. The SI was recalculated based on the optimal number of classifications and a Gaussian function was constructed to describe the relationship between the SI and each environmental factor. The boosted regression tree algorithm was used to determine the weight of each environmental factor in the model according to the relative contribution rate of the environmental variables.

[0017] Preferably, in the step of combining the suitability index of each environmental factor based on the calculated weights, respectively constructing two weighted habitat suitability index models, and selecting the optimal model, the two weighted habitat suitability index HSI models are as follows:

[0018]

[0019]

[0020] Where, HSI A and HSI G are the HSI calculated based on the arithmetic mean method and the geometric mean method respectively; w i is the weight assigned to environmental factor i; n is the total number of environmental factors included in the HSI model.

[0021] Preferably, 80% of the data are randomly selected as the training set and the other 20% of the data are selected as the test set, and the prediction performance of the two models is evaluated by a cross-validation method to select the optimal model.

[0022] Preferably, the step of using the optimal model to calculate the HSI value of the habitat suitability index and verifying it with the help of production data specifically includes: using the optimal model to calculate the HSI value of the habitat suitability index and visualize it, comparing the difference between the average latitude and average longitude corresponding to the area with HSI>0.5 and the latitude center of gravity and longitude center of gravity of the catch, and the calculation formula of the latitude center of gravity and longitude center of gravity of the catch is as follows:

[0023]

[0024]

[0025] In the formula, G Lat and G Lon are the latitude and longitude centroids of the catch, Lat i and Lon i is the longitude and latitude of grid i, C i and is the total catch in grid i.

[0026] Furthermore, the present invention also provides a computer device, which includes a processor and a memory for storing executable instructions of the processor, and the processor is configured to perform the marine fishery habitat assessment method as described above by executing the executable instructions.

[0027] Furthermore, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store program codes, and the program codes are used to execute the marine fishery habitat assessment method as described above.

[0028] Compared with the prior art, the marine fishery habitat assessment method based on the fish finder data of light-luring fishing vessels of the present invention proposes to use the acoustic data collected during the navigation and operation of light-luring fishing vessels as a new parameter to describe the relative abundance of fish and apply it to habitat suitability modeling. It not only solves the shortcoming of poor randomness in the distribution of relative abundance index data such as fishing effort and catch per unit fishing effort in traditional methods, but also makes full use of the data of fish finders on board light-luring fishing vessels, reducing the required investigation cost and time and energy; uses a generalized additive model to screen key response variables in advance, determines the weight of each environmental factor in the habitat model with the help of a boosted regression tree, and verifies the effectiveness of the fish finder data of light-luring fishing vessels in modeling in combination with fishery production data, so that the assessment accuracy is effectively improved; through the comprehensive and integrated use of large-spatial-range and long-term series acoustic data, environmental data and fishery production data collected by all light-luring fishing vessels in a certain sea area, the habitat distribution of the Northwest Pacific fishing grounds is evaluated, which helps to accurately locate the location of the fishing grounds and promote the high-quality development of my country's Northwest Pacific light-luring fishery. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0030] Figure 1 It is a flowchart of the marine fishery habitat assessment method of the present invention;

[0031] Figure 2 It is a spatial distribution diagram of the scattering coefficient of the whole nautical mile area obtained by converting the acoustic data of the fish finder in a specific embodiment;

[0032] Figure 3 The SI curve diagram is obtained by Gaussian fitting of four environmental variables SST, SSTG, SSH and Chl in a specific embodiment;

[0033] Figure 4 In a specific embodiment, HSI A and HSI G The intercept, slope, and R obtained from 500 cross-validation of the model 2 and AIC plot;

[0034] Figure 5 It is a spatial distribution characteristic diagram of the saury HSI and the actual saury catch in the same voyage in a specific embodiment;

[0035] Figure 6 The figure is a graph showing the proportion of each HSI group and the corresponding saury catch in a specific embodiment. DETAILED DESCRIPTION

[0036] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0037] Specifically, the present invention provides a method for assessing marine fishery habitat based on fish finder data of light-induced fishing vessels, such as Figure 1 As shown, the method comprises the following steps:

[0038] S1: Acquire the acoustic data collected by the fish finder on the light-attracting fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the data of marine environmental elements in the research area;

[0039] Specifically, the acoustic backscattering intensity data within a certain water depth range collected by the fish finder onboard the light-luring fishing vessel during the voyage is obtained, the fishery production data of the fishing vessel with the same source as the acoustic data is obtained, including the catch, longitude, latitude and date, with a time resolution of 1 day, and the marine environmental element data in the study sea area are obtained, and the environmental elements include sea surface temperature (Sea surface temperature, SST), sea surface temperature gradient (Sea surface temperature gradient, SSTG), sea surface height (Sea surface height, SSH), chlorophyll a concentration (Chlorophyll aconcentration, Chl), with a time resolution of 1 day.

[0040] Using the route length and the area of ​​the research sea area contained in the acoustic data, the acoustic survey route coverage formula is used to calculate whether the acoustic data meets the survey coverage requirements. The calculation formula is as follows:

[0041]

[0042] Where Λ is the coverage rate of the acoustic survey route, D is the route length (in nmi), and A is the area of ​​the research sea area (in nmi 2 ).

[0043] S2: After processing the data, the number of single targets, the nautical area scattering coefficient (NASC) of the fish school and the NASC of the whole voyage are obtained;

[0044] Specifically, after preprocessing the acoustic backscatter intensity data by removing background noise, impulse noise, and transmission noise, the single-target detection algorithm, fish school detection algorithm, and echo integration method were used to obtain the number of single-target targets, fish school NASC, and full-range NASC.

[0045] The number of individual targets, fish school NASC, full-voyage NASC, catch and environmental data are gridded, and the corresponding catch and environmental data are extracted and accurately matched according to the date and longitude and latitude corresponding to the grids of the number of individual targets, fish school NASC and full-voyage NASC.

[0046] S3: Three optimal generalized additive models were established with the number of single targets, fish NASC, and full-course NASC as response variables and various environmental factors as explanatory variables. The optimal response variable was selected and the weights of various environmental factors in the model were calculated.

[0047] Specifically, the variance inflation factor was used to determine whether there was multicollinearity between environmental factors. Then, three optimal generalized additive models were established with the number of single targets, fish NASC, and full-course NASC as response variables and each environmental factor as an explanatory variable. The significance of the impact of each environmental factor on the above three response variables was analyzed. The deviation explanation rate, R 2 The significant results determine the response variables that can be used for subsequent habitat modeling, i.e., the optimal response variables;

[0048] The best response variable is used to calculate the suitability index (SI) of each environmental factor under different classification numbers, and the optimal number of classifications for each environmental factor is selected according to the determination coefficient; SI is recalculated based on the optimal number of classifications and a Gaussian function is constructed to describe the relationship between SI and each environmental factor;

[0049] The boosted regression tree algorithm was used to determine the weight of each environmental factor in the habitat model to be established according to the relative contribution rate of environmental variables.

[0050] S4: Combine the suitability index of each environmental factor based on the calculated weights, construct two weighted habitat suitability index (HSI) models, and select the optimal model;

[0051] Specifically, the SI of each environmental factor was combined based on the calculated weights to construct two weighted habitat suitability index HSI models, which are as follows:

[0052]

[0053]

[0054] Where, HSI A and HSI G are the HSI calculated based on the arithmetic mean method and the geometric mean method respectively; w i is the weight assigned to environmental factor i; n is the total number of environmental factors included in the HSI model;

[0055] 80% of the data were randomly selected as the training set and the other 20% of the data were used as the test set. The cross-validation method was used to calculate the regression intercept, slope, and R 2 The prediction performance of the two models was evaluated by the Akaike information criterion (AIC) and the optimal model was selected.

[0056] S5: Calculate the HSI value of the habitat suitability index using the optimal model and verify it with the help of production data;

[0057] Specifically, the optimal model was used to calculate and visualize the HSI value of the habitat suitability index, and the HSI was divided into 5 groups (0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1). The effectiveness of the optimal HSI model was tested according to the proportion of catch in each group. The difference between the average latitude and average longitude corresponding to the area with HSI>0.5 and the latitude and longitude centroids of the catch was compared. The latitude and longitude centroids of the catch were calculated as follows:

[0058]

[0059] In the formula, G Lat and G Lon are the latitude and longitude centroids of the catch, Lat i and Lon i is the longitude and latitude of grid i, C i and is the total catch in grid i.

[0060] The marine fishery habitat assessment method of the present invention is described below with a specific embodiment, comprising:

[0061] (1) Acquire the acoustic backscatter intensity data within a certain water depth range collected by the onboard fish finder of the light-attracting fishing vessel during the voyage; obtain the fishery production data of the fishing vessel with the same source as the acoustic data; obtain the data of marine environmental elements in the research sea area. The acoustic backscatter intensity data in this embodiment comes from the acoustic backscatter intensity data of 5 to 88 m and 28 kHz collected by the onboard fish finder during the saury side-lifting operation of the Chinese distant-water saury light-attracting fishing vessel "Minghua" from May 31, 2021 to July 17, 2021 in the northwest Pacific Ocean at 40.4-48.2°N, 163.7-171.9°E. The saury production data of this voyage comes from the Pacific Saury Technical Group of the China Ocean Fisheries Association. The data spans from May 31, 2021 to July 17, 2021, with a time resolution of 1 day. Environmental factors SST, SSH and Chl data were obtained by downloading public satellite remote sensing data with a time resolution of 1 day. SSTG was calculated using the gradient amplitude formula, as follows:

[0062]

[0063] Where i and j are positive integers, representing the row and column numbers of the data grid, respectively; SSTG i,j is the SSTG value of the current grid; SST i+1,j ,SST i-1,j ,SST i,j+1 and SST i,j-1Represents the SST values ​​of the four grid points above, below, left and right adjacent to the current grid; Δx represents the longitude interval between the j+1 column and the j-1 column; Δy represents the latitude interval between the i+1 row and the i-1 row.

[0064] (2) Using the route length and the area of ​​the research sea area contained in the acoustic data, the acoustic survey route coverage formula is used to calculate whether the acoustic data meets the survey coverage requirements. The formula is: In the formula, Λ is the coverage rate of the acoustic survey route, D is the route length, and A is the area of ​​the studied sea area. In this embodiment, the route length of the light-lured fishing boat for saury is 2753 nmi, and the area of ​​the investigated sea area is 164887 nmi 2 , calculated according to the coverage formula, the coverage is 6.78, which meets the survey requirements.

[0065] (3) After preprocessing the acoustic backscatter intensity data by removing background noise, pulse noise, and transmission noise, the single-target detection algorithm, fish school detection algorithm, and echo integration method are used to obtain the number of single-target targets, fish school NASC, and full-range NASC.

[0066] Using the single-body detection algorithm, the minimum threshold is set to -60dB to detect the number of single fish in the water layer of 5 to 88m. Using the fish school detection algorithm, the minimum threshold is set to -45dB to detect the number of fish schools in the water layer of 5 to 88m, and the echo integration is used to calculate the NASC value of each fish school. The basic integration unit is set to 1nmi in the horizontal direction and 88m in the vertical direction, and the integration threshold is set to -76dB to integrate the backscatter intensity data of the entire range of 5 to 88m.

[0067] (4) The number of individual targets, fish school NASC, full-voyage NASC, catch, and environmental data were gridded at a spatial resolution of 0.1°×0.1°. The corresponding catch and environmental data were extracted and accurately matched based on the date and longitude and latitude corresponding to the grids of the number of individual targets, fish school NASC, and full-voyage NASC.

[0068] (5) The variance inflation factor was used to determine whether there was multicollinearity between environmental factors. Then, three optimal generalized additive models were established with the number of individual targets, fish NASC, and full-course NASC as response variables and each environmental factor as an explanatory variable. The significance of the impact of each environmental factor on the above three response variables was analyzed. The optimal model was used to determine the deviation explanation rate, R 2 The significant results determine the response variables that can be used for subsequent habitat modeling, i.e., the optimal response variables;

[0069] The generalized additive models of the three response variables SST, SSH, SSTG and Chl that passed the multicollinearity test were constructed in turn by stepwise regression method. The Akaike information criterion AIC was used to balance the model goodness of fit and complexity, and the optimal models of each response variable were obtained as follows:

[0070] lnTarget=s(SSH)+s(SST)+ε

[0071] In the formula, Target represents the number of single targets. The AIC of the optimal model is 813.2856, the deviation explanation rate is 12%, and R 2 =0.0921. Environmental factors SST and SSH were not significantly correlated with Target (p>0.05). They were not suitable as response variables for subsequent habitat modeling.

[0072] lnSchool=s(SSH)+s(SST)+ε

[0073] In the formula, School represents the NASC of the fish school. The AIC of the optimal model is 447.5274, the deviation explanation rate is 36.7%, and R 2 =0.357. Only the environmental factor SSH was significantly correlated with School (p<0.001). It is not suitable as a response variable for subsequent habitat modeling.

[0074] lnNASC=s(SST)+s(SSTG)+s(SSH)+s(Chl)+ε

[0075] Where NASC represents the NASC of the entire flight. The AIC of the optimal model is 1194.64, the deviation explanation rate is 44.5%, and R 2 =0.414. Environmental factors SST, SSTG, SSH and Chl are all significantly correlated with NASC (p<0.001). They are suitable as response variables for subsequent habitat modeling, so the full-course NASC is the optimal response variable. Its distribution is shown in the figure below. Figure 2 shown.

[0076] (6) The boosted regression tree algorithm is used to determine the weight of each environmental factor in the habitat model to be established based on the relative contribution rate of environmental variables.

[0077] In this embodiment, the parameters of the boosted regression tree algorithm: the number of trees, the depth of the tree, the learning rate, the minimum number of observations and the bagging fraction are set to 1000, 5, 0.05, 20 and 0.75 respectively. The optimal boosted regression tree model is established using this set of parameters, and the SSH weight is calculated to be 46.23%, the SST weight is 26.87%, the Chl weight is 19.10%, and the SSTG is 7.8%.

[0078] (7) Use the full-range NASC to calculate the suitability index SI of each environmental factor under different classification numbers, and select the optimal number of classifications for each environmental factor based on the determination coefficient; recalculate the SI based on the optimal number of classifications and construct a Gaussian function to describe the relationship between the SI and each environmental factor.

[0079] SI is calculated as follows:

[0080]

[0081] Where, SI i,k is the SI of the kth category of environmental variable i; NASC i,k is the average NASC of the kth category of environmental variable i; NASC i,min and NASC i,max are the minimum and maximum values ​​of the average NASC corresponding to environmental variable i, respectively.

[0082] The number of groups is set to 10 to 20. According to the maximum value of the determination coefficient, the optimal number of classifications for SST, SSTG, SSH and Chl can be determined to be 12, 11, 14 and 10 groups respectively;

[0083] Recalculate SI and construct Gaussian function as follows:

[0084]

[0085] Where, SI i is the SI of environmental variable i; a and b are parameters to be estimated and solved by the least squares method; X is the value of the environmental variable.

[0086] SI values ​​range from 0 to 1. Since saury gather in a wide range during migration, SI values ​​above 0.5 are defined as suitable sea areas and obtained. Figure 3 The suitability index curves and optimal ranges of SST, SSTG, SSH and Chl are shown.

[0087] (8) Based on the weights calculated in step (6), the SI of each environmental factor is combined to construct two weighted habitat suitability index HSI models, respectively. The two models are as follows:

[0088]

[0089] Where, HSI A and HSI G are the HSI calculated based on the arithmetic mean method and the geometric mean method respectively; w i is the weight assigned to environmental factor i according to the calculation result of step (6); n is the total number of environmental factors included in the HSI model, and in this embodiment, n=4.

[0090] (9) Randomly select 80% of the data as the training set and the other 20% of the data as the test set, and use the cross-validation method to calculate the regression intercept, slope, and R 2 And AIC are used to evaluate the prediction performance of the two models described in step (8) and select the optimal model.

[0091] About HSI A Model and HSI G The model was cross-validated 500 times. In each cross-validation, the HSI prediction value calculated based on the training set and the HSI observation value based on the test set were subjected to linear regression analysis, and the degree of approximation between the two was described by constructing a linear relationship.

[0092] In this embodiment, from Figure 4 According to the values ​​shown in A The model intercept is closer to 0, the slope is closer to 1, and R 2 Higher and lower AIC value, HSI A The model is the optimal model.

[0093] (10) Using the optimal model HSI described in step (9) A The model calculates and visualizes discrete HSI values, and divides HSI into five groups (0-0.2, 0.2-0.4, 0.4-0.6, 0.6-0.8, 0.8-1). The effectiveness of the optimal HSI model is tested based on the proportion of catch in each group. The difference between the average latitude and average longitude corresponding to the area with HSI>0.5 and the latitude and longitude centroids of the catch is compared. The latitude and longitude centroids of the catch are calculated as follows:

[0094]

[0095] In the formula, G Lat and G Lon are the latitude and longitude centroids of the catch, Lat i and Lon i is the longitude and latitude of grid i, C i and is the total catch in grid i.

[0096] like Figure 5It is shown that the proportion of suitable habitats for saury (HSI>0.5) in the study area is 27.38%, and 38.32% of the areas have HSI values ​​between 0.2 and 0.4; among the suitable habitats, the areas with HSI values ​​between 0.6 and 0.8 are the largest, reaching 22.72%; the habitat suitability of saury in the south is significantly higher than that in the middle and north. Specifically, these suitable habitats are mainly located in the area south of 44°N and west of 172°E; there are two HSI high value areas (HSI>0.8) at 42.2°N, 166.5°E and 43.3°N, 168.5°E, respectively, accounting for 1.33% of the entire study area;

[0097] Most high catches occurred in areas with higher HSI, and the two were in good agreement; the average latitude of suitable habitats with HSI>0.5 was 42.4°N, the average longitude was 168.5°E, and the latitude centroid of catch was 44.1°N, and the longitude centroid was 168.4°E.

[0098] like Figure 6 As shown in the figure, according to the different HSI groups and the corresponding catch percentages of each group, the catch tends to increase with the stratification of HSI values, and the catch caught in the sea areas with higher HSI is also higher; most of the catch comes from areas with HSI>0.5, accounting for 61.23% of the total catch, among which the catch in the HSI range of 0.6-0.8 accounts for the largest proportion, which is 32.6%.

[0099] According to another aspect of an embodiment of the present application, a computer device is provided, which includes a processor and a memory for storing executable instructions of the processor, and the processor is configured to execute the marine fishery habitat assessment method described in the embodiment of the present application by executing the executable instructions.

[0100] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the marine fishery habitat assessment method described in the embodiments of the present application.

[0101] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A method for assessing marine fishery habitats, characterized in that: The method comprises the following steps: Acquire the acoustic data collected by the fish finder on board the light-attracting fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the data of marine environmental elements in the research area; After processing the data, the number of individual targets, the NASC of the fish school and the NASC of the entire voyage are obtained; Three optimal generalized additive models were established with the number of single targets, fish NASC, and full-course NASC as response variables and various environmental factors as explanatory variables. The optimal response variable was selected and the weights of various environmental factors in the model were calculated. The suitability index of each environmental factor was combined based on the calculated weights, and two weighted habitat suitability index models were constructed, and the optimal model was selected. The optimal model was used to calculate the habitat suitability index (HSI) value and verified with the help of production data.

2. The marine fishery habitat assessment method according to claim 1, characterized in that: The steps of obtaining the acoustic data collected by the fish finder onboard the light-luring fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the marine environmental element data in the study sea area specifically include: obtaining the acoustic backscattering intensity data within a certain water depth range collected by the fish finder onboard the light-luring fishing vessel during the voyage, obtaining the fishery production data of the fishing vessel with the same source as the acoustic data, including the catch, longitude, latitude and date, and obtaining the marine environmental element data in the study sea area, wherein the environmental elements include sea surface temperature, sea surface temperature gradient, sea surface height and chlorophyll a concentration.

3. The marine fishery habitat assessment method according to claim 1, characterized in that: In the step of obtaining the acoustic data collected by the fish finder on the light-attracting fishing vessel during the voyage, the fishery production data of the fishing vessel with the same source as the acoustic data, and the marine environmental element data in the research sea area, the route length and the area of ​​the research sea area contained in the acoustic data are used to calculate whether the acoustic data meets the survey coverage requirement through the acoustic survey route coverage formula, and the calculation formula is: Where Λ is the coverage rate of the acoustic survey route, D is the length of the route, and A is the area of ​​the studied sea area.

4. The marine fishery habitat assessment method according to claim 1, characterized in that: The step of obtaining the number of single targets, the NASC of fish schools and the NASC of the entire voyage after processing the data specifically includes: performing background noise removal, impulse noise removal and transmission noise removal preprocessing on the acoustic backscatter intensity data, and then using a single detection algorithm, a fish school detection algorithm and an echo integration method to obtain the number of single targets, the NASC of fish schools and the NASC of the entire voyage.

5. The marine fishery habitat assessment method according to claim 1, characterized in that: The method comprises the following steps: establishing three optimal generalized additive models with the number of single target, NASC of fish school and NASC of whole voyage as response variables and various environmental factors as explanatory variables, selecting the optimal response variable and calculating the weight of each environmental factor in the model, and specifically comprising: establishing three optimal generalized additive models with the number of single target, NASC of fish school and NASC of whole voyage as response variables and various environmental factors as explanatory variables, analyzing the significance of the influence of each environmental factor on the three response variables, and calculating the weight of each environmental factor in the model according to the deviation explanation rate, R 2 The optimal response variable was determined based on the significance results, and the suitability index SI of each environmental factor under different classification numbers was calculated using the optimal response variable. The optimal number of classifications for each environmental factor was selected according to the determination coefficient. The SI was recalculated based on the optimal number of classifications and a Gaussian function was constructed to describe the relationship between the SI and each environmental factor. The boosted regression tree algorithm was used to determine the weight of each environmental factor in the model according to the relative contribution rate of the environmental variables.

6. The marine fishery habitat assessment method according to claim 1, characterized in that: In the step of combining the suitability index of each environmental factor based on the calculated weights, respectively constructing two weighted habitat suitability index models, and selecting the optimal model, the two weighted habitat suitability index HSI models are as follows: Where, HSI A and HSI G are the HSI calculated based on the arithmetic mean method and the geometric mean method respectively; w i is the weight assigned to environmental factor i; n is the total number of environmental elements included in the HSI model.

7. The marine fishery habitat assessment method according to claim 6, characterized in that: 80% of the data were randomly selected as the training set and the other 20% of the data were used as the test set. The prediction performance of the two models was evaluated using the cross-validation method and the optimal model was selected.

8. The marine fishery habitat assessment method according to claim 1, characterized in that: The step of using the optimal model to calculate the HSI value of the habitat suitability index and verifying it with the help of production data specifically includes: using the optimal model to calculate the HSI value of the habitat suitability index and visualize it, comparing the difference between the average latitude and average longitude corresponding to the area with HSI>0.5 and the latitude center of gravity and longitude center of gravity of the catch, and the calculation formula of the latitude center of gravity and longitude center of gravity of the catch is as follows: In the formula, G Lat and G Lon are the latitude and longitude centroids of the catch, Lat i and Lon i is the longitude and latitude of grid i, C i and is the total catch in grid i.

9. A computer device, characterized in that: The computer device includes a processor and a memory for storing executable instructions of the processor, and the processor is configured to perform the marine fishery habitat assessment method according to any one of claims 1 to 8 by executing the executable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program code, and the program code is used to execute the marine fishery habitat assessment method described in any one of claims 1-8.

Images