CN110399676B - Northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system - Google Patents

Abstract

The invention discloses a northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system, which comprises an oil spill information rapid processing module, an environmental information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an ensemble forecasting module, a geographic information data module, a visual analysis module and a system control center module; the method can realize three-dimensional short-term and medium-term numerical simulation prediction of deep sea oil spill in the northwest Pacific ocean and application comparison of different vertical diffusion schemes under the action of sea waves; the method can be used for preprocessing various oil source information and oil spilling types and developing prediction; the external forced field of the oil spill model can be optimized by utilizing field observation data and an optimal interpolation assimilation method in real time, so that the forecasting accuracy is improved; the ensemble forecasting result can be provided; has important practical value for the scientific research of marine ecological disaster prevention and reduction, the design of emergency response systems of management departments, and the like.

Description

Northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system

Technical Field

The invention relates to a numerical simulation technology of offshore oil spill drifting diffusion, in particular to a short-term and medium-term prediction technology of three-dimensional numerical simulation in a deep sea area, and realizes the field of business application in the northwest Pacific sea area.

Background

In the existing research of oil spill transportation and diffusion simulation, the transportation and diffusion of sea surface oil spill are mainly affected by sea surface wind field, ocean current, wave and turbulent flow, and the transportation and diffusion of underwater oil spill are mainly affected by ocean current flow velocity and turbulent flow. Reed et al (1994a) suggested that 3.5% of the wind speed gives the oil film drift velocity in the absence of a breakup wave and in the case of a breeze. However, when the wind speed increases, the spilled oil will be entrained into the water, the shearing of the sea current and the breaking action of the waves are not negligible, and studies have already demonstrated the importance of the vertical movement of the spilled oil. The results of various field, laboratory or numerical simulation results of scholars such as Johansen (1984), Elliot et al (1986), Delvignone and Sweeney (1988), Reed et al (1994a) and the like show the importance of vertical movement of oil spill, and the natural entrapment process of oil spill plays an important role in oil spill homing simulation and also determines the space-time distribution of oil spill on the sea surface. Therefore, the vertical diffusion motion of oil spill under the influence of waves becomes the leading edge research of the current three-dimensional oil spill numerical simulation, and four vertical diffusion simulation methods appear in the existing Lagrangian random walk oil spill simulation research, such as Giovanni coppinci et al (2011), Wang S.D. et al (2008), Craig (1994), Mellor (2004), Lonin (1999) and the like, wherein different vertical diffusion schemes are adopted in the simulation research. However, sensitivity analysis of vertical diffusion on distribution of oil spill transportation in space and time is not carried out, and relevant documents do not exist for carrying out comparative analysis on simulation results of different vertical diffusion simulation schemes, and further do not disclose that the simulation results of which scheme are adopted are more consistent with observation results in practical application.

With the improvement of field observation technology and monitoring level, the development of satellite technology and the strength for handling the accidents are enhanced. Wind field observation data at the offshore oil well platform are gradually accumulated. Meanwhile, after an accident happens, the emergency department can monitor the oil spill drifting condition and the nearby ocean current condition by using satellite, aerial remote sensing and ship field observation. It is expected that the observation data of offshore wind, current and field accidents are more and more, how to improve the accuracy of the oil spill transportation diffusion prediction by using the observation data becomes a new direction for researching an oil spill numerical prediction model, and the new problem on the schedule is also a problem concerned by a plurality of experts and oil spill emergency decision-making system application units in the process of checking and accepting the task of the national 'one-fiftieth' scientific and technical attack project, and is also a further direction for researching the oil spill prediction model pointed out by experts.

The accuracy of the oil spill numerical simulation depends on whether the formulation solution of the physical process of the mode is reasonable or not, the accuracy of input data information (such as information of wind, flow and oil spill sources) and the use of a simulation result. The input data bring errors from various sources, which affect the prediction accuracy, and the sources of the errors are very important to the prediction accuracy and can not be ignored. Sensitivity tests of various parameters in the model of oil spill transport diffusion (e.g., Elliott, 200) have been performed by a number of researchers4; jorda et al, 2007; ana j. abasca et al, 2010), which revealed many factors that influence the accuracy of the spill transport diffusion simulation. Besides the errors caused by the data of various input modes, a series of research works are carried out on the influence of factors such as mode calculation time step, mode integration method and mode resolution on the oil spill transportation simulation forecasting result by a plurality of scholars (James M.et al, 2004; Elliott and Jones, 2000; Reinaldo and Henry, 1999). In the face of so many factors affecting simulation prediction accuracy of oil spill transportation, in order to improve prediction accuracy to meet actual requirements, we should avoid or weaken various error sources from those aspects, so as to achieve the purpose of improving simulation accuracy? Different scholars have conducted various research works aiming at different error sources. Such as Mariano a.j.et al. (2011), the ocean current forecast error is weakened through ensemble forecasting, so that the oil spill transport diffusion forecasting accuracy is improved, and the method is applied to forecasting of the oil spill event of the NOAA in the gulf of mexico.

And the guidelines Soares (2006,2007) developed methods to determine the uncertainty in the oil spill forecast for coastal and open sea areas, respectively. However, no study has been made to determine the main error source in the spill oil transportation diffusion prediction, and no scholars have developed experimental analysis for susceptibility of spill oil transportation medium-term prediction in a certain region to find the main error source for medium-term prediction.

In the prior art, the oil spill prediction mostly adopts a two-dimensional model, only sea level is considered, and the consideration on the deep sea oil spill and the vertical motion process of oil particles is not sufficient. The advanced OSCAR model in Norway abroad and the OILMAP model in the United states can realize three-dimensional simulation of oil spill and deep sea oil spill simulation, but on one hand, the technology is not disclosed, and on the other hand, the effect of vertical motion of different water layer ocean currents on oil particles is not considered in the vertical direction in the technology, which is fully considered in some sea areas, particularly places with obvious vertical flow.

And the vertical turbulence diffusion part of the existing oil spill three-dimensional simulation is basically an empirical formula, and the influence of the turbulence characteristics of the ocean current model on the oil spill diffusion and the due characteristics of the oil spill vertical turbulence diffusion are rarely considered.

Disclosure of Invention

The invention provides a northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system.

The scheme of the invention is as follows:

the northwest Pacific ocean three-dimensional oil spill business emergency forecasting and evaluating system comprises an oil spill information rapid processing module, an environmental information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an aggregation forecasting module, a geographic information data module, a visual analysis module and a system control center.

As a preferred technical scheme, the oil spill information rapid processing module rapidly processes different types of accident source information and accident categories according to an instruction sent by the system control center, and provides information of an oil spill source for the system control center;

the environment information processing module provides the sea surface wind, ocean current, ocean temperature and ocean wave field data information processing cost system with different sources and data formats to the system control center for storage according to the system control center instruction, so that the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module can read and use the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module;

the oil spill transportation module simulates the time-space behavior of oil spill in the marine environment by adopting an oil particle model method; the oil particle model method comprises an advection process and a diffusion process, wherein the oil spill three-dimensional transportation diffusion motion comprises the advection process, the diffusion process and the action of sea waves in the vertical direction on the water entering, the buoyancy, the turbulence and the vertical sea current of oil particles; the advection process is solved by adopting an Euler method with second-order precision, the diffusion process is simulated by adopting a random walk method, and a vertical diffusion scheme A, a vertical diffusion scheme B and a vertical diffusion scheme C are respectively adopted in the vertical direction in the diffusion process; the oil spill transportation module acquires information required by calculation from the system control center and provides a calculation result to the system control center; the model boundary condition in the oil particle model method adopts a non-reflection condition, wherein the non-reflection condition is that when the oil particles rise to the sea surface in water, the oil particles become surface particles and translate and diffuse along with surface flow together with other surface particles; the oil particles reach a land boundary in the drifting process, which is equal to the fact that the oil particles are adhered to the land, the particles are declared to be dead in calculation and do not participate in calculation any more, the oil particles are considered to be dead when the oil particles reach the boundary of a calculation area and do not participate in calculation any more, and the condition that the oil particles are in the bank is judged by adopting a grid method;

the oil spill weathering module calculates evaporation, emulsification and density of oil spill through an empirical model to simulate and predict the change conditions of the marine residual oil quantity and the oil spill density, acquires information required by calculation from the system control center and provides the calculation result to the system control center;

the data assimilation module is started under the condition that a field observation result is read from the oil spill information rapid processing module, and environment field data are optimized by using an optimal interpolation assimilation method, so that the forecasting accuracy is improved; the control button of the data assimilation module is in a system control center and is triggered by the system control center according to whether field observation data exist or not;

the ensemble forecasting module selects simulation results under parameters of ocean currents, wind fields, sea waves and oil spillage from different sources to carry out ensemble forecasting according to instructions of the system control center, defaults or manually selects an external forcing scheme and an oil spillage model parameter scheme, defaults or manually inputs weights of all schemes to obtain ensemble forecasting results, and provides the ensemble forecasting results to the system control center;

the geographic information data module processes the Chinese and western Pacific ocean geographic information data in the module database into water depth and sea-land grid data in a format required by current calculation according to an instruction of the system control center, and provides the data to the system control center;

the visual analysis module outputs the forecast result in a picture and animation mode according to the command of the system control center, and the forecast content comprises an oil drift track, a diffusion range, the thickness and the distribution of an oil film, an influence bank section range and shore arrival time, an influence sensitive area and arrival time and a sea sweeping area;

the system control center is a control instruction center and a data exchange center, controls an oil spill information rapid processing module, an environment information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an ensemble prediction module, a geographic information data module and a visual analysis module, and selects an external forced field file and an input oil spill parameter manually or in a default mode; and the system control center completes data exchange among the modules according to the setting, realizes simulation calculation, sends a command to the visualization module, and draws a simulation result product.

As a preferred technical scheme, the oil spill information rapid processing module comprises an oil spill source initialization module and an oil spill field observation information processing module, wherein the oil spill source initialization module comprises processing of picture oil source information and data oil source information, for the picture oil source information, the module collects the coordinate position of a pollution area outline based on matlab software, digitalizes the oil spill source observation picture information by using the functions in the matlab, and for the data oil source information, the oil spill source initialization module converts the oil source information of point source, surface source continuous oil spill or instantaneous oil spill, and moving oil source continuous or instantaneous oil spill into oil spill digitalized information of a format required by a model; the oil spill site observation information processing module processes oil spill observation time, place, range, oil spill amount, oil spill mode, oil product type, oil source moving speed, wind field, flow field, sea wave observed in the oil spill site, latest observation position, shape and time information, provides the information to the system control center for storage, and reads and uses the information for the data assimilation module, the oil spill transportation module and the oil spill weathering module.

As a preferred technical scheme, the picture oil source information comprises a remote sensing picture and an aerial picture.

As a preferred technical scheme, the diffusion process is simulated by adopting a random walk method, an A vertical diffusion scheme is adopted in the vertical direction in the diffusion process, and the following equation is adopted

Xi is standard white Gaussian noise, K_hIs a vertical diffusion coefficient, Δ t is a time step, and C' is a constant;

the K in the A vertical diffusion scheme_hIs K_waveSaid K is_waveFor vertical vortex viscosity, said K_waveThe semi-empirical formula for the Reynolds stress caused by sea surface waves is as follows

Wherein H_sThe effective wave height is represented, kappa is the wave number, T is the average period, and Z is the distance between the position of the oil particles and the sea surface; said H_sThe following formula is adopted for obtaining the effective wave height

As a preferred technical scheme, the diffusion process is simulated by a random walk method, a vertical diffusion scheme B is adopted in the vertical direction in the diffusion process, and the following equation is adopted:

xi is standard white Gaussian noise, K_hIs vertical diffusion coefficient, delta t is time step length, C' is constant, the above equation directly adopts vertical turbulence coefficient in the ocean current model as K in the equation_h。

As a preferred technical scheme, the diffusion process is simulated by a random walk method, a C vertical diffusion scheme is adopted in the vertical direction in the diffusion process, and the following equation is adopted:

the above formula is a Langeven equation, where α' and λ are coefficients, determined by the variance and dispersion of the Stochartic process, and the relationship between the Langeven equation and the Markov chain is as follows

Wherein A, B and C in the formula are solved according to the following formula;

where σ is the root mean square of the turbulent velocity, T_LFor Lagrange integration time scale, the T_LSolving for Lagrange integration time scale according to the following formula;

b is turbulent kinetic energy, epsilon is turbulent energy dissipation rate, and l is turbulent characteristic scale.

And comparing and judging the optimal schemes of the vertical scheme A, the vertical scheme B and the vertical scheme C, and selecting the optimal scheme for providing.

Wave and vertical flow action and A, B, C vertical diffusion scheme are selected as oil spill three-dimensional motion simulation technology.

The movement of oil spill in the ocean water is mainly represented by two processes: bulk displacement under advection and diffusion under shear and turbulence. The surface diffusion process of the oil spill itself is of short duration, while the motion patterns of longer duration are mainly manifested as advection transport and turbulent diffusion. These two motion mechanisms are subject to "advection" and "turbulence", respectively. Both processes always exist simultaneously, and are often referred to as the "advection-diffusion" problem.

Based on the above, the invention adopts an oil particle method to simulate the space-time behavior of oil spill in the marine environment. The particle model method divides the motion process into two main parts, namely a advection process and a diffusion process, solves the advection process of oil spill (particle cloud) by adopting an Euler method with second-order precision, and simulates the diffusion process of the oil spill by adopting a random walk method.

The three-dimensional transportation and diffusion motion of the spilled oil is divided into an advection process and a diffusion process, the action of sea waves in the vertical direction on the entering of oil particles into water, buoyancy, turbulence and vertical sea current is given as follows:

u_oand v_oIs the horizontal velocity, w, of the oil particles_oIs the vertical velocity of the oil particles. The first term to the right of the equation on the equal sign is the velocity of the ocean current, u_cAnd v_cRepresenting the flow velocity in the horizontal direction at the oil particle position, w_cThe velocity of the flow in the vertical direction at the position of the oil particle is shown, and the velocity is derived from the simulation result of the ocean current mode and represents the motion condition of the ocean current at the position of the oil particle. The second term on the right of the equal sign in the equation represents the dragging effect of wind on oil particles, u_aAnd v_aIs the wind speed at 10 m height above the sea surface where the oil particles are located, which is also obtained from the simulation results of the atmospheric numerical model, and this effect is not taken into account when the oil particles are in the water. Alpha and beta are wind drift factors and declination angles, the wind drift factors are usually 1% -6%, the absolute value of the declination angles is usually 0-45 degrees, the right declination is a negative value in the northern hemisphere, and the left declination is a positive value in the southern hemisphere. u. of_wAnd v_wThe oil particle drift velocity due to the wave residual flow generated by the nonlinear wave. The surface tension is increased due to the existence of the oil film, so that the sea surface tends to be smooth, the nonlinear effect of sea waves is greatly weakened, and the corresponding wave residual current is also greatly reduced, therefore, the method is used for solving the problems that the surface tension is increased, the sea surface is smooth, the nonlinear effect of the sea waves is greatly weakened, and the corresponding wave residual current is greatly reducedThe oil particle motion is calculated to be negligible. w is a_okIs the vertical velocity under the action of buoyancy, according to Stokes law:

d radius of oil particles, d_cCritical radius of oil particles. The oil particle radius is lognormal distribution between 10-1000 μm. V is the viscosity of water, ρ_oAnd ρ_wOil and water densities, respectively, the oil density assumed in the modeling process was 830kg m^-3(determined in the actual case according to the type of oil manually input), the density of water was set to 1025kg m^-3The viscosity of water is 1.311mm³s^-1。

The turbulent velocity (< u ' >, < v ' >, < w ' >) was determined by a random walk method:

xi is standard Gauss 'white noise', A_mTo a horizontal diffusion coefficient, K_hFor vertical diffusivity, Δ t is the time step and C' is a constant.

The following three vertical diffusion calculations were performed and can be compared.

Scheme A uses vertical vortex viscosity (K)_wave) Represents K_h。K_waveSemi-empirical formula derived from reynolds stress due to sea surface waves (Ichiye, 1967):

H_sis the effective wave height, κ is the wave number, T is the average period, and Z is the distance from the sea surface where the oil particles are located. Turbulent flow expanderThe dispersion includes vortex and molecular diffusion, generally the vortex diffusion is far larger than the molecular diffusion, so we can neglect the molecular diffusion and directly use the vortex diffusion to represent the diffusion coefficient. The effective wave height adopts the formula suggested by Neumann and Pierson (1996):

scheme B

In the above equation, the vertical turbulence coefficient in the ocean current model is directly adopted as K in the equation_h。

Scheme C

From Lonin (1999). The vertical diffusion process should not be solved as well as the horizontal diffusion process, and the vertical diffusion process should be refined. The main reasons are three points: first, the vertical diffusion process is a rapidly changing transient process; second, the "life times" of vertical turbulent eddies are very short, i.e., the time required for the eddies to collapse from formation is very short; third, vertical shear of the ocean current is important in vertical diffusion simulations. The Langeven equation is used to solve the vertical diffusion process. Oil spill simulation occurs using this method to solve for turbulent motion of the oil spill in the vertical direction.

Langeven's equation is as in equation (4.1), α' and λ are coefficients, determined by the variance and dispersion of the Stochartic process, and the relationship between Langeven's equation and Markov chain (Markov's chain) is as in equation (4.2). Wherein A, B and C are solved according to the formula (4.3). σ is the root mean square of the turbulent velocity, T_LAnd (4) solving according to the formula (4.4) for Lagrange integration time scale. b is turbulent kinetic energy, epsilon is turbulent energy dissipation rate, and l is turbulent characteristic scale. These parameters can all be obtained from ocean current mode.

The vertical diffusion sensitivity experimental analysis is mainly carried out aiming at the three schemes, the influence of vertical mixing on oil spill space-time distribution can be revealed through comparing and analyzing test results, and the optimal vertical diffusion scheme is judged and selected.

And optimizing an external forced field of the oil spill model by using field observation data and an optimal interpolation assimilation method so as to provide prediction accuracy.

In the existing oil spill business emergency forecast, the forecast system is required to be started at any time and to give a forecast result quickly. The method is characterized in that the calculation of a business numerical forecasting system of a wind field and a flow field is time-consuming, so that the calculation time is saved, the numerical forecasting of wind and flow is usually carried out at a fixed point every day, then the forecasting results of three days in the future are stored in a database, and when an oil spill accident occurs, an oil spill transportation diffusion numerical forecasting model is started to read. However, this has led to new problems: when the observation of the wind field and the ocean current of the accident site is received within one day, the data is not suitable to be added into a sea surface wind field and ocean current forecasting system for assimilation and recalculation, so that the calculation burden is greatly increased, time is consumed, and the characteristics of rapidness and rapidness in meeting the emergency forecasting requirement cannot be met; therefore, how to optimize wind field and flow field forecast data by using existing observation data and then provide the data to the oil spill mode under the condition that observation data of a wind field and a flow field on site can be acquired after an accident occurs?

The invention provides a method for optimizing an external forcing field by using field observation data and an optimal interpolation assimilation method, provides the accuracy of an environmental field, reduces the error introduction of an oil spill model, and improves the prediction accuracy of the oil spill model.

An Optimal Interpolation method (OI) is an analysis method based on a statistical theory, and adopts a statistical least square method to determine Optimal weight, so that observation is directly interpolated to a background field to obtain an Optimal Interpolation result, namely, the obtained analysis field is a linear combination of the background field and the observation field under the meaning of minimum variance. The analytical equation of OI is as follows:

x^a＝x^b+BH^T(HBH^T+R)^-1(y^o-H[x^b])

wherein x^bAs background field (i.e. mode prediction result), y^oTo observe a field (i.e., an observation), we generally assume:

K＝BH^T(HBH^T+R)^-1

and K is called a gain matrix or a weight matrix, wherein R is an observation error covariance matrix, diagonal elements of the matrix are observation error covariance values, other elements are 0, and B is a background error covariance matrix:

σ²is a static background error covariance constant value, L is a relevant scale factor, also a constant, Δ x²,Δy²For the distance matrix of each grid point of the background field and other grid points, H is a linear observation operator, x^aThe analysis field is obtained after the background field is corrected by an OI assimilation technology. Considering the observation error and the forecast error, the sigma of B is set as 0.2 and R is empirically²＝4。

Therefore, the problem to be solved here is how much L value can obtain the minimum value of the whole field error for each field observation data number and position distribution feature. In the module, a cross-correction (cross-correct) method is adopted to determine the value of L. Here, the cross correction method is to take the values of N-1 observation stations in N observation stations in turn to make assimilation to obtain an analysis field, and then find the 1 observation station observation in the same time of the analysis fieldThe resulting error value. Thus, under the condition of the fixed value L, N error values are obtained, and then the N values are averaged to be used as the average error value (sigma) of the analysis field under the L value_c ²) Then sigma obtained by this method is applied to each observation time_c ²Averaging to obtain time average error (σ) of the analysis field under the L value_cm ²). And respectively carrying out tests on the U component and the V component of the wind speed from 10 meters to 300 meters by using the N observation point data and the environmental background forcing field data and adopting a cross correction method, and selecting an optimal L selection value from test results to carry out assimilation.

And analyzing the model error, developing ensemble prediction and providing reference for comprehensive study and judgment of a forecaster.

The simulation results under different ocean currents, wind fields and oil spill parameter selections are utilized to carry out error analysis, ensemble prediction is carried out, an external forcing scheme can be manually selected, or error disturbance is carried out on the wind currents, an oil spill model parameter scheme can be selected, weights of all schemes can be manually input to obtain ensemble prediction results, and the ensemble prediction results can be used for a forecaster to carry out comprehensive study and judgment and provide reference information.

The digital processing technology of the information of various oil spilling sources and the oil spilling class data.

The invention can convert the picture information and the data file into the formatted digital information required by the oil spill model, thereby realizing the numerical simulation prediction of the oil spill types monitored by ships, platforms and moving oil wheels, point source surface sources and moving oil sources, continuous or instantaneous oil spill, satellites and aerial remote sensing. FIG. 3 shows the process of the present module digitizing pictures; based on matlab software, collecting the coordinate position of the profile line of the pollution area, uniformly generating oil particles in the profile line area by using some functions in the matlab, and obtaining the coordinate position of each oil particle.

The invention processes the oil spilling time, the oil spilling place, the oil spilling range, the oil spilling mode, the oil product type, the oil source moving speed, the wind field, the flow field, the sea wave and the latest observation position, shape and time information observed on the oil spilling site, so as to facilitate the reading and the use of the data assimilation module, the oil spilling transportation module and the oil spilling weathering module.

And establishing a northwest Pacific three-dimensional oil spill business forecasting system and realizing business operation.

According to engineering standards and software copyright application standards, the modules are connected with an environmental information processing module, an oil spill weathering module, a geographic information data module and a visual analysis module, and a system control center module is established to control the operation sequence and data exchange of the modules, so that a northwest Pacific three-dimensional oil spill business forecasting system is established, and business operation is realized. The system structure is shown in fig. 1, the business operation flow of the system is shown in fig. 2, and the functions and data exchange conditions of the modules are described as follows:

the oil spilling information rapid processing module: the module rapidly processes different types of accident source information and accident categories according to instructions of the system control center, digitizes the oil spilling source information data of different categories according to system format requirements, and provides the digitized oil spilling source information data to the system control center.

The environment information processing module: the module processes the sea surface wind, ocean current, ocean temperature and ocean wave field data information with different sources and data formats into the format required by the system according to the instruction of the system control center and provides the format to the system control center.

And (3) an oil spill transportation module: the module calculates the time-space behavior of the oil spill in the marine environment according to the control center instruction and the data provided by other modules to the control center, and provides the calculation result to the control center.

Oil spill weathering module: the module calculates evaporation, emulsification and density of spilled oil through an empirical model according to a control center instruction and the provided environment field condition to simulate and predict the change conditions of the offshore residual oil quantity and the spilled oil density, and provides the calculation result to the control center.

A data assimilation module: the module is started according to the instruction of the control center, optimizes the environmental field data by using an optimal interpolation assimilation method, and provides the optimized environmental field data for the control center.

An ensemble forecasting module: the module selects simulation results of ocean currents, wind fields, sea waves and oil spilling parameters from different sources to develop ensemble prediction according to instructions of a system control center, can manually select an external forcing scheme and an oil spilling model parameter scheme, can manually input weights of all schemes to obtain ensemble prediction results, and returns the results to the control center for storage.

A geographic information data module: and processing the sea geographic information data of the Chinese and western Pacific ocean in the module database into water depth and sea and land grid data in a format required by current calculation according to instructions of a system control center, and returning the data to the control center for storage.

A visualization analysis module: the module outputs the forecast result in the form of pictures and animations according to the command of the system control center, and the forecast content comprises an oil drift track, a diffusion range, the thickness and the distribution of an oil film, the range of an affected bank section and the shore arrival time, the affected sensitive area and the arrival time and the sea sweeping area.

The system control center: the module is a control command center and a data exchange center and controls other 8 modules. The system control center selects an external forced field file and an input oil spilling parameter manually or in a default mode; and the system control center completes data exchange among the modules according to the setting, realizes simulation calculation, sends an instruction to the visualization module, and draws a simulation result product.

The invention has the advantages that:

1. the method is completely independent of intellectual property rights and does not depend on foreign technologies and the provision of foreign marine environment compelling fields.

2. The three-dimensional simulation of deep sea oil spill and the application selection and comparison of different vertical schemes are realized, and the three-dimensional simulation technology reaches the international leading level.

3. The business application of northwest tai is realized, and the important role is played in the actual case application process.

4. And the method can process and predict information of various oil spilling sources.

5. By utilizing an assimilation technology, the optimization of real-time field observation data on the forced field data of the oil spill model is realized, and the forecasting accuracy is improved.

6. The oil spill simulation results under the environmental compelling fields of different sources can be contrasted and displayed, error analysis is convenient to carry out, the ensemble forecasting technology processing and displaying of all simulation results can be realized after the weight factors are input, and ensemble forecasting result analysis can be provided for forecasters.

The invention has the following beneficial effects:

the invention establishes an optional northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system considering the action of sea waves and vertical flow and three vertical diffusion schemes based on an oil particle method. The system can be applied to the deep sea area of the northwest Pacific ocean; the external forced field of the oil spill model can be optimized by using field observation data and an optimal interpolation assimilation method so as to improve the forecasting accuracy; considering the problem of introduced errors of an external forced field, the forced field can be disturbed or forced fields from different sources can be selected to drive an oil spill model, model errors are analyzed, ensemble prediction is developed, and reference is provided for comprehensive study and judgment of a forecaster; the system can carry out digital processing on various oil spilling source information and oil spilling class data. All functions of the system are applied in a business mode, needed prediction information such as oil spill drifting tracks, extended ranges, oil film thickness, sea sweeping areas and time of oil spill reaching a bank or a sensitive area is timely and quickly provided for field treatment in an actual oil spill case, and prediction results can be displayed in a picture or animation mode. The prediction result determines a targeted, efficient and scientific field cleaning scheme for field disposal, and protects a fragile ecological area and a tourism breeding area in time; the method has important guiding significance for evaluating the influence of marine environment disasters in coastal sea areas on marine ecological environment, protecting the marine ecological environment and seawater quality, protecting coastal breeding industry and implementing sustainable development strategy; the method also has important practical value for scientific research of marine ecological disaster prevention and reduction, design of emergency response systems of management departments and the like.

Drawings

FIG. 1 is a system configuration diagram of embodiment 1;

FIG. 2 is a flow chart of the business operation of embodiment 1;

FIG. 3 is a diagram showing a process of digitizing information in example 1;

FIG. 4 is a schematic diagram of linear interpolation in time in example 1;

FIG. 5 is a diagram illustrating spatial horizontal bilinear interpolation in example 1;

FIG. 6 is a schematic diagram of linear interpolation in the vertical direction in space in example 1;

FIG. 7 is a technical diagram for determining landing of oil particles in example 1;

FIG. 8 is a main interface of the northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system in embodiment 2;

fig. 9 is satellite picture information for selecting an oil spill type of the northwest pacific three-dimensional oil spill business emergency forecasting and evaluating system in embodiment 2;

FIG. 10 is a point source information for selecting a type of oil spill in the northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system in accordance with example 2;

FIG. 11 is a graph showing the shift trajectory of the center position of oil particles in example 2;

FIG. 12 is a graph of the area of influence of example 2;

FIG. 13 is a sea area sweeping view according to example 2;

FIG. 14 is a graph of oil film concentration for example 2;

FIG. 15 is a prediction test chart of example 2.

Detailed Description

In order that the objects and advantages of the invention will be more clearly understood, the invention is further described in detail below with reference to examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The northwest Pacific ocean three-dimensional oil spill business emergency forecasting and evaluating system comprises an oil spill information rapid processing module, an environmental information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an aggregation forecasting module, a geographic information data module, a visual analysis module and a system control center.

As a preferred technical scheme, the oil spill information rapid processing module rapidly processes different types of accident source information and accident categories according to an instruction sent by the system control center, and provides information of an oil spill source for the system control center;

the environment information processing module provides the sea surface wind, ocean current, ocean temperature and ocean wave field data information processing cost system with different sources and data formats to the system control center for storage according to the system control center instruction, so that the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module can read and use the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module;

the oil spill transportation module simulates the time-space behavior of oil spill in the marine environment by adopting an oil particle model method; the oil particle model method comprises an advection process and a diffusion process, the spilled oil three-dimensional transportation diffusion movement comprises the advection process and the diffusion process, the action of sea waves on the vertical direction on the water entering, the buoyancy, the turbulence and the vertical sea current of oil particles is realized, the advection process is solved by adopting an Euler method with second-order precision, the diffusion process is simulated by adopting a random walk method, and the vertical direction in the diffusion process respectively adopts an A vertical diffusion scheme, a B vertical diffusion scheme and a C vertical diffusion scheme; the oil spill transportation module acquires information required by calculation from the system control center and provides a calculation result to the system control center; the model boundary condition in the oil particle model method adopts a non-reflection condition, the non-reflection condition is that when the oil particles in water rise to the sea surface, the oil particles become surface particles and move and diffuse along with surface flow together with other surface particles, the oil particles arrive at a land boundary in the drifting process, which is equal to that the oil particles are adhered to the land, the particles are also declared to be dead in the calculation and do not participate in the calculation any more, when the oil particles arrive at the boundary of a calculation area, the oil particles are also considered to be dead and do not participate in the calculation any more, and a grid method is adopted to judge the condition that the oil particles are in the bank;

the oil spill weathering module calculates evaporation, emulsification and density of oil spill through an empirical model to simulate and predict the change conditions of the marine residual oil quantity and the oil spill density, acquires information required by calculation from the system control center and provides the calculation result to the system control center;

the data assimilation module is started under the condition that a field observation result is read from the oil spill information rapid processing module, and environment field data are optimized by using an optimal interpolation assimilation method, so that the forecasting accuracy is improved; the control button of the data assimilation module is in a system control center and is triggered by the system control center according to whether field observation data exist or not;

the ensemble forecasting module selects simulation results under parameters of ocean currents, wind fields, sea waves and oil spillage from different sources to carry out ensemble forecasting according to instructions of the system control center, defaults or manually selects an external forcing scheme and an oil spillage model parameter scheme, defaults or manually inputs weights of all schemes to obtain ensemble forecasting results, and provides the ensemble forecasting results to the system control center;

the geographic information data module processes the Chinese and western Pacific ocean geographic information data in the module database into water depth and sea-land grid data in a format required by current calculation according to an instruction of the system control center, and provides the data to the system control center;

the visual analysis module outputs the forecast result in a picture and animation mode according to the command of the system control center, and the forecast content comprises an oil drift track, a diffusion range, the thickness and the distribution of an oil film, an influence bank section range and shore arrival time, an influence sensitive area and arrival time and a sea sweeping area;

the system control center is a control instruction center and a data exchange center, controls an oil spill information rapid processing module, an environment information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an ensemble prediction module, a geographic information data module and a visual analysis module, and selects an external forced field file and an input oil spill parameter manually or in a default mode; and the system control center completes data exchange among the modules according to the setting, realizes simulation calculation, sends a command to the visualization module, and draws a simulation result product.

The oil spill information rapid processing module comprises an oil spill source initialization module and an oil spill field observation information processing module, wherein the oil spill source initialization module comprises processing of picture oil source information and data oil source information, for the picture oil source information, the module collects the coordinate position of a pollution area outline based on matlab software, digitalizes oil spill source observation picture information by using functions in the matlab, and for the data oil source information, the oil spill source initialization module converts the oil source information of point source, surface source continuous oil spill or instantaneous oil spill, continuous or instantaneous oil spill of a motion oil source into oil spill digital information in a format required by a model; the oil spill site observation information processing module processes oil spill observation time, place, range, oil spill amount, oil spill mode, oil product type, oil source moving speed, wind field, flow field, sea wave observed in the oil spill site, latest observation position, shape and time information, provides the information to the system control center for storage, and reads and uses the information for the data assimilation module, the oil spill transportation module and the oil spill weathering module.

The picture oil source information comprises a remote sensing picture and an aerial picture.

The diffusion process is simulated by adopting a random walk method, a vertical diffusion scheme A is adopted in the vertical direction in the diffusion process, and the equation is as follows:

xi is standard white Gaussian noise, K_hIs a vertical diffusion coefficient, Δ t is a time step, and C' is a constant;

the K in the A vertical diffusion scheme_hIs K_waveSaid K is_waveFor vertical vortex viscosity, said K_waveThe semi-empirical formula for the Reynolds stress caused by sea surface waves is as follows

Wherein H_sThe effective wave height is represented, kappa is the wave number, T is the average period, and Z is the distance between the position of the oil particles and the sea surface; said H_sThe following formula is adopted for obtaining the effective wave height

The diffusion process is simulated by adopting a random walk method, a vertical diffusion scheme B is adopted in the vertical direction in the diffusion process, and the equation is as follows:

xi is standard white Gaussian noise, K_hIs vertical diffusion coefficient, delta t is time step length, C' is constant, the above equation directly adopts vertical turbulence coefficient in the ocean current model as K in the equation_h。

The diffusion process is simulated by a random walk method, a C vertical diffusion scheme can be adopted in the vertical direction in the diffusion process, and the equation is as follows:

the above formula is a Langeven equation, where α' and λ are coefficients, determined by the variance and dispersion of the Stochartic process, and the relationship between the Langeven equation and the Markov chain is as follows

Wherein A, B and C in the formula are solved according to the following formula;

where σ is the root mean square of the turbulent velocity, T_LFor Lagrange integration time scale, the T_LSolving for Lagrange integration time scale according to the following formula;

b is turbulent kinetic energy, epsilon is turbulent energy dissipation rate, and l is turbulent characteristic scale.

In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.

Example 1

The embodiment of the invention provides a northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system, and realizes business operation of the system, wherein the system structure is shown in figure 1, and the business operation flow is shown in figure 2. The system implementation comprises the following steps:

1. system control center

The module is a control command center and a data exchange center and controls other 8 modules. Before the system is started, important parameters of a system control center are set manually or by selecting a default mode, for example, the processing mode of oil spill source information is selected manually, whether an assimilation module is started or not is selected manually, whether an ensemble prediction module is started or not is selected manually, an external force field file and oil spill parameters are selected manually or by the default mode, an instruction is sent to a visualization module, and a simulation result product needing to be drawn is selected. And starting the system to run after the system parameters are set.

2. Oil spill information rapid processing module

And operating an oil spill information rapid processing module to rapidly process different types of accident source information and provide the information of the oil spill source for the system control center. The oil spill information rapid processing module comprises an oil spill source initialization module and an oil spill site observation information processing module.

2.1 oil spill source initialization processing module

The module is responsible for realizing the digital conversion of oil spill positions and oil film distribution information, so that oil spill of ships, platforms and ports and the digitization of various oil spill source information types of oil spill monitored by satellites and aerial remote sensing are realized, and the digitalized information required by the oil spill model is provided for the function of the module. The types of oil sources which can be treated by the module are mainly as follows: the method comprises the steps of obtaining a satellite remote sensing picture, an aviation remote sensing picture and a facsimile picture, and obtaining a plurality of shapes of surface sources or line sources, instantaneous or continuous oil spilling of a platform and instantaneous or continuous oil spilling of a moving ship.

FIG. 3 shows the process of the present module digitizing pictures; based on matlab software, collecting the coordinate position of the profile line of the pollution area, uniformly generating oil particles in the profile line area by using some functions in the matlab, and obtaining the coordinate position of each oil particle.

2.2 oil spill site observation information processing module

The module processes the oil spilling time, place, range, oil spilling amount, oil spilling mode, oil product type, oil source moving speed, wind field and flow field observed on the oil spilling site, latest observation position, latest observation shape and latest observation time information, and the information is read and used by the data assimilation module, the oil spilling transportation module and the oil spilling weathering module.

3. Data assimilation module

And judging whether the assimilation module needs to be started according to whether field observation environment data exists in the system control center, optimizing the environment field data by the assimilation module, and outputting the optimized environment field data to the control center.

The assimilation module directly interpolates observation into a background field by adopting an Optimal Interpolation method (OI) to obtain an Optimal Interpolation result, namely, the obtained analysis field is a linear combination of the background field and the observation field under the meaning of minimum variance. The analytical equation of OI is as follows:

x^a＝x^b+BH^T(HBH^T+R)^-1(y^o-H[x^b])

wherein x^bAs background field (i.e. mode prediction result), y^oFor observing the field(i.e., observations), we generally assume:

K＝BH^T(HBH^T+R)^-1

and K is called a gain matrix or a weight matrix, wherein R is an observation error covariance matrix, diagonal elements of the matrix are observation error covariance values, other elements are 0, and B is a background error covariance matrix:

σ²is a static background error covariance constant value, L is a relevant scale factor, also a constant, Δ x²,Δy²For the distance matrix of each grid point of the background field and other grid points, H is a linear observation operator, x^aThe analysis field is obtained after the background field is corrected by an OI assimilation technology. Considering the observation error and the forecast error, the sigma of B is set as 0.2 and R is empirically²＝4。

And the data assimilation module determines the value of the L by adopting a cross-correction (cross-correction) method according to the number of field observation data and the position distribution characteristic condition. In the module. The cross correction method is to take the values of N-1 observation stations in N observation stations in turn to obtain an analysis field, and then calculate the error value of the observation result of the remaining 1 observation station in the analysis field. Thus, under the condition of the fixed value L, N error values are obtained, and then the N values are averaged to be used as the average error value (sigma) of the analysis field under the L value_c ²) Then sigma obtained by this method is applied to each observation time_c ²Averaging to obtain time average error (σ) of the analysis field under the L value_cm ²). And respectively carrying out tests on the U component and the V component of the wind speed from 10 meters to 300 meters by using the N observation point data and the environmental background forcing field data and adopting a cross correction method, and selecting an optimal L selection value from test results to carry out assimilation. After the optimal L value is selected, the background error covariance matrix can be determined, and therefore the optimal interpolation field result is obtained.

4. Ensemble forecasting module

The module can utilize simulation results under different ocean currents, wind fields and oil spilling parameter selections to carry out ensemble prediction, can manually select an external forcing scheme and an oil spilling model parameter scheme, and can manually input weights of the schemes to obtain ensemble prediction results.

Assume a set of N samples and a sample weight of W_iIf the ensemble prediction result is:

5. environment information processing module

The module can process sea surface wind, ocean current, ocean temperature and ocean wave field data information with different sources and data formats into a format required by a system for mode reading;

the module acquires the data of the marine environment field required by the current time mode by adopting a linear interpolation mode on the data information in time. In space, the marine environment data of the positions of the particles are obtained by adopting a bilinear interpolation mode on a horizontal plane, and the marine environment data of the positions of the particles are obtained by adopting linear interpolation on different water layers in a vertical direction.

As shown in fig. 4, linear interpolation is to solve the marine environmental field data at the current time T, and the specific formula is:

where T represents time, i represents the mode calculation step, and y represents marine environmental field data.

As shown in fig. 5, the marine environment field data of the position (P) where the oil particles are located is solved by bilinear interpolation, and the specific formula is as follows:

where x represents longitude and y represents latitude.

As shown in fig. 6Showing that the position P of the particle is solved by vertical linear interpolation_(xi,yi,z)The specific formula of the marine environmental field data is as follows:

X_i,Y_iand Z represents the three-dimensional grid position where the oil particle P is located. X_i,Y_iRepresenting the longitude and latitude of the projection of P to the sea surface horizon, respectively.

6. Spilled oil transport module

And simulating the space-time behavior of the oil spill in the marine environment by adopting an oil particle model method. The oil particle model method divides the process of motion into two main parts, namely advection and diffusion.

6.1 oil particle size and Density settings

The concept of "particle diffusion" is to model the concentration field as a "cloud" consisting of a large number of particles, each carrying a certain amount of a tracer substance. Each model particle is translated under specific flow field conditions, and displaced in the vertical direction (surging or sedimentation) under the action of gravity and buoyancy, so-called advection (convection), which is suitably simulated by the lagrangian method. The diffusion process of the model particles is caused by shear flow and turbulence, and a random walk method is suitably adopted to simulate the diffusion process of the particle cloud. Turbulence can be considered as a random flow field, and the motion of each model particle in the turbulent flow field is similar to the brownian motion of the fluid molecules. The dispersion process of the entire cloud in the water body is caused by the random motion of each particle. This method is actually a combination of deterministic and stochastic methods, i.e., the advection process is simulated using a deterministic method (numerical solution) and the diffusion process is simulated using a stochastic method.

The oil particles are defined as small spherical balls with a diameter of 10-100 um. The actual number of particles required to accurately represent an oil spill film should be quite large (e.g., 1m for a diameter equal to 100 um) in view of the range of variation in oil drop diameter³The oil of (A) is equivalent to 1.9X 10¹²Oil for preventing and treating diabetesParticles) it is not possible to simultaneously accumulate such many coordinate points and characteristic parameters in a computer. Therefore, the computer capacity and the length of the run time determine the maximum possible number of particles, and the simulation of the oil particle characteristics is achieved by means of additional volume parameters. Considering a particular particle, with a diameter d, the true volume of this particle is:

it accounts for the total volume of the oil film_iComprises the following steps:

wherein n is the total number of oil particles.

The characteristic volume of each oil particle (additional volume parameter) is thus defined as: v_i＝f_i ^·V₀

Wherein, V₀Is the initial volume of the spill. Thus, each oil particle represents a portion of the volume of oil spill that is proportional to the portion of the total population that it represents.

Since several behaviors of simulated oil droplets, including flotation and mixing, take into account the size and density of the oil droplets, the size spectrum of the oil particles should reflect the reality as much as possible. On-site observations have shown that oil droplet sizes vary between 10 and 1000um (Forrester, 1971). The most likely distribution of oil droplets mixed in a water body is the lognormal distribution (Johansen, 1985).

The standard normal distribution is:

where Φ (x) is the normalized distribution function,

d is the log of the particle diameter₁₀,μ_dLog of mean value of particle diameter₁₀，σ_dLog of standard deviation of oil droplet diameter₁₀. The average size of the oil droplets was taken to be 100 μm (μm)_d2) standard deviation of 3.16(σ)_d0.5), which corresponds to a log normal spectrum with 95% of the oil droplets distributed between 10 and 1000 μm. Delvigne and Sweeney (1994,1989) performed a series of oil spill experiments in the laboratory and found that the average radius of the oil particles in the water was 250 μm with a deviation of 75 μm.

The three-dimensional transportation and diffusion motion of the spilled oil can be divided into an advection process and a diffusion process, the action of sea waves in the vertical direction on the entering of oil particles into water, buoyancy, turbulence and vertical sea current is given as follows:

u_oand v_oIs the horizontal velocity, w, of the oil particles_oIs the vertical velocity of the oil particles. The first term to the right of the equation on the equal sign is the velocity of the ocean current, u_cAnd v_cRepresenting the flow velocity in the horizontal direction at the oil particle position, w_cThe velocity of the flow in the vertical direction at the position of the oil particle is shown, and the velocity is derived from the simulation result of the ocean current mode and represents the motion condition of the ocean current at the position of the oil particle. The second term on the right of the equal sign in the equation represents the dragging effect of wind on oil particles, u_aAnd v_aIs the wind speed at 10 m height above the sea surface where the oil particles are located, which is also obtained from the simulation results of the atmospheric numerical model, and this effect is not taken into account when the oil particles are in the water. Alpha and beta are wind drift factors and declination angles, the wind drift factors are usually 1% -6%, the absolute value of the declination angles is usually 0-45 degrees, the right declination is a negative value in the northern hemisphere, and the left declination is a positive value in the southern hemisphere. u. of_wAnd v_wThe oil particle drift velocity due to the wave residual flow generated by the nonlinear wave. The surface tension is increased due to the existence of oil film, so that the sea is pollutedThe surface tends to be smooth, the nonlinear effect of the sea wave is greatly weakened, and the corresponding wave residual flow is also greatly reduced, so that the motion of oil particles can be ignored when calculating. w is a_okIs the vertical velocity under the action of buoyancy, according to Stokes law:

d radius of oil particles, d_cCritical radius of oil particles. The oil particle radius is lognormal distribution between 10-1000 μm. V is the viscosity of water, ρ_oAnd ρ_wOil and water densities, respectively, the oil density assumed in the modeling process was 830kg m^-3(in the actual case, it is determined by the type of oil manually input), the density of water is assumed to be 1025kg m^-3The viscosity of water is 1.311mm³s^-1。

The turbulent velocity (< u ' >, < v ' >, < w ' >) was determined by a random walk method:

xi is standard Gauss 'white noise', A_mTo a horizontal diffusion coefficient, K_hFor vertical diffusivity, Δ t is the time step and C' is a constant.

The following three vertical diffusion calculations were performed and compared

Scheme A uses vertical vortex viscosity (K)_wave) Represents K_h。K_waveSemi-empirical formula derived from reynolds stress due to sea surface waves (Ichiye, 1967):

H_sis provided withThe effective wave is high, kappa is the wave number, T is the average period, and Z is the distance between the oil particle position and the sea surface. Turbulent diffusion involves vortex and molecular diffusion, and usually vortex diffusion is much larger than molecular diffusion, so we can ignore molecular diffusion and directly use vortex diffusion to represent diffusion coefficient. The effective wave height adopts the formula suggested by Neumann and Pierson (1996):

scheme B

In the above equation, the vertical turbulence coefficient in the ocean current model is directly adopted as K in the equation_h。

Scheme C

From Lonin (1999). The vertical diffusion process should not be solved as well as the horizontal diffusion process, and the vertical diffusion process should be refined. The main reasons are three points: first, the vertical diffusion process is a rapidly changing transient process; second, the "life times" of vertical turbulent eddies are very short, i.e., the time required for the eddies to collapse from formation is very short; third, vertical shear of the ocean current is important in vertical diffusion simulations. The Langeven equation is used to solve the vertical diffusion process. Oil spill simulation occurs using this method to solve for turbulent motion of the oil spill in the vertical direction.

Langeven's equation is as in equation (4.1), α' and λ are coefficients, determined by the variance and dispersion of the Stochartic process, and the relationship between Langeven's equation and Markov chain (Markov's chain) is as in equation (4.2). Wherein A, B and C are solved according to the formula (4.3). σ is the root mean square of the turbulent velocity, T_LAnd (4) solving according to the formula (4.4) for Lagrange integration time scale. b is turbulent kinetic energy, epsilon is turbulent energy dissipation rate, and l is turbulent characteristic scale. These parameters can all be obtained from ocean current mode.

Vertical diffusion sensitivity experimental analysis is mainly carried out according to the three schemes, the influence of vertical mixing on oil spill space-time distribution is revealed through comparison analysis test results, and an optimal vertical diffusion scheme is judged.

6.2 equation solving method

The solution of the advection part in the transport diffusion equation can be solved by adopting an Euler method with second-order precision.

The Euler method is an explicit one-step method with first order accuracy, which is the simplest and most common method of solving differential equations. Solving the trajectory equation by the euler method can be written as: x is the number ofⁿ⁺¹≈xⁿ+v_a(xⁿ,tⁿ) Δ t, where Δ t is the integration step, n is the time index, t_n＝nΔt。

The modified Euler method approximates an integral equation using a trapezoidal equation in a numerical integration method, which has second-order precision. Solving the trajectory equation with the improved euler method can be written as:

7. geographic information data module

The module extracts shoreline and water depth data required by mode calculation from a reserved northwest Pacific ocean geographic information database for oil particle diffusion boundary condition processing (shore-approaching judgment).

In this mode, the motion of the oil particle cloud is three-dimensional, and during the motion, the oil particle cloud may reach a land boundary, float out of the sea surface, or drift out of a forecast area.

In the sea surface boundary processing mode, a reflection condition and a non-reflection condition exist. The reflection condition is that the particles are reflected back to the inside of the water body when sinking to the sea bottom, landing on the shore or floating to the sea surface. The non-reflective condition is that when the downward settling particle reaches the sea floor or shore it will be "permanently" adhered to the sea floor or shore (forming a sediment). When the floating particles reach the sea surface, the floating particles are retained on the sea surface and move horizontally along with the surface flow.

Because the viscosity of oil drops is large, the model adopts a non-reflection condition when oil particles are processed to reach the sea surface and the coast. I.e. when oil drops in the water rise to the surface, they become surface particles and are displaced and spread with the surface flow, together with other surface particles. Oil particles may reach the land boundary during the drift, and at this time, the particles are also considered to be stuck on the land, and the particles are also declared to be "dead" in the calculation and not to participate in the calculation. When a particle reaches the boundary of the calculation region, the particle is also considered to be dead and not to participate in the calculation.

The handling of the seabed and coastal boundaries has certain limitations. Because of this, the bottoming and landing oil cake is likely to be returned to the water area by the turbulent bottom. However, it is difficult to describe these processes accurately and quantitatively, so the model adopts the above processing method.

This relates to a technique for determining landing of oil particles. As can be seen from fig. 4, in this mode, a grid method is used to determine the condition of landing of the oil particles, the land line grid is consistent with the flow field grid, the land value is set to be 1 in the grid, the sea value is set to be 0, if the land line grid where the particles are located is determined to be 1, the particles are considered to be landing, otherwise, the particles are considered not to be landing, and the particles that are landed will not participate in the calculation in the model.

8. Oil spill weathering module

The module calculates evaporation, emulsification and density of spilled oil through an empirical model to simulate and predict the change conditions of the offshore residual oil quantity and the spilled oil density, and provides the calculation result to the transportation module.

The weathering process of the spill is quite complicated. The model mainly researches the evaporation process and the emulsification process of oil spilling. And are not considered for processes such as biodegradation and photooxidation.

8.1 Evaporation (Evaporation)

Evaporation is a very important process of oil spill and physicochemical changes within the first hour after the oil spill occurs. The proportion of evaporation is determined by oil, seawater temperature, wind speed and other processes (such as expansion and emulsification), and the system adopts a parameterized formula proposed by a driver and Mackay (1985) to calculate the evaporation rate and the evaporation amount.

The evaporation coefficient is defined as:

wherein k' is 0.78 × 2.5 × 10-3Uw, Uw is the wind speed 10 meters above sea surface, a is the area of oil film, V0 is the initial volume of oil spill, and t is time.

The evaporation rate is then a function of factors such as evaporation coefficient, boiling point temperature, etc.:

where FV is the evaporation rate, a 'is 6.3, B' is 10.3, TG is the gradient of the boiling point curve, T is the temperature of the oil, and T0 is the initial boiling point temperature of the oil (when FV is 0).

8.2 Emulsification (Emulsification)

During the emulsification process of the spilled oil, dispersed oil drops and seawater form a water-in-oil emulsion which is in a dark brown foam structure, and the water content of the emulsion can reach 80%. This emulsion has a high density and viscosity, so that it can influence the diffusion process of the oil spill. And presents difficulties in removing spilled oil. Light volatile oils rarely form emulsions, and heavy fuel oil or crude oil forms a substantial amount of emulsions. The emulsification process is influenced by factors such as wind speed, waves, oil thickness, ambient temperature, oil weathering process and the like, and the emulsification degree is generally represented by the water content YW. The formula for calculating the water content of the emulsion (Mackay et al (1980)) is:

wherein, Y_WIs the water content (%) of the emulsion, K_A＝4.5×10-8，U_wIs the wind speed, K _B1/YFW ≈ 1.25, YFW is the final moisture content, and t is time.

8.3 Density Change

The change in density will affect the change in buoyancy experienced by the oil particles and thus their vertical movement.

The effect of emulsification on oil density is expressed as:

ρ_e＝(1-Y_W)ρ₀+Y_W·ρ_W

where ρ is_eDensity of the emulsified oil, p₀Initial density of oil before emulsification, p_wSea water density, Y_W-emulsion moisture content.

The effect of evaporation on oil density is expressed as:

ρ＝(0.6·ρ₀-0.34)F_V+ρ₀

combining the effects of both, the density of the oil is expressed as:

ρ＝(1-Y_W)((0.6·ρ₀-0.34)F_V+ρ₀)+Y_W·ρ_W

the viscosity of the oil changes with temperature, and the model assumes that the temperature changes are small, so the influence of the temperature change on the viscosity can be ignored.

9. Visual analysis module

And analyzing the forecast data result based on matlab software, and drawing into pictures and animation modes, wherein the forecast content of analysis processing comprises the oil spill drifting track, the expansion range, the oil film thickness, the sea sweeping area and the time of the oil spill abutting against the bank or reaching a sensitive area.

Drawing the average positions of all oil particles per hour on a map by adopting a plot command in matlab on a drift track; the diffusion range adopts plot commands in matlab to draw the position of each oil particle in each hour on a map; the oil film thickness adopts matlab software to calculate the number of oil particles in each grid in a mode area range, and the relative thickness distribution condition of the oil film is drawn on a map by using a contourf command in the matlab; drawing the position of each oil particle on a map at all times by adopting a plot command in matlab in the sea sweeping area; and drawing the position of oil particles arriving at a shoreline and marking the corresponding time on a map by adopting a plot command in matlab when the spilled oil arrives at the shore or reaches a sensitive area.

Example 2

15 minutes in 20 days in 6 months in 2018, the Panama oil tanker 'Sangji wheel' and the Chinese hong Kong bulk cargo ship 'Changfeng crystal' wheel collide with the ship at the east longitude of 124 degrees 56.7 'and the northern latitude of 30 degrees 42.7' N in the open of Changjiang river, 32 people fall into water, oil stains are diffused, and the 'Sangji' oil tanker fires. 16 minutes at 14 months and 14 days, the tanker sinks at the east longitude 125 degrees 58.4 'and the north latitude 28 degrees 21.9', and the oil spilled from the ship is still burnt in the sinking sea area. 58 minutes after 09 days 15, the field naked fire disappears completely, the oil stains on the sea surface begin to drift in a large area, and the oil stains on the sea surface are changed into sporadic distribution by satellite monitoring until 25 days 1 month. By using the system, the oil spill drifting track and diffusion prediction of the oil spill accident are carried out by taking the site report sunken ship position and the oil spill position of the first satellite picture as the starting position and the starting time, the sunken ship position is set to be continuous oil spill, the satellite picture position is set to be instantaneous leakage, the prediction time is 20 minutes at 1 month, 15 months and 6 months in 2018, 20 minutes at 1 month, 16 months and 5 days in 2018, and is 24 hours in total, and the prediction time interval is 1 hour. Starting the system, selecting the oil spill type as a satellite picture on a main interface (figure 8) according to the oil spill type as the joint simulation of the satellite picture and a point source, setting satellite picture information (figure 9) and point source information (figure 10), and not starting assimilation and ensemble prediction in the prediction, so that only corresponding prediction environment information field data needs to be selected, and a model is operated. An oil particle center position drift trajectory diagram, an affected area diagram, a sweep area diagram, and an oil film thickness diagram were obtained (fig. 11 to 14).

And (3) according to the satellite picture obtained in the later stage, the prediction result is tested (figure 15), and because the satellite picture obtained in the later stage is 23 hours, the result obtained in 23 hours is tested, the distance error of the oil film center point is less than 10KM, the error is within an acceptable range, and the prediction result is good.

The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (3)

1. A northwest Pacific three-dimensional oil spill business emergency forecasting and evaluating system is characterized in that: the system comprises an oil spill information rapid processing module, an environmental information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an ensemble forecasting module, a geographic information data module, a visual analysis module and a system control center;

the oil spill information rapid processing module rapidly processes different types of accident source information and accident categories according to the instruction sent by the system control center and provides information of the oil spill source for the system control center;

the environment information processing module provides the sea surface wind, ocean current, ocean temperature and ocean wave field data information processing cost system with different sources and data formats to the system control center for storage according to the system control center instruction, so that the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module can read and use the spilled oil transportation module, the spilled oil weathering module, the data assimilation module, the ensemble forecasting module and the visualization analysis module;

the oil spill transportation module simulates the time-space behavior of oil spill in the marine environment by adopting an oil particle model method; the oil particle motion equation of the oil particle model is as follows:

u_oand v_oIs the horizontal velocity, w, of the oil particles_oIs the vertical velocity of the oil particles,<u′>，<v′>，<w′>the middle and front two are horizontal turbulence speed, and w' is vertical turbulence speed which is vertical turbulence diffusion speed; the first term on the right of the equal sign in the equation of motion of the oil particles is the flow velocity of the ocean current, u_cAnd v_cRepresenting the flow velocity in the horizontal direction at the oil particle position, w_cRepresenting the flow velocity in the vertical direction at the position of the oil particle, and representing the motion condition of the ocean current at the position of the oil particle from the simulation result of the ocean current mode; the second term on the right of the equal sign in the motion equation of the oil particles represents the dragging effect of wind on the oil particles, u_aAnd v_aThe wind speed is the wind speed at 10 m height above the sea surface where the oil particles are located, the wind speed is also obtained through the simulation result of an atmospheric numerical mode, and the effect is not considered when the oil particles are in water; alpha and beta are wind drift factors and deflection angles, the wind drift factors are usually 1% -6%, the absolute value of the deflection angle is usually 0-45 degrees, the right deflection is a negative value in the northern hemisphere, and the left deflection is a positive value in the southern hemisphere; u. of_wAnd v_wThe drift velocity of oil particles brought by the residual flow generated by the nonlinear wave; the surface tension is increased due to the existence of the oil film, so that the sea surface tends to be smooth, the nonlinear effect of sea waves is greatly weakened, and the corresponding wave residual flow is also greatly reduced, so that the motion of oil particles can be ignored when being calculated; w is a_okIs the vertical velocity under the action of buoyancy, according to Stokes law:

d radius of oil particles, d_cCritical radius of oil particles; the oil particle radius is 10-1000 muA log-normal distribution between m; v is the viscosity of water, ρ_oAnd ρ_wThe density of oil and water are respectively adopted, the density of oil is determined according to the type of oil product manually input in the modeling process, and the density of water is set to 1025kg m^-3The viscosity of water is 1.311mm³s^-1(ii) a The oil spill transportation module acquires information required by calculation from the system control center and provides a calculation result to the system control center; the model boundary condition in the oil particle model method adopts a non-reflection condition, the non-reflection condition is that when the oil particles in water rise to the sea surface, the oil particles become surface particles and move and diffuse along with surface flow together with other surface particles, the oil particles arrive at a land boundary in the drifting process, which is equal to that the oil particles are adhered to the land, the particles are also declared to be dead in the calculation and do not participate in the calculation any more, when the oil particles arrive at the boundary of a calculation area, the oil particles are also considered to be dead and do not participate in the calculation any more, and a grid method is adopted to judge the condition that the oil particles are in the bank;

the oil spill weathering module calculates evaporation, emulsification and density of oil spill through an empirical model to simulate and predict the change conditions of the marine residual oil quantity and the oil spill density, acquires information required by calculation from the system control center and provides the calculation result to the system control center;

the data assimilation module is started under the condition that a field observation result is read from the oil spill information rapid processing module, and environment field data are optimized by using an optimal interpolation assimilation method, so that the forecasting accuracy is improved; the control button of the data assimilation module is in a system control center and is triggered by the system control center according to whether field observation data exist or not; the optimal interpolation method is to adopt a statistical least square method to determine the optimal weight, directly interpolate the observation to the background field and obtain the optimal interpolation result, namely the obtained analysis field is a linear combination of the background field and the observation field under the meaning of minimum variance; the analytical equation of the optimal interpolation method is as follows:

x^a＝x^b+BH^T(HBH^T+R)^-1(y^o-H[x^b])

wherein x^bFor ambient fields, i.e. mode prediction results, y^oFor the observation field, i.e. the observation result, let:

K＝BH^T(HBH^T+R)^-1

and K is called a gain matrix or a weight matrix, wherein R is an observation error covariance matrix, diagonal elements of the matrix are observation error covariance values, other elements are 0, and B is a background error covariance matrix:

σ²is a static background error covariance constant value, L is a relevant scale factor, also a constant, Δ x²,Δy²For the distance matrix of each grid point of the background field and other grid points, H is a linear observation operator, x^aAn analysis field is obtained after the background field is corrected by an OI assimilation technology;

the ensemble forecasting module selects simulation results under parameters of ocean currents, wind fields, sea waves and oil spillage from different sources to carry out ensemble forecasting according to instructions of the system control center, defaults or manually selects an external forcing scheme and an oil spillage model parameter scheme, defaults or manually inputs weights of all schemes to obtain ensemble forecasting results, and provides the ensemble forecasting results to the system control center;

the geographic information data module processes the Chinese and western Pacific ocean geographic information data in the module database into water depth and sea-land grid data in a format required by current calculation according to an instruction of the system control center, and provides the data to the system control center;

the visual analysis module outputs the forecast result in a picture and animation mode according to the command of the system control center, and the forecast content comprises an oil drift track, a diffusion range, the thickness and the distribution of an oil film, an influence bank section range and shore arrival time, an influence sensitive area and arrival time and a sea sweeping area;

the system control center is a control instruction center and a data exchange center, controls an oil spill information rapid processing module, an environment information processing module, an oil spill transportation module, an oil spill weathering module, a data assimilation module, an ensemble prediction module, a geographic information data module and a visual analysis module, and selects an external forced field file and an input oil spill parameter manually or in a default mode; and the system control center completes data exchange among the modules according to the setting, realizes simulation calculation, sends a command to the visualization module, and draws a simulation result product.

2. The northwest pacific three-dimensional oil spill business emergency forecasting and evaluation system of claim 1 wherein: the oil spill information rapid processing module comprises an oil spill source initialization module and an oil spill field observation information processing module, wherein the oil spill source initialization module comprises processing of picture oil source information and data oil source information, for the picture oil source information, the module collects the coordinate position of a pollution area outline based on matlab software, digitalizes oil spill source observation picture information by using functions in the matlab, and for the data oil source information, the oil spill source initialization module converts the oil source information of point source, surface source continuous oil spill or instantaneous oil spill, continuous or instantaneous oil spill of a motion oil source into oil spill digital information in a format required by a model; the oil spill site observation information processing module processes oil spill observation time, place, range, oil spill amount, oil spill mode, oil product type, oil source moving speed, wind field, flow field, sea wave observed in the oil spill site, latest observation position, shape and time information, provides the information to the system control center for storage, and reads and uses the information for the data assimilation module, the oil spill transportation module and the oil spill weathering module; the picture oil source information comprises a remote sensing picture and an aerial picture.

3. The northwest pacific three-dimensional emergency oil spill forecasting and evaluation system of claim 1 wherein w' in the oil spill transportation module is a vertical turbulent diffusion velocity, which can be calculated by three vertical diffusion schemes, and can be compared, and the influence of vertical mixing on the temporal and spatial distribution of oil spill is revealed by comparing and analyzing the test results, and an optimal vertical diffusion scheme is determined and selected, wherein the three vertical diffusion schemes include a vertical diffusion scheme, B vertical diffusion scheme and C vertical diffusion scheme, and the a vertical diffusion scheme is as follows:

xi is standard white Gaussian noise, K_hIs a vertical diffusion coefficient, Δ t is a time step, and C' is a constant; w' is the vertical turbulent diffusion velocity;

in the A vertical diffusion scheme, the K_hIs K_waveSaid K is_waveFor vertical vortex viscosity, said K_waveThe semi-empirical formula for the Reynolds stress caused by sea surface waves is as follows

Wherein H_sThe effective wave height is represented, kappa is the wave number, T is the average period, and Z is the distance between the position of the oil particles and the sea surface; said H_sThe following formula is used for obtaining the effective wave height:

u_aand v_aThe wind speed is 10 m high above the sea surface where the oil particles are located;

the B vertical diffusion protocol was as follows:

xi is standard white Gaussian noise, K_hIs vertical diffusion coefficient, delta t is time step length, C' is constant, the above equation directly adopts vertical turbulence coefficient in the ocean current model as K in the equation_h(ii) a w' is the vertical turbulent diffusion velocity;

the C vertical diffusion protocol is as follows:

the above formula is a Langeven equation, where α' and λ are coefficients, determined by the variance and dispersion of the Stochartic process, and the relationship between the Langeven equation and the Markov chain is as follows

Wherein A, B and C in the formula are solved according to the following formula;

where σ is the root mean square of the turbulent velocity, T_LFor Lagrange integration time scale, the T_LSolving for Lagrange integration time scale according to the following formula;

b is turbulent kinetic energy, epsilon is turbulent energy dissipation rate, and w' is vertical turbulent diffusion speed.

Images