A three-layer nested hydrodynamic and ecological coupling modeling method

Abstract

The present application relates to the technical field of numerical simulation of marine environment, and discloses a three-layer nested water power and ecological coupling modeling method, comprising: S1, constructing an outer model of the Beibu Gulf and calculating; S2, constructing a middle model of the Guangxi coastal waters and a boundary connection; S3, constructing an inner model of the Qinzhou Bay or a typical harbor and a boundary connection; S4, performing online coupling of water power and ecology; S5, performing parameter calibration and result verification; and S6, outputting three-layer nested results. Through three-layer one-way nesting, step-by-step boundary transmission and online coupling of water power and ecology, the present application realizes continuous connection between different spatial scale models, and improves the simulation accuracy and reliability of the water power field, the temperature and salinity field and the ecological variable field in the coastal and harbor areas.

Description

Technical Field

[0001] This invention relates to the field of technology, and in particular to a three-layer nested hydrodynamic and ecological coupling modeling method. Background Technology

[0002] Hydrodynamic and ecological processes in nearshore waters and typical harbors are jointly influenced by factors such as ocean currents, meteorological forcing, river runoff, temperature-salinity structure, and nutrient input, exhibiting significant spatial scale differences and process coupling characteristics. For areas such as the Beibu Gulf, the coastal waters of Guangxi, and Qinzhou Bay or typical harbors, it is necessary to reflect the impact of the large-scale ocean background field on nearshore areas, while also obtaining high-resolution hydrodynamic, temperature-salinity, and ecological variable fields within the harbor areas. Therefore, establishing continuous and reliable hydrodynamic and ecological simulation relationships across different spatial scales is a crucial problem that needs to be addressed in modeling such marine areas.

[0003] In existing technologies, hydrodynamic models or hydrodynamic-ecological coupled models with a single computational domain are commonly used for simulation. While large-scale models can cover offshore and nearshore areas and reflect regional-scale background changes in tidal currents, temperature, and salinity, their limited grid resolution makes it difficult to finely characterize the distribution of local hydrodynamic and ecological variables near harbors, estuaries, and complex coastlines. Conversely, while small-scale high-resolution models can improve the local simulation accuracy of harbor areas, their open boundary conditions often rely on external data or coarse-resolution data, making it difficult to guarantee the continuous transmission of hydrodynamic and ecological variables from the offshore to the nearshore and from the nearshore to the harbor.

[0004] Furthermore, in the existing modeling process, there is sometimes a lack of sufficient online coupling between hydrodynamic calculations and ecological variable calculations. The response process, horizontal transport process, and vertical mixing and diffusion process of ecological variables cannot be coordinated and completed under the same computational framework. This can easily lead to a mismatch between the simulation results of ecological variables such as nutrients and chlorophyll and hydrodynamic environmental variables such as water level, flow velocity, temperature, and salinity, thereby affecting the simulation accuracy of the ecological variable field.

[0005] Meanwhile, the marine areas, observational data, and dominant processes corresponding to models at different scales vary. If uniform parameters are used for the outer, middle, and inner models, or if regional validation is lacking, the model may perform well at a certain scale or in a certain region, but exhibit significant simulation deviations in other regions. Therefore, a three-layer nested modeling method is needed that can progressively transfer the large-scale background field to nearshore and harbor areas, and perform online coupling of hydrodynamics and ecology, regional calibration, and validation in each model layer. Summary of the Invention

[0006] The purpose of this invention is to solve the problem of difficulty in balancing regional continuity and local accuracy in the prior art, and to propose a three-layer nested hydrodynamic and ecological coupling modeling method.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a three-layer nested hydrodynamic and ecological coupling modeling method, comprising the following steps: S1: Construct an outer unstructured grid model covering the Beibu Gulf, inputting data on the outer sea boundary, meteorological forcing, river runoff, and ecological observations, and calculate the outer hydrodynamic field, temperature and salinity field, and ecological variable field at the first resolution; S2: Construct a mid-level unstructured grid model covering the nearshore waters of Guangxi, interpolate the output of the outer-level model as the open boundary conditions of the mid-level model, and calculate the mid-level hydrodynamic field, temperature and salinity field and ecological variable field at the second resolution. S3: Construct an inner-layer unstructured grid model covering Qinzhou Bay or a typical harbor, interpolate the output of the middle-layer model as the open boundary conditions of the inner-layer model, and calculate the inner-layer hydrodynamic field, temperature and salinity field and ecological variable field at the third resolution. S4: Perform online coupling of hydrodynamics and ecology in the outer, middle and inner layer models respectively; S5: Perform parameter calibration and result verification for the outer layer model, middle layer model, and inner layer model respectively; S6: Output the result of a three-level nested structure.

[0008] The beneficial effects of the technical solution provided by this invention include at least the following: The invention constructs an outer, middle, and inner unstructured grid model covering the Beibu Gulf, the coastal waters of Guangxi, and Qinzhou Bay or typical harbors. The three-layer model is nested unidirectionally from the outside to the inside, with the resolution increasing step by step. This allows it to simultaneously meet the needs of large-scale offshore background fields and high-resolution simulation of local harbors, solving the problem that a single model cannot balance computational range and local accuracy.

[0009] This invention interpolates the output of the outer layer model as the open boundary condition of the middle layer model, and interpolates the output of the middle layer model as the open boundary condition of the inner layer model. The interpolation includes planar interpolation, temporal interpolation, and vertical interpolation. This enables the continuous transfer of tidal level, current velocity, temperature, salinity, and ecological variables between different levels of models, and improves the adaptability between the open boundary conditions of the sub-layer model and the output of the parent layer model.

[0010] This invention couples hydrodynamics and ecology online in the outer, middle, and inner models, respectively. This allows the hydrodynamic model to transmit water level, three-dimensional velocity, temperature, salinity, water depth, and vertical diffusion coefficient to the ecological dynamics model, and the ecological dynamics model to return ecological variable response terms. This enables the biogeochemical reactions, horizontal transport, and vertical mixing and diffusion of ecological variables to be completed in a coordinated manner in the same calculation process, thereby improving the consistency between the ecological variable field and the hydrodynamic environmental variables.

[0011] This invention calibrates parameters and verifies results using observational data from corresponding regions for the outer, middle, and inner layer models, and outputs water level, velocity, temperature, salinity, nutrient, and chlorophyll fields for the outer, middle, and inner layers, respectively. This improves the reliability of model results at different spatial scales and provides a data foundation for hydrodynamic process analysis, ecological variable simulation, and ecological environment assessment in nearshore waters and typical harbors. Attached Figure Description

[0012] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is an overall flowchart of the method provided in the embodiments of the present invention;

[0014] Figure 2 A schematic diagram of a three-layer nested hydrodynamic and ecological coupling model provided in an embodiment of the present invention. Detailed Implementation

[0015] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a three-layer nested hydrodynamic and ecological coupling modeling method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0017] The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0018] The following description, in conjunction with the accompanying drawings, details the specific scheme of the three-layer nested hydrodynamic and ecological coupling modeling method provided by this invention.

[0019] This invention aims to address the shortcomings of existing single-scale marine models, which struggle to simultaneously accommodate both large-scale open ocean background fields and the high-resolution simulation requirements of local harbor areas. It also addresses the insufficient integration between existing hydrodynamic and ecological variable calculations, leading to inadequate simulation accuracy for hydrodynamic, temperature-salinity, and ecological variable fields in nearshore and harbor areas. This invention constructs a three-layered nested model (outer, middle, and inner layers) and couples hydrodynamic and ecological data online within each layer. This allows hydrodynamic and ecological background information at the open ocean scale to be progressively transferred to nearshore and harbor areas, resulting in simulation results with progressively increasing resolution.

[0020] Please see Figure 1 and Figure 2 This illustrates a three-layer nested hydrodynamic and ecological coupling modeling method provided by an embodiment of the present invention, comprising the following steps: S1: Construct an outer unstructured grid model covering the Beibu Gulf, inputting data on the outer sea boundary, meteorological forcing, river runoff, and ecological observations, and calculate the outer hydrodynamic field, temperature and salinity field, and ecological variable field at the first resolution.

[0021] Specifically, the outer unstructured grid model refers to a numerical model that uses the Beibu Gulf as the computational region and divides the water space using irregular triangular grids. Unlike regular rectangular grids, unstructured grids can flexibly vary in size and shape according to the location of coastlines, islands, estuaries, shoals, and open boundaries, making them suitable for describing marine areas like the Beibu Gulf with complex coastlines and significant nearshore topographic variations. The outer sea boundary refers to the open boundary where the outer model connects to the external sea area. At this boundary, conditions representing the influence of the external sea, such as tide level, temperature, and salinity, need to be input. Meteorological forcing refers to the driving effect of atmospheric factors such as wind speed, wind direction, air pressure, temperature, and radiation on the sea surface. River runoff refers to the volume of river water flowing into the computational sea area and its freshwater impact. Ecological observation data refers to field observation or survey data related to ecological variables such as nitrates, phosphates, silicates, ammonium nitrogen, and chlorophyll.

[0022] It is worth noting that the hydrodynamic field mainly includes the spatial distribution and temporal variation of physical quantities such as water level, flow velocity, and flow direction; the temperature and salinity field mainly includes the distribution of temperature and salinity in the horizontal and vertical directions; and the ecological variable field mainly includes the distribution of ecological variables such as nutrients and chlorophyll within the model area. The first resolution is the spatial resolution of the outer layer model, which is used to reflect the large-scale background field of the Beibu Gulf. It is not required to reach the finest resolution at the harbor scale, but it should be able to provide continuous and stable boundary conditions for the middle layer model.

[0023] In one specific implementation, the Beibu Gulf is used as the calculation area for the outer model. An unstructured triangular mesh is established using coastline and water depth data. Tidal boundaries can be obtained using tidal boundary data generation tools, temperature and salinity boundaries can be obtained from external ocean reanalysis data, meteorological forcing can be achieved using hourly or daily meteorological reanalysis data, and river runoff can be obtained using measured runoff, monthly average runoff, or climatological runoff. The above data are input into the outer model after time standardization, unit standardization, and outlier removal. After the model runs, results such as water level, three-dimensional flow velocity, temperature, salinity, nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll are obtained within the outer model area.

[0024] When performing step S1, it is essential to ensure that the time ranges of the offshore boundary, meteorological forcing, river runoff, and ecological observation data correspond to each other. When the time resolutions of different data sources are inconsistent, they can be unified to the input time interval required by the model. When the spatial locations of different data sources are inconsistent, they can be mapped to outer model grid nodes or grid cells. For anomalous data that significantly deviates from the reasonable physical or ecological range, they can be removed before being input into the model to avoid abnormal inputs affecting model stability and simulation results.

[0025] S2: Construct a mid-layer unstructured grid model covering the nearshore waters of Guangxi, interpolate the output of the outer layer model as the open boundary conditions of the mid-layer model, and calculate the mid-layer hydrodynamic field, temperature and salinity field and ecological variable field with second resolution.

[0026] Specifically, the intermediate-layer unstructured mesh model refers to a sub-layer model with the Guangxi coastal waters as its computational domain. This intermediate-layer model is located within the computational domain of the outer-layer model, with a smaller spatial extent but a higher mesh resolution. Open boundary conditions refer to the external driving conditions required by the intermediate-layer model at its open boundaries, including tidal level, current velocity, temperature, salinity, and ecological variables. Since the boundary nodes, time steps, and vertical stratification of the intermediate-layer model typically differ from those of the outer-layer model, the output results of the outer-layer model need to be interpolated and converted into open boundary conditions that can be directly read by the intermediate-layer model.

[0027] It is worth noting that the output of the outer model in step S2 is not simply used as reference data, but rather as the actual driving conditions for the middle model. In this way, the hydrodynamic field, temperature and salinity field, and ecological variable field of the middle model can inherit the large-scale offshore background influences reflected by the outer model, while simultaneously utilizing its higher grid resolution to characterize local processes in the Guangxi coastal area. The second resolution being higher than the first resolution means that the grid scale of the middle model is smaller than that of the outer model, enabling the acquisition of more refined spatial simulation results.

[0028] In one specific implementation, after the outer model is completed, the results of water level, flow velocity, temperature, salinity, and ecological variables covering the open boundary of the Guangxi coastal area are selected from the outer model. Then, based on the location of the open boundary node of the middle model, the corresponding neighboring grid nodes or grid cells in the outer model are found, and the outer model results are transferred to the open boundary of the middle model. If the output time of the outer model is different from the boundary update time of the middle model, time interpolation is performed on the results of adjacent times in the outer model; if the vertical stratification of the outer model is different from that of the middle model, vertical interpolation is performed on the temperature, salinity, flow velocity, and ecological variables.

[0029] When performing step S2, it is important to ensure that the coordinate datum, vertical datum, and time datum of the outer layer model and the middle layer model remain consistent. For two-dimensional variables such as tide level, the main transformations are the boundary node positions and time directions; for three-dimensional variables such as flow velocity, temperature, salinity, and ecological variables, a vertical layer transformation is also required. The boundary file after transformation should be consistent with the mesh file, time file, and variable names of the middle layer model to ensure that the middle layer model can be stably read and run.

[0030] S3: Construct an inner-layer unstructured grid model covering Qinzhou Bay or a typical harbor, interpolate the output of the middle-layer model as the open boundary conditions of the inner-layer model, and calculate the inner-layer hydrodynamic field, temperature and salinity field and ecological variable field at the third resolution.

[0031] Specifically, the resolution model. The inner layer model is located within the computational domain of the middle layer model. Its computational range is further reduced, but its spatial resolution is further improved. It is used to simulate fine hydrodynamic processes, temperature and salinity distribution processes, and ecological variable changes near harbors, estuaries, and complex shorelines. The third resolution is higher than the second resolution, meaning that the grid scale of the inner layer model is smaller than that of the middle layer model, enabling it to represent the local topography and shoreline influences of harbors in greater detail.

[0032] It is noteworthy that the open boundary conditions of the inner model are derived from the output of the middle-layer model, rather than directly using coarse-resolution external data. This allows the inner model to inherit both the large-scale background influences of the outer model and the nearshore process influences of the middle-layer model, thereby improving the continuity and reliability of the simulation results for the harbor area. The inner model can focus on outputting the distribution of variables such as water level, flow velocity, temperature, salinity, nutrients, and chlorophyll in the harbor interior and adjacent areas.

[0033] In one specific implementation, Qinzhou Bay is used as the calculation region for the inner layer model. A local unstructured triangular mesh is established based on the bay's shoreline, water depth, channel, estuary, and open boundary location. After the middle layer model is run, the tidal level, current velocity, temperature, salinity, and ecological variables near the open boundary of the inner layer model are extracted from the middle layer model and interpolated in the plane, time, and vertical directions to generate the open boundary file of the inner layer model. The inner layer model runs under this open boundary condition to obtain a higher resolution hydrodynamic field, temperature and salinity field, and ecological variable field within the bay area.

[0034] When performing step S3, it is important to note that the inner layer model mesh can be appropriately densified in areas such as narrow waterways in harbors, estuaries, shoreline bends, and regions with significant water depth variations to improve local simulation capabilities. Simultaneously, the open boundary of the inner layer model should not be set in locations with excessively drastic hydrodynamic changes or overly complex topography to minimize the impact of boundary errors on the inner layer computational domain. The initial temperature, salinity, and ecological variable fields of the inner layer model can be obtained through interpolation of the corresponding time-series results from the middle layer model, or processed observation data or historical simulation results can be used as the initial fields.

[0035] S4: Perform online coupling of hydrodynamics and ecology in the outer, middle and inner models respectively.

[0036] Specifically, online coupling of hydrodynamics and ecology refers to the exchange of variables between the hydrodynamic model and the ecological dynamics model during the same computational process, rather than running the hydrodynamic model separately first and then inputting the hydrodynamic results offline into the ecological model. The hydrodynamic model is responsible for calculating hydrodynamic environmental variables such as water level, three-dimensional flow velocity, temperature, salinity, water depth, and vertical diffusion coefficient; the ecological dynamics model is responsible for calculating the biogeochemical reaction processes of ecological variables based on the above hydrodynamic environmental variables and conditions such as light and nutrients. Online coupling ensures that the reaction processes of ecological variables are consistent with the hydrodynamic transport processes in time.

[0037] It is noteworthy that in this invention, the relationship between the hydrodynamic model and the ecodynamic model is not a simple post-processing relationship, but rather a continuous exchange of variables during model operation. The hydrodynamic model transmits environmental conditions affecting ecological processes to the ecodynamic model, which in turn returns ecological variable response terms. These ecological variable response terms can be understood as the increases or decreases in ecological variables caused by biogeochemical processes such as growth, consumption, transformation, death, and remineralization. Subsequently, the hydrodynamic model performs horizontal transport and vertical mixing and diffusion of ecological variables based on flow velocity and diffusion conditions.

[0038] In one specific implementation, a finite-volume coastal-ocean hydrodynamic model can be used as the hydrodynamic model, and an ecological dynamics model can be used as the ecological calculation module. Online coupling between the two can be achieved through an aquatic biogeochemical model framework. During model execution, the hydrodynamic model calculates water level, flow velocity, temperature, salinity, and vertical diffusion coefficient in each preset calculation step or coupling step, and transmits these variables to the ecological dynamics model. The ecological dynamics model then calculates the reaction terms for ecological variables such as nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll. The hydrodynamic model then updates the concentrations of ecological variables based on the reaction terms and completes the calculations for horizontal transport and vertical mixing and diffusion.

[0039] When performing step S4, it is essential to ensure that the hydrodynamic model and the ecological dynamics model use the same or corresponding grid, time axis, and vertical layering. If the computation time step of the ecological dynamics model differs from that of the hydrodynamic model, a reasonable coupling time interval should be set to ensure the stability of variable exchange. Furthermore, it is necessary to check whether the units, variable names, and boundary input formats of the ecological variables are consistent with the model configuration file to avoid errors in calculation results due to inconsistencies in variable units or names.

[0040] S5: Perform parameter calibration and result verification for the outer layer model, middle layer model, and inner layer model respectively.

[0041] Specifically, parameter calibration refers to adjusting some parameters in the model based on observational data to make the model's calculation results as close as possible to the actual observation results. Result validation refers to comparing the model's output results with observational data that were not involved in or only partially involved in calibration to determine whether the model's simulation results are reliable. The outer, middle, and inner models correspond to different spatial scales, and their dominant processes and observational data also differ; therefore, parameter calibration and result validation should be performed using observational data from the corresponding regions, respectively.

[0042] It is worth noting that parameter calibration and result validation should not be limited to a single model level, but should cover the outer, middle, and inner layers respectively. The outer layer model can focus on validating large-scale changes in water level, flow velocity, temperature, and salinity; the middle layer model can focus on validating changes in flow velocity, temperature, salinity, and ecological variables in the coastal areas of Guangxi; and the inner layer model can focus on validating local changes in water level, flow velocity, nutrients, and chlorophyll in harbor areas. Regional calibration and validation can improve the adaptability of the three-layer models at their respective scales.

[0043] In one specific implementation, flow velocity, temperature, salinity, nutrient levels, and chlorophyll data can be used to calibrate and validate the model. For hydrodynamic parameters, bottom friction parameters, vertical mixing parameters, or horizontal diffusion parameters can be adjusted; for ecological parameters, sensitive parameters related to nutrient uptake, phytoplankton growth, mortality, sedimentation, or remineralization can be adjusted. The model results and observed data can be evaluated using indicators such as model skill score, correlation coefficient, and root mean square error. If the three-layer nested boundary conditions result in a higher model skill score, higher correlation coefficient, or lower root mean square error compared to results using conventional external data as boundary conditions, it indicates that the three-layer nested boundary transfer improves the model simulation effect.

[0044] When performing step S5, it is essential to first check whether the time, location, unit, and variable names of the observation data correspond to the model output results. Observation data that is clearly abnormal can be discarded. For point-based observation data, it can be mapped to the nearest unstructured grid node, and then the model results corresponding to that grid node at that time can be extracted for comparison. For vertical profile observation data, the observation depth and model vertical stratification should be matched simultaneously to avoid evaluation errors caused by inconsistent depths.

[0045] S6: Output the result of a three-level nested structure.

[0046] Specifically, the three-layer nested results refer to the hydrodynamic field, temperature and salinity field, and ecological variable field output by the outer, middle, and inner models, respectively. The outer layer results reflect the large-scale background processes in the Beibu Gulf, the middle layer results reflect the more detailed processes in the coastal waters of Guangxi, and the inner layer results reflect the local high-resolution processes in Qinzhou Bay or typical harbors. The output results can be hourly results, daily results, monthly average results, typical time-time results, or results at preset time intervals.

[0047] It is worth noting that the three-layer nested results include not only the final inner-layer model results, but also the outer and middle-layer model results. The outer and middle-layer results can serve as boundary inputs to the sub-layer models and can also be used to analyze hydrodynamic and ecological processes at different scales. The variables, time intervals, and spatial ranges of the output results can be set according to actual application needs, but should at least reflect key variables such as water level, flow velocity, temperature, salinity, nutrients, and chlorophyll.

[0048] In one specific implementation, the model output can be saved as a gridded data file, a text table, a graphical result, or a visualization. The outer layer model outputs the water level field, current velocity field, temperature field, salinity field, and ecological variable field within the Beibu Gulf region; the middle layer model outputs the corresponding results within the coastal area of ​​Guangxi; and the inner layer model outputs high-resolution results within the Qinzhou Bay or typical harbor areas. These results can be used to draw spatial distribution maps, time series plots, vertical profiles, and regional average variation curves.

[0049] When performing step S6, ensure that the variable names, units, and time formats of the three-layer model outputs are consistent or mutually convertible. For outputs that need to be further used for boundary transfer, retain sufficiently high temporal resolution and complete vertical layer information. For outputs used for presentation or analysis, time averaging, spatial averaging, or regional statistics can be performed according to the research purpose, but the physical meaning of the original simulation results should not be altered.

[0050] In steps S1 to S3, the outer unstructured mesh model, the middle unstructured mesh model, and the inner unstructured mesh model all adopt a unidirectional nesting method from the outside to the inside. The output results of the outer model are used to drive the middle model, and the output results of the middle model are used to drive the inner model.

[0051] Specifically, unidirectional nesting refers to the transmission of model information only along the direction from the outer layer to the middle layer and from the middle layer to the inner layer. The outer layer model does not receive feedback from the middle layer model, and the middle layer model does not receive feedback from the inner layer model. The outer layer model runs first and forms a large-scale background field. The middle layer model then reads the results of the outer layer model as boundary conditions, and the inner layer model then reads the results of the middle layer model as boundary conditions. This approach reduces model coupling complexity and ensures the stable driving force of the parent layer model's results on the child layer model.

[0052] It is important to note that unidirectional nesting does not mean that the three-layer models are independent of each other, but rather that the model results are passed down step by step in a fixed direction. The outer layer model is used to provide large-scale ocean impacts, the middle layer model is used to inherit the outer background and refine nearshore processes, and the inner layer model is used to further refine harbor processes based on the middle layer results. Since the spatial extent and grid resolution of each layer model are different, unidirectional nesting can avoid local perturbations in the fine-scale model from having a reverse impact on the large-scale model, thus maintaining the clarity and stability of the computational process.

[0053] In one specific implementation, the outer layer model covering the Beibu Gulf is first run to obtain results for water level, current velocity, temperature, salinity, and ecological variables within the Beibu Gulf region. Then, results from the outer layer model corresponding to the open boundary position of the mid-layer model in the Guangxi coastal area are selected and interpolated to generate the mid-layer model boundary file. After the mid-layer model is completed, results from the inner layer model corresponding to the open boundary position of the Qinzhou Bay or a typical harbor are selected and interpolated to generate the inner layer model boundary file. Through this method, a hierarchical nesting is achieved from the Beibu Gulf to the Guangxi coastal area, and then to Qinzhou Bay or a typical harbor.

[0054] In steps S1 to S3, the first resolution is lower than the second resolution, and the second resolution is lower than the third resolution, so that the mesh scale of the outer layer model, the middle layer model and the inner layer model decreases step by step from the outer layer to the inner layer.

[0055] Specifically, resolution represents a model's ability to represent spatial details. Higher resolution generally indicates a smaller mesh scale, allowing the model to depict more subtle variations in shorelines, topography, and current fields. The first resolution is lower than the second, and the second resolution is lower than the third, meaning the outer layer model has the largest mesh scale, the middle layer model has a smaller mesh scale, and the inner layer model has the smallest mesh scale. This configuration ensures that the outer model covers a large area of ​​sea while giving the inner model high local simulation accuracy.

[0056] It is worth noting that this invention does not limit the first, second, and third resolutions to fixed values. In practice, the grid scale of the three-layer model can be adjusted according to the size of the computational domain, the complexity of the shoreline, water depth variations, computational resources, and simulation objectives. As long as the grid scale decreases progressively from the outermost to the innermost layer and the resolution increases progressively, it falls under the three-layer nested resolution setting method of this invention.

[0057] In one specific implementation, the outer layer model can use a kilometer-scale grid to cover a large area of ​​the Beibu Gulf, the middle layer model can use a grid scale of several hundred meters to cover the coastal areas of Guangxi, and the inner layer model can use a smaller-scale grid to cover Qinzhou Bay or typical harbor areas. For areas such as estuaries, shoreline bends, narrow waterways, or areas with large water depth variations, further local densification can be achieved in the corresponding layer model to improve the model's ability to represent local processes.

[0058] In steps S2 and S3, the output results of the outer layer model and the middle layer model both include tidal level, current velocity, temperature, salinity and ecological variables, and the tidal level, current velocity, temperature, salinity and ecological variables are used as the open boundary variables of the sub-layer model.

[0059] Specifically, tidal level represents the change in the height of the free surface of seawater, mainly reflecting tides and large-scale sea surface changes; current velocity represents the speed at which seawater moves in the horizontal and vertical directions; temperature and salinity are used to describe the thermohaline structure of seawater; and ecological variables are used to describe ecological processes such as nutrients and chlorophyll. Using these variables as open boundary variables of the sublayer model allows the sublayer model to simultaneously receive hydrodynamic, temperature-salinity, and ecological information at its open boundary, rather than receiving only a single water level or current velocity condition.

[0060] It is worth noting that if the sub-layer model only inputs the hydrodynamic boundary without inputting the temperature, salinity, and ecological variable boundaries, it may lead to a discontinuity between the temperature-salinity structure or the distribution of ecological variables in the sub-layer model and the parent model. By simultaneously transmitting tidal level, current velocity, temperature, salinity, and ecological variables, the consistency of the variable fields between the parent and sub-layer models can be improved, and the impact of boundary abrupt changes on the simulation results can be reduced.

[0061] In one specific implementation, the outer layer model transmits tidal level, three-dimensional current velocity, temperature, salinity, nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll to the middle layer model; the middle layer model transmits the same type of variables to the inner layer model. For two-dimensional variables, such as tidal level, interpolation can be performed at horizontal boundary nodes and in the time direction; for three-dimensional variables, such as current velocity, temperature, salinity, and ecological variables, simultaneous interpolation transformations at horizontal location, time, and vertical stratification are required.

[0062] In steps S2 and S3, interpolation includes planar interpolation, time interpolation, and vertical interpolation;

[0063] Among them, the plane interpolation uses nearest neighbor interpolation, the time interpolation uses linear interpolation, and the vertical interpolation uses linear interpolation.

[0064] Specifically, planar interpolation refers to transforming the results of the parent model to the open boundary nodes of the child model in a horizontal spatial position. Nearest neighbor interpolation refers to finding the nearest grid node or result point in the parent model to a given open boundary node of a child layer and using the variable value of that nearest point as the variable value of the open boundary node of that child layer. Temporal interpolation refers to calculating the variable value at the target time based on the variable values ​​of two adjacent parent model output times when the output time of the parent model is inconsistent with the boundary update time required by the child model. Vertical interpolation refers to transforming the variables on the vertical layer of the parent model to the vertical layer of the child model when the vertical layers or depth positions of the parent and child models are different.

[0065] It is worth noting that linear interpolation assumes that variables change linearly between two known points and calculates the variable values ​​based on the proportion of the target location or time between the two known points. Temporal linear interpolation is used to preserve the continuous change of boundary variables over time, while vertical linear interpolation is used to preserve the continuous change of temperature, salinity, current velocity, and ecological variables in the direction of water depth. Nearest neighbor interpolation is simple and stable to calculate and is suitable for planar location matching between unstructured grids.

[0066] In one specific implementation, for a given open boundary node in the middle-layer model, the nearest outer-layer mesh node is first located in the outer-layer model, and the tidal level, current velocity, temperature, salinity, and ecological variables of that outer-layer mesh node at two adjacent output times are obtained. Then, temporal linear interpolation is performed based on the required times for the middle-layer model boundary. For three-dimensional variables, vertical linear interpolation is then performed between adjacent vertical layers in the outer-layer model based on the depth of the middle-layer model's vertical layers. The same processing method is used when the middle-layer model transfers boundary variables to the inner-layer model.

[0067] In step S4, the online coupling of hydrodynamics and ecology includes: calculating water level, three-dimensional flow velocity, temperature, salinity, water depth and vertical diffusion coefficient by the hydrodynamic model, and transferring water level, three-dimensional flow velocity, temperature, salinity, water depth and vertical diffusion coefficient as hydrodynamic environmental variables to the ecological dynamics model.

[0068] Specifically, water level is used to determine the location of the free surface of the water body, three-dimensional flow velocity is used to describe the movement of the water body in the horizontal and vertical directions, temperature and salinity are used to describe the thermohaline environment of the water body, water depth is used to determine the vertical spatial extent of the water body, and the vertical diffusion coefficient is used to describe the mixing strength between the upper and lower layers of the water body. These hydrodynamic environmental variables affect nutrient transport, phytoplankton growth, chlorophyll distribution, and the diffusion of ecological variables, and therefore need to be transferred to the ecological dynamics model for ecological calculations.

[0069] It is noteworthy that the hydrodynamic model transmits not only static results at a specific moment to the ecological dynamics model, but also environmental variables that are continuously updated during the model's calculation process. In this way, when calculating the responses of ecological variables, the ecological dynamics model can consider changes in the flow field, temperature-salinity structure, and mixing conditions in real time, thereby avoiding a disconnect between ecological calculations and the hydrodynamic environment.

[0070] In one specific implementation, within each model layer, the hydrodynamic model first calculates the current water level, three-dimensional velocity, temperature, salinity, water depth, and vertical diffusion coefficient based on open boundary conditions, meteorological forcing, river runoff, and the initial field. Then, these variables are transferred to the ecological dynamics model in units of grid nodes or grid cells. Upon receiving these variables, the ecological dynamics model, combining ecological parameters and initial values ​​of the ecological variables, continues to calculate the ecological variable response terms.

[0071] In step S4, the ecological dynamics model calculates the ecological variable response terms based on the hydrodynamic environmental variables and returns the ecological variable response terms to the hydrodynamic model, which then completes the horizontal transport and vertical mixing and diffusion of the ecological variables based on the ecological variable response terms.

[0072] Specifically, ecological variable response terms refer to the source-sink changes in ecological variables caused by biogeochemical processes. Taking nutrients and chlorophyll as examples, nitrates, phosphates, silicates, and ammonium nitrogen may decrease due to phytoplankton uptake or increase due to organic matter decomposition or remineralization; chlorophyll can be used to characterize phytoplankton biomass, and its changes are related to processes such as growth, death, feeding, and sedimentation. After calculating these response terms, the ecological dynamics model returns them to the hydrodynamic model for transport and diffusion calculations.

[0073] It is worth noting that horizontal transport refers to the movement of ecological variables in the ocean along horizontal current fields, while vertical mixing and diffusion refers to the mixing and diffusion of ecological variables between the upper and lower water layers. If only the ecological response term is calculated without considering hydrodynamic transport and diffusion, the spatial distribution of ecological variables in the ocean cannot be accurately described; conversely, if only transport and diffusion are calculated without considering the ecological response term, the biogeochemical changes of the ecological variables themselves cannot be reflected. Therefore, this invention uses online coupling of these two aspects to allow both the response and transport / diffusion processes of ecological variables to participate in the simulation.

[0074] In one specific implementation, the ecological dynamics model calculates the reaction terms for nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll in each coupling step; after receiving the above reaction terms, the hydrodynamic model performs horizontal advection transport, horizontal diffusion, vertical mixing diffusion, and reaction term updates on the above ecological variables in the same grid and at the same time progression, and finally obtains the updated ecological variable field at the current time.

[0075] In steps S1 to S4, the ecological variables include one or more of nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll.

[0076] Specifically, nitrate and ammonium nitrogen are important forms of inorganic nitrogen in water bodies, phosphate is an important phosphorus source required for phytoplankton growth, silicates are closely related to the growth of phytoplankton groups such as diatoms, and chlorophyll can be used to reflect phytoplankton biomass and the primary productivity of water bodies. Incorporating these variables into the ecological variable field can provide a more complete description of the nutrient structure and phytoplankton response in nearshore and harbor waters.

[0077] It is worth noting that the ecological variables actually involved in the simulation may vary depending on the sea area and the purpose of the simulation. When the observation data is relatively complete, nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll can be simulated simultaneously. When reliable observation or boundary data for some variables is lacking, one or more of these variables can be selected for simulation, but the selected ecological variables should be consistent with the input-output relationship of the ecological dynamics model.

[0078] In one specific implementation, the outer, middle, and inner models all use nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll as ecological variables. The outer model forms a large-scale ecological variable background field based on the offshore boundary and river input; the middle model simulates the distribution of ecological variables in the Guangxi coastal area based on the outer boundary; and the inner model simulates the fine distribution of nutrients and chlorophyll in Qinzhou Bay or typical harbors based on the middle boundary.

[0079] In step S5, parameter calibration and result verification are performed using observation data from the corresponding region. The observation data includes one or more of the following: flow velocity, temperature, salinity, nutrient salts, and chlorophyll data.

[0080] When verifying the results, the point observation data is mapped to the nearest unstructured grid node.

[0081] Specifically, the observation data for the corresponding region refers to the field survey data, buoy data, site monitoring data, or remote sensing inversion data that match the calculation regions of the outer, middle, and inner models, respectively. Since unstructured grid nodes usually do not completely overlap with actual observation sites, when comparing model results and site observation data, the unstructured grid node closest to the observation site can be found, and the model results at that grid node can be extracted and compared with the observation data.

[0082] It is worth noting that calibration and validation can use the same type of data, but the evaluation process should reflect the model's ability to simulate real-world ocean processes as closely as possible. For current velocity data, velocity magnitude and direction can be compared; for temperature and salinity data, surface values, bottom values, or vertical profiles can be compared; for nutrient and chlorophyll data, station concentrations, regional averages, or seasonal trends can be compared. Evaluation metrics may include correlation coefficients, root mean square error, and model skill scores.

[0083] In one specific implementation, the outer layer model can be calibrated and validated using flow velocity, temperature, salinity, and chlorophyll data within the Beibu Gulf region; the middle layer model can be calibrated and validated using nearshore observation data from Guangxi; and the inner layer model can be calibrated and validated using observation data from stations in Qinzhou Bay or typical harbors. For each observation station, the nearest unstructured grid node to that station is first found in the corresponding model grid. Then, the simulated values ​​of the model at the same or similar times are extracted, and the error and correlation between them and the observed values ​​are calculated.

[0084] In step S6, the three-layer nested results include the water level field, flow velocity field, temperature field, salinity field, nutrient salt field, and chlorophyll field corresponding to the outer, middle, and inner layers, respectively.

[0085] Specifically, the water level field represents the distribution of water surface height at different spatial locations and times; the velocity field represents the spatial distribution of water movement speed; the temperature and salinity fields represent the thermo-salinity structure of the water body; the nutrient field represents the concentration distribution of nutrients such as nitrate, phosphate, silicate, and ammonium nitrogen; and the chlorophyll field represents the chlorophyll concentration distribution. These results can be output from the outer, middle, and inner layer models respectively to form a simulation result system at different spatial scales.

[0086] It is worth noting that the three nested results serve different purposes. The outer layer results are mainly used to represent the large-scale hydrodynamic and ecological background of the Beibu Gulf; the middle layer results are mainly used to represent transitional-scale processes in the coastal areas of Guangxi; and the inner layer results are mainly used to represent local fine-scale processes in Qinzhou Bay or typical harbors. The three layers are interconnected and can be used to analyze how the influences of the open sea are progressively transmitted to the nearshore and harbor areas.

[0087] In one specific implementation, the outer layer model outputs hourly or daily water level, current velocity, temperature, salinity, nutrient, and chlorophyll fields within the Beibu Gulf region; the middle layer model outputs similar results within the Guangxi coastal area; and the inner layer model outputs similar results at higher resolution within the Qinzhou Bay or typical harbor areas. These results can be further used to draw contour maps, flow field maps, profile maps, time series curves, and regional statistical maps, and can also serve as a data foundation for ecological environment assessment, nearshore marine management, and harbor ecological process analysis.

[0088] Example 1: The outer layer model, middle layer model, and inner layer model are calculated using different internal and external modal parameters and vertical layering parameters.

[0089] The internal and external modes are a computational separation method in hydrodynamic models. External modes are primarily used to describe rapidly propagating hydrodynamic processes such as water level changes and barotropic flow; internal modes are primarily used to describe changes in three-dimensional flow velocity, temperature, salinity, and density structure. By setting the ratio of internal to external modes and the time step of the internal mode, the model's operational efficiency can be improved while ensuring computational stability.

[0090] Specifically, the outer layer model is the Beibu Gulf model, with an inner-outer mode ratio of 1:5, an inner mode time step of 2 seconds, and 40 vertical layers; the middle layer model is the Guangxi coastal model, with an inner-outer mode ratio of 1:2, an inner mode time step of 1 second, and 20 vertical layers; and the inner layer model is the Qinzhou Bay or a typical harbor model, with an inner-outer mode ratio of 1:5, an inner mode time step of 1 second, and 10 vertical layers.

[0091] It is worth noting that the aforementioned ratio of inner and outer modes, time step of the inner mode, and number of vertical layers are specific operating parameters in this embodiment and are not intended to limit the scope of protection of this invention. In other embodiments, the ratio of inner and outer modes, time step of the inner mode, and number of vertical layers can be adjusted according to the water depth of the computational domain, grid scale, tidal propagation speed, computational stability requirements, and computational resource conditions. As long as the outer layer model, middle layer model, and inner layer model can achieve unidirectional nesting from the outside to the inside, progressively increase resolution, and perform online coupling of hydrodynamics and ecology, they all fall under the embodiments of this invention.

[0092] Example 2: To verify the simulation capability of the inner-layer model established in this invention for the tidal level change process, hourly tidal level data from two tidal level stations, Qinzhou Station and Sanniang Bay Station, in July 2024 were selected as verification data. The tidal level results output by the inner-layer model were compared with the measured tidal level data from the above stations to evaluate the model's simulation effect on the tidal level change process in Qinzhou Bay.

[0093] Specifically, the tidal type in Qinzhou Bay is diurnal. Comparative analysis shows that the model-output tide level and the measured tide level exhibit good consistency in terms of high tide, low tide, and tidal peak-valley variations. The overall model skill score for tide level is 0.9940, the coefficient of determination is 0.9888, and the root mean square error is 0.1827 meters. The closer the model skill score and coefficient of determination are to 1, the higher the consistency between the model results and the measured results; the smaller the root mean square error, the smaller the deviation between the model's calculated value and the measured value. These results indicate that the three-layer nested model established in this embodiment can accurately simulate the tidal level variation process in Qinzhou Bay.

[0094] It is worth noting that tidal level validation is mainly used to demonstrate the boundary condition transfer capability and local water level simulation capability of the hydrodynamic model. Since water level and flow velocity are important foundations for the calculation of ecological variable transport and diffusion, the reliability of tidal level simulation results can further support the credibility of subsequent simulations of variables such as temperature, salinity, nutrients, and chlorophyll.

[0095] Example 3: To verify the improvement effect of three-layer nested boundary propagation compared with directly using conventional external data as boundary conditions, an example and a comparative example are set up for comparison.

[0096] The embodiment employs the three-layer nesting method of this invention: first, the outer layer model covering the Beibu Gulf is run to obtain the outer hydrodynamic field, temperature and salinity field, and ecological variable field; then, the outer layer output is interpolated as the open boundary condition of the middle layer model of the Guangxi nearshore area; finally, the middle layer output is interpolated as the open boundary condition of the inner layer model of Qinzhou Bay or a typical harbor. The comparative example does not use the hierarchical nesting boundary transfer from outer to middle and from middle to inner layers, but directly uses conventional external oceanographic data as the boundary input of the target model.

[0097] In the evaluation of the power flow simulation results, the flow velocity model skill score of the embodiment was 0.7648, and the flow direction model skill score was 0.8760; the flow velocity correlation coefficient was 0.7650, and the flow direction correlation coefficient was 0.8875; the root mean square error of flow velocity was 0.0930 m / s, and the root mean square error of flow direction was 33.9327 degrees. The flow velocity model skill score of the comparative embodiment was 0.6934, and the flow direction model skill score was 0.7659; the flow velocity correlation coefficient was 0.6184, and the flow direction correlation coefficient was 0.7937; the root mean square error of flow velocity was 0.3485 m / s, and the root mean square error of flow direction was 46.9731 degrees.

[0098] The comparison results above show that, compared to directly using conventional external data as boundary input, the present invention, by generating open boundary conditions for sub-layer models through a three-layer nesting method, improves the model skill scores and correlation coefficients for flow velocity and direction, while reducing the root mean square errors for both flow velocity and direction. These results demonstrate that the three-layer nested boundary transfer method of the present invention can improve the accuracy and reliability of tidal current simulation results in nearshore and harbor areas.

[0099] Please see Figure 2 In one specific embodiment, the three-layer nested model includes an outer Beibu Gulf hydrodynamic and ecological model, a middle Guangxi nearshore water dynamic and ecological model, and an inner Qinzhou Bay or typical harbor hydrodynamic and ecological model. The outer, middle, and inner models all include hydrodynamic calculation components and ecological dynamics calculation components.

[0100] The hydrodynamic calculation section is used to calculate water level, three-dimensional flow velocity, temperature, salinity, water depth, and vertical diffusion coefficient; the ecodynamic calculation section is used to calculate the response terms of ecological variables such as nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll. The hydrodynamic calculation section transfers hydrodynamic environmental variables to the ecodynamic calculation section, which in turn returns the ecological variable response terms to the hydrodynamic calculation section, which then further completes the horizontal transport and vertical mixing and diffusion of ecological variables.

[0101] After the outer Beibu Gulf hydrodynamic and ecological model is run, its output tidal level, current velocity, temperature, salinity, and ecological variables are processed through planar interpolation, temporal interpolation, and vertical interpolation, and used as the open boundary conditions for the middle-layer Guangxi nearshore hydrodynamic and ecological model. Similarly, after the middle-layer Guangxi nearshore hydrodynamic and ecological model is run, its output tidal level, current velocity, temperature, salinity, and ecological variables are processed through planar interpolation, temporal interpolation, and vertical interpolation, and used as the open boundary conditions for the inner-layer Qinzhou Bay or typical harbor hydrodynamic and ecological model. This forms a hierarchical nested modeling process from the Beibu Gulf to the Guangxi nearshore waters, and then to Qinzhou Bay or typical harbors.

[0102] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A three-layer nested hydrodynamic and ecological coupling modeling method, characterized in that, Includes the following steps: S1: Construct an outer unstructured grid model covering the Beibu Gulf, inputting data on the outer sea boundary, meteorological forcing, river runoff, and ecological observations, and calculate the outer hydrodynamic field, temperature and salinity field, and ecological variable field at the first resolution; S2: Construct a mid-level unstructured grid model covering the nearshore waters of Guangxi, interpolate the output of the outer-level model as the open boundary conditions of the mid-level model, and calculate the mid-level hydrodynamic field, temperature and salinity field and ecological variable field at the second resolution. S3: Construct an inner-layer unstructured grid model covering Qinzhou Bay or a typical harbor, interpolate the output of the middle-layer model as the open boundary conditions of the inner-layer model, and calculate the inner-layer hydrodynamic field, temperature and salinity field and ecological variable field at the third resolution. S4: Perform online coupling of hydrodynamics and ecology in the outer, middle and inner layer models respectively; S5: Perform parameter calibration and result verification for the outer layer model, middle layer model, and inner layer model respectively; S6: Output the result of a three-level nested structure.

2. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In steps S1 to S3, the outer unstructured mesh model, the middle unstructured mesh model, and the inner unstructured mesh model all adopt a unidirectional nesting method from the outside to the inside. The output of the outer model is used to drive the middle model, and the output of the middle model is used to drive the inner model.

3. A three-layer nested hydrodynamic and ecological coupling modeling method as described in claim 1, characterized in that, In steps S1 to S3, the first resolution is lower than the second resolution, and the second resolution is lower than the third resolution, so that the mesh scale of the outer layer model, the middle layer model and the inner layer model decreases step by step from the outer layer to the inner layer.

4. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In steps S2 and S3, the output results of the outer layer model and the middle layer model both include tidal level, current velocity, temperature, salinity and ecological variables, and the tidal level, current velocity, temperature, salinity and ecological variables are used as the open boundary variables of the sub-layer model.

5. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In steps S2 and S3, the interpolation includes planar interpolation, time interpolation, and vertical interpolation; Among them, the plane interpolation uses nearest neighbor interpolation, the time interpolation uses linear interpolation, and the vertical interpolation uses linear interpolation.

6. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In step S4, the online coupling of hydrodynamics and ecology includes: calculating water level, three-dimensional flow velocity, temperature, salinity, water depth and vertical diffusion coefficient by the hydrodynamic model, and transferring water level, three-dimensional flow velocity, temperature, salinity, water depth and vertical diffusion coefficient as hydrodynamic environmental variables to the ecological dynamics model.

7. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In step S4, the ecological dynamics model calculates the ecological variable response terms based on the hydrodynamic environmental variables and returns the ecological variable response terms to the hydrodynamic model, which then completes the horizontal transport and vertical mixing and diffusion of the ecological variables based on the ecological variable response terms.

8. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In steps S1 to S4, the ecological variables include one or more of nitrate, phosphate, silicate, ammonium nitrogen, and chlorophyll.

9. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In step S5, the parameter calibration and result verification are performed using observation data from the corresponding region. The observation data includes one or more of the following: flow velocity, temperature, salinity, nutrient salts, and chlorophyll data. When verifying the results, the point observation data is mapped to the nearest unstructured grid node.

10. The three-layer nested hydrodynamic and ecological coupling modeling method according to claim 1, characterized in that, In step S6, the three-layer nested result includes the water level field, flow velocity field, temperature field, salinity field, nutrient salt field, and chlorophyll field corresponding to the outer layer, middle layer, and inner layer, respectively.

Images