Ozone simulation method and device based on multi-source data assimilation, terminal and medium

Abstract

The invention discloses an ozone simulation method and device based on multi-source data assimilation, a terminal and a medium. The method comprises the following steps: acquiring spatial and temporal distribution data of ozone column content of a geostationary orbit satellite; according to the spatial and temporal distribution data, ozone column content data of the geostationary orbit satellite are obtained through inversion, and the ozone column content data comprise ozone column content distribution conditions, observation values, inversion errors, priori values and an average kernel function matrix; inputting ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilation data, and performing ozone simulation according to the multi-source assimilation data. The invention develops a multi-source data assimilation system for data of ground monitoring, low-orbit satellites and geostationary orbit satellites, and ozone observation data of the geostationary orbit satellites are taken as one of reference factors of ozone simulation, so that the accuracy of ozone simulation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ozone simulation, and in particular to an ozone simulation method, device, terminal and medium based on multi-source data assimilation. Background Art

[0002] Numerical models play an important role in the prevention and control of atmospheric ozone pollution. Since numerical simulation of ozone is easily limited by factors such as meteorological fields, chemical initial fields, emission source lists, and immature physical and chemical parameterization schemes of models, it often has inevitable uncertainties. Therefore, improving the effect of ozone simulation is of great practical significance.

[0003] Data assimilation technology can reduce the prediction error of simulation results and improve the accuracy of ozone forecasts by correlating observation data (ground monitoring, satellite remote sensing, lidar, sensors, aircraft, sounding observations, etc.) and numerical models. However, most of the current research on ozone data assimilation uses low-orbit satellite data. Due to the long revisit period, the temporal resolution of low-orbit satellite data is low, and the improvement effect on ozone simulation is limited. With the successful launch of the GEMS satellite, mankind has entered a new era of geostationary satellite monitoring of air quality. Compared with low-orbit satellites, geostationary satellites can realize real-time daytime monitoring of atmospheric pollutants, which greatly enriches the source of assimilation data and provides new opportunities for improving ozone numerical simulation.

[0004] However, the existing technology does not consider ground monitoring, low-orbit satellite and geostationary orbit satellite data at the same time, and cannot implement an ozone simulation solution that assimilates multi-source stereo data, resulting in low accuracy of ozone simulation. Therefore, the existing technology needs to be improved and enhanced. Summary of the invention

[0005] The technical problem to be solved by the present invention is that, in view of the above-mentioned defects of the prior art, an ozone simulation method, device, terminal and medium based on multi-source data assimilation are provided, aiming to solve the problem that the prior art does not simultaneously consider ground monitoring, low-orbit satellite and geostationary orbit satellite data, and cannot realize the ozone simulation scheme of multi-source stereo data assimilation, resulting in low accuracy of ozone simulation.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, the present invention provides an ozone simulation method based on multi-source data assimilation, wherein the method comprises:

[0008] Obtain the temporal and spatial distribution data of ozone column content from geostationary orbit satellites;

[0009] According to the spatiotemporal distribution data, inverting to obtain ozone column content data of a geostationary orbit satellite, the ozone column content data including ozone column content distribution, observation value, inversion error, prior value and average kernel function matrix;

[0010] The ground monitoring data, low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilation data, and ozone simulation is performed based on the multi-source assimilation data.

[0011] In one implementation, the step of obtaining the spatiotemporal distribution data of the ozone column content of a geostationary orbit satellite includes:

[0012] Determine a target monitoring area of ​​a geostationary orbit satellite, and use an atmospheric chemistry online coupling model to simulate the atmosphere of the target monitoring area to obtain a real atmosphere;

[0013] The target observation time corresponding to the target monitoring area is determined, and satellite sampling of the real atmosphere is performed based on the target observation time to obtain the spatiotemporal distribution data of the ozone column content.

[0014] In one implementation, the ozone column content data of the geostationary orbit satellite is inverted according to the spatiotemporal distribution data, and the ozone column content data includes ozone column content distribution, observation value, inversion error, prior value and average kernel function matrix, including:

[0015] Inputting the spatiotemporal distribution data into a preset radiation transfer model to obtain a real atmospheric ozone column content profile, wherein the real atmospheric ozone column content profile reflects an observed value;

[0016] The climatological ozone column content profile and the real atmospheric ozone column content profile are input into a preset inversion model to invert the ozone column content data of the geostationary orbit satellite, wherein the climatological ozone column content profile reflects a priori values.

[0017] In one implementation, the step of inputting the climatological ozone column content profile and the real atmospheric ozone column content profile into a preset inversion model to invert and obtain the ozone column content data of a geostationary orbit satellite includes:

[0018] The priori value and the observed value are input into the inversion model for inversion calculation to obtain the ozone column content data, wherein the inversion calculation process is: p =x s +A(x ap -x s )+ε,x′ p is the inversion profile of the pth layer surface, x sis the observed value, ε is the inversion error, x ap is the prior value, and A is the average kernel function matrix.

[0019] In one implementation, the ground monitoring data, the low-orbit satellite data, and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and ozone simulation is performed based on the multi-source assimilated data, including:

[0020] Determining an observation error of the observation data based on the ground monitoring data, the low-orbit satellite data and the ozone column content data;

[0021] Inputting the priori error of the forecast field and the observation error into a preset multi-source data assimilation system for data assimilation to obtain a model error;

[0022] An optimal estimate is determined according to the observation error and the model error, multi-source assimilation data is obtained based on the optimal estimate, and ozone simulation is performed based on the multi-source assimilation data to obtain an ozone concentration analysis field.

[0023] In one implementation, determining the optimal estimate value according to the observation error and the pattern error includes:

[0024] Construct observation error covariance matrix and model error covariance matrix;

[0025] Determining a Kalman gain according to the observation error covariance matrix and the pattern error covariance matrix;

[0026] An optimal estimation value is determined based on the observation error, the model error and the Kalman gain.

[0027] In one implementation, after performing ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field, the method includes:

[0028] determining a frequency of the data assimilation, and determining a first assimilation time and a second assimilation time based on the frequency;

[0029] The forecast field at the second assimilation time is updated to the ozone concentration analysis field at the first assimilation time.

[0030] In a second aspect, an embodiment of the present invention further provides an ozone simulation device based on multi-source data assimilation, wherein the device comprises:

[0031] A geostationary orbit satellite data acquisition module is used to obtain the temporal and spatial distribution data of the ozone column content of the geostationary orbit satellite;

[0032] An ozone column content data inversion module is used to invert the ozone column content data of the geostationary orbit satellite according to the spatiotemporal distribution data, wherein the ozone column content data includes the distribution of ozone column content, observation value, inversion error, prior value and average kernel function matrix;

[0033] The multi-source data assimilation and ozone simulation module is used to input ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and perform ozone simulation based on the multi-source assimilated data.

[0034] In a third aspect, an embodiment of the present invention further provides a terminal, wherein the terminal includes a memory, a processor, and an ozone simulation program based on multi-source data assimilation stored in the memory and executable on the processor, and when the processor executes the ozone simulation program based on multi-source data assimilation, the steps of the ozone simulation method based on multi-source data assimilation of any one of the above-mentioned schemes are implemented.

[0035] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein an ozone simulation program based on multi-source data assimilation is stored on the computer-readable storage medium, and when the ozone simulation program based on multi-source data assimilation is executed by a processor, the steps of the ozone simulation method based on multi-source data assimilation described in any one of the above schemes are implemented.

[0036] Beneficial effects: Compared with the prior art, the present invention provides an ozone simulation method, device, terminal and medium based on multi-source data assimilation. When implemented specifically, the present invention first obtains the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite. Then, based on the spatiotemporal distribution data, the ozone column content data of the geostationary orbit satellite is inverted, and the ozone column content data includes the distribution of the ozone column content, the observation value, the inversion error, the prior value and the average kernel function matrix. Finally, the ground monitoring data, the low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and ozone simulation is performed based on the multi-source assimilated data. It can be seen that the present invention has developed a multi-source data assimilation system for ground monitoring, low-orbit satellite and geostationary orbit satellite data, and uses the ozone observation data of the geostationary orbit satellite as one of the reference factors for ozone simulation, thereby improving the accuracy of ozone simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart of a specific implementation of the ozone simulation method based on multi-source data assimilation provided in an embodiment of the present invention.

[0038] Figure 2A framework diagram of the overall technical route of the ozone simulation method based on multi-source data assimilation provided in an embodiment of the present invention.

[0039] Figure 3 A comparison chart of simulated ozone concentration and observed ozone concentration in different regions provided by an embodiment of the present invention.

[0040] Figure 4 The inversion value and inversion error diagram of the ozone column content profile provided by the embodiment of the present invention.

[0041] Figure 5 This is a diagram of the average kernel function matrix provided by an embodiment of the present invention.

[0042] Figure 6 A Kalman filter assimilation flow chart provided for an embodiment of the present invention.

[0043] Figure 7 A comparison diagram of the geostationary orbit satellite ozone column content profile before and after assimilation provided in an embodiment of the present invention.

[0044] Figure 8 A functional schematic diagram of an ozone simulation device based on multi-source data assimilation provided in an embodiment of the present invention.

[0045] Fig. 9 A functional block diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0047] Those skilled in the art will appreciate that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Unless specifically stated, the singular forms "one", "said", and "the" used herein may also include plural forms, and the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or their groups. It should be understood that when an element is "connected" or "coupled" to another element, it may be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0048] The present embodiment provides an ozone simulation method, device, terminal and medium based on multi-source data assimilation. When implemented specifically, the present embodiment first obtains the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite. Then, based on the spatiotemporal distribution data, the ozone column content data of the geostationary orbit satellite is inverted, and the ozone column content data includes the distribution of the ozone column content, the observation value, the inversion error, the prior value and the average kernel function matrix. Finally, the ground monitoring data, the low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and ozone simulation is performed based on the multi-source assimilated data. Therefore, the present embodiment develops a multi-source data assimilation system for ground monitoring, low-orbit satellite and geostationary orbit satellite data, and uses the ozone observation data of the geostationary orbit satellite as one of the reference factors for ozone simulation, thereby improving the accuracy of ozone simulation.

[0049] The ozone simulation method based on multi-source data assimilation of this embodiment can be applied to a terminal, which can be an external device connected to the ozone simulation device based on multi-source data assimilation, or a device built into the ozone simulation device based on multi-source data assimilation. Figure 1 As shown in , the ozone simulation method based on multi-source data assimilation of this embodiment includes the following steps:

[0050] Step S100: Acquire the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite.

[0051] Since the present embodiment uses the ozone observation data of the geostationary orbit satellite (i.e., the GEMS satellite) as one of the reference factors for ozone simulation, the present embodiment first needs to obtain the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite, and then invert the ozone column content data of the geostationary orbit satellite based on the spatiotemporal distribution data, and use it as the ozone observation data, thereby providing a reference for subsequent ozone simulation.

[0052] In one implementation, step S100 of this embodiment specifically includes the following steps:

[0053] Step S101, determining a target monitoring area of ​​a geostationary orbit satellite, and using an atmospheric chemistry online coupling mode to perform atmospheric simulation on the target monitoring area to obtain a real atmosphere;

[0054] Step S102: determine the target observation time corresponding to the target monitoring area, perform satellite sampling on the real atmosphere based on the target observation time, and obtain the spatiotemporal distribution data of the ozone column content.

[0055] In specific implementation, since it is impossible to know the ground-based ozone monitoring data (including the physical and chemical state of the atmosphere) before the geostationary orbit satellite data is officially released, this embodiment adopts the observation system simulation test method to "construct the real atmosphere" in order to predict the ground-based ozone monitoring data of the geostationary orbit satellite in advance, such as Figure 2 As shown, the observation system simulation test in this embodiment chooses to adopt the atmospheric chemistry online coupling mode (i.e., WRF-Chem mode). The WRF-Chem mode is a new generation of regional air quality mode developed by the Forecast System Laboratory (FSL) of the National Oceanic and Atmospheric Administration (NOAA) of the United States. The biggest feature of this mode is that it realizes the online coupling of the meteorological mode (WRF) and the chemical mode (Chem). The chemical process and the meteorological process use the same horizontal and vertical coordinate system and the same physical parameterization scheme. There is no interpolation in time and space, and the feedback effect of the chemical process on the meteorological process can be considered.

[0056] Specifically, the present embodiment first needs to determine the target monitoring area of ​​the geostationary orbit satellite, and then use the atmospheric chemistry online coupling mode to simulate the atmosphere of the target monitoring area to obtain the real atmosphere. The present embodiment has compared the monthly average ozone concentration between the monthly average ozone data of the Chinese region obtained by the WRF-Chem mode and the monthly average ozone data actually obtained by the station observation. It can be seen that the ozone data simulated by the WRF-Chem mode in the present embodiment is basically consistent with the actual detected monthly average ozone data, which verifies that the real atmosphere obtained by the atmospheric chemistry online coupling mode in the present embodiment has a good simulation effect.

[0057] In addition, this embodiment also selects typical seasons or months in recent years when ozone pollution exceeds the standard seriously (such as April-September 2019) as the research period of numerical simulation, and the research area with more serious ozone pollution as the target monitoring area. The target monitoring area in this embodiment is the Guangdong-Hong Kong-Macao Greater Bay Area, which is in the field of view of geostationary orbit satellites (5°S to 45°N, 75°E to 145°E). The monthly average ground ozone in the Guangdong-Hong Kong-Macao Greater Bay Area obtained in this way is exactly the "real atmosphere" obtained by the WRF-Chem model in this embodiment.

[0058] Figure 3 In this example, the simulated ozone concentrations in several different areas of the Guangdong-Hong Kong-Macao Greater Bay Area are compared with the observed ozone concentrations. Figure 3 The broken line formed by the dots in the figure represents the measured ozone concentration on the ground, and the broken line formed by the black segments represents the ozone concentration simulated by the WRF-Chem model. The comparison results of ozone concentrations in Shenzhen, Guangzhou, Zhuhai and Zhongshan in the Guangdong-Hong Kong-Macao Greater Bay Area also verify that the ozone concentration obtained by the atmospheric chemistry online coupling model in this embodiment has a good simulation effect.

[0059] Among them, the settings of the WRF-Chem model in this embodiment are as follows: the simulation adopts two layers of nesting, with resolutions of 60 and 20 km from the outer layer to the inner layer respectively. The meteorological field uses the meteorological grid data provided by the NCEP FNL reanalysis data with a resolution of 0.25°×0.25°. The simulation adopts the CBMZ gas phase chemical mechanism and the MOSIAC 4bin aerosol chemical mechanism. The anthropogenic emissions in the Chinese region adopt the 2016 MEIC emission inventory with a spatial resolution of 0.25°×0.25°, and the anthropogenic emissions outside the Chinese region adopt the 2010 EDGAR-HTAP emission inventory with a spatial resolution of 0.1°×0.1°. Natural source and biomass burning emissions are from MEGAN and FINN v1.5 respectively. The WRF-Chem chemical initial field and boundary field are from the simulation output of the WACCM global model.

[0060] Furthermore, the present embodiment also needs to determine the target observation time corresponding to the target monitoring area. Since the observation fields of the GEMS satellites in different months and at different times are different, the present embodiment needs to select the time when the corresponding observation field of the GEMS satellite covers the Chinese region according to the research period, that is, the target observation time. For example, for summer (June-August), the target observation time is 01-08 Universal Time, and the spatial resolution of the GEMS satellite observation is 7km×8km. Then, the present embodiment performs satellite sampling of the real atmosphere based on the target observation time, that is, according to the ozone concentration, meteorological parameters and grid parameters of the "real atmosphere", the layered column content of ozone is calculated, and it is interpolated to the observation pixels of the GEMS satellite to obtain the spatiotemporal distribution data of the ozone column content consistent with the spatial resolution of the GEMS satellite. The spatiotemporal distribution data will serve as important input information for the subsequent radiation transfer model calculation.

[0061] Step S200: invert the ozone column content data of the geostationary orbit satellite according to the spatiotemporal distribution data, wherein the ozone column content data includes ozone column content distribution, observation value, inversion error, prior value and average kernel function matrix.

[0062] like Figure 2 As shown, in this embodiment, after obtaining the spatiotemporal distribution data through satellite sampling, the spatiotemporal distribution data is input into the radiation transfer model and the inversion model, and then the ozone column content data of the geostationary orbit satellite can be obtained through the inversion calculation of the model.

[0063] In one implementation, step S200 of this embodiment specifically includes the following steps:

[0064] Step S201, inputting the spatiotemporal distribution data into a preset radiation transfer model to obtain a real atmospheric ozone column content profile, wherein the real atmospheric ozone column content profile reflects an observed value;

[0065] Step S202: input the climatological ozone column content profile and the real atmospheric ozone column content profile into a preset inversion model, and invert to obtain the ozone column content data of the geostationary orbit satellite, wherein the climatological ozone column content profile reflects a priori values.

[0066] Specifically, the GEMS satellite ozone column profile is divided into 24 layers in the vertical direction, starting from the surface (P 0 :0km) to the 23rd floor (P 23 : about 55km), each layer is about 2.5km thick. 24 ) is set to about 65km, the top layer thickness is about 10km, and each layer (P i ) is expressed in logarithmic units of air pressure: Here, atm refers to standard atmospheric pressure. In this embodiment, the distribution of ozone column content in the surface layer is used to show the "satellite sampling" results of the "real atmosphere".

[0067] During the specific implementation process, this embodiment inputs the spatiotemporal distribution data into a preset radiation transfer model to obtain the ozone column content profile of the real atmosphere (i.e., the observed value), and then obtains the climatic ozone column content profile (i.e., the prior value) by summarizing the typical climate characteristics of a certain area over a long period of time. Finally, the climatic ozone column content profile and the ozone column content profile of the real atmosphere are input into a preset inversion model, and the ozone column content data of the geostationary orbit satellite can be inverted. Among them, the ozone column content data in this embodiment can also reflect four-dimensional ozone column content data.

[0068] In one implementation, step S202 of this embodiment specifically includes the following steps:

[0069] Step S212: Input the priori value and the observed value into the inversion model for inversion calculation to obtain the ozone column content data, wherein the inversion calculation process is: p =x s +A(x ap -x s )+ε,x′ p is the inversion profile of the pth layer surface, x s is the observed value, ε is the inversion error, x ap is the prior value, and A is the average kernel function matrix.

[0070] Specifically, the inversion model of this embodiment is based on the optimal estimation method, and uses an iterative method under the prior value constraint to find the solution when the cost function reaches the minimum value, that is, x′ p =x s +A(x ap -x s )+ε. In the inversion algorithm, this embodiment adopts the principal component analysis (PCA) and table lookup correction method to improve the calculation speed of the ozone column content profile inversion. Through the inversion calculation, this embodiment can obtain the ozone column content data, including the inversion profile of the p-th layer of the surface, the observation value, the inversion error, the prior value and the average kernel function matrix. For example, this embodiment will first obtain an example of the inversion of the ozone column content in the GEMS satellite troposphere (below 200hPa) in China in the spectral range of 300nm to 340nm based on the inversion model. Then, Figure 4 The inversion value and inversion error diagram of the ozone column content profile obtained according to the inversion model in this embodiment are as follows: Figure 4 (1) is the inversion value and inversion error of the GEMS satellite ozone column profile for Shenzhen, a representative city in the Guangdong-Hong Kong-Macao Greater Bay Area. Figure 4 (2) is the inversion value and inversion error of the GEMS satellite ozone column profile for Guangzhou, a representative city in the Guangdong-Hong Kong-Macao Greater Bay Area. Figure 4 (1) and Figure 4 The black solid line, red solid line and blue solid line in (2) represent the prior ozone profile, the real atmospheric ozone profile and the GEMS satellite inverted ozone profile, respectively. The black dashed line, red dashed line and blue dashed line represent the prior error, inversion solution error and random error, respectively. Figure 5 The average kernel function matrix diagram of the GEMS satellite ozone column content profile obtained according to the inversion model in this embodiment (normalized according to the height and prior error), Figure 5 (1) Shenzhen area, Figure 5 (2) is the Guangzhou area, where the lines of different colors represent the sensitivity of the GEMS satellite ozone inversion values ​​at different altitudes to the observation values ​​at different altitudes, and the numerical values ​​represent the air pressure (P, unit: hPa) and altitude (Z, unit: km) of different layers.

[0071] Step S300: input the ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and perform ozone simulation based on the multi-source assimilated data.

[0072] Specifically, after obtaining the ozone column content data about the geostationary orbit satellite in this embodiment, a multi-source data assimilation system is also developed. It is only necessary to input the ground monitoring, low-orbit satellite and geostationary orbit satellite data into the multi-source data assimilation system to complete the assimilation between the multi-source data, so that in the subsequent process, ozone simulation can be performed based on the multi-source assimilated data to improve the accuracy of ozone simulation.

[0073] In one implementation, step S300 of this embodiment specifically includes the following steps:

[0074] Step S301, determining the observation error of the observation data according to the ground monitoring data, the low-orbit satellite data and the ozone column content data;

[0075] Step S302, inputting the forecast field prior error and the observation error into a preset multi-source data assimilation system for data assimilation to obtain a model error;

[0076] Step S303: determining an optimal estimate value according to the observation error and the model error, obtaining multi-source assimilation data based on the optimal estimate value, and performing ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field.

[0077] Specifically, since the global three-dimensional atmospheric chemical transmission model (GEOS-Chem v12.9.3) can also be used for the calculation of atmospheric quality factors in the development of ozone satellite remote sensing products and the later verification of ozone satellite remote sensing products when simulating ozone profiles, compared with other models such as NU-WRF and CMAQ, it has the effect of multi-source stereo ozone observation data assimilation, so the global three-dimensional atmospheric chemical transmission model is selected as the assimilation platform in this embodiment for the development of a multi-source data assimilation system. Those skilled in the art may also select other modes according to current needs, which are not specifically limited in this embodiment.

[0078] like Figure 6 As shown, this embodiment can first determine the observation error of the observation data based on ground monitoring data, low-orbit satellite data and ozone column content data. Then it is determined based on the ozone data monitored at the forecast site, and finally the a priori error and the observation error are input into a preset multi-source data assimilation system for data assimilation to obtain the model error. Optionally, the Kalman filter method is used in this embodiment to complete data assimilation. The advantage of the Kalman filter method is that it can provide sequential estimation of the optimal state of the system and information on the system error, and is suitable for assimilation applications from real-time observation data to forecast models.

[0079] In one implementation, this embodiment determines the optimal estimation value according to the observation error and the pattern error, specifically including the following steps:

[0080] Step S331, constructing an observation error covariance matrix and a pattern error covariance matrix;

[0081] Step S332, determining the Kalman gain according to the observation error covariance matrix and the pattern error covariance matrix;

[0082] Step S333: determine an optimal estimate value according to the observation error, the pattern error and the Kalman gain.

[0083] In the specific implementation process, the observation error ε′ and the model error ε of this embodiment are a It is described by the error covariance matrix, so this embodiment first needs to construct the observation error covariance matrix and the pattern error covariance matrix, where the observation error covariance matrix is:

[0084] S ε =E[ε′ε′ T ], E is the expected function, ε′ is the observation error, ε′ T is the transposed matrix of the observation error. The model error covariance matrix is:

[0085] ε a is the model error, is the transposed matrix of the mode error, ε a =(ε a，1 , ..., ε a，n ) T , ε a,i represents the error of the mode grid i, and can also be understood as:

[0086] Among them, the diagonal terms of the error covariance matrix are the initial errors, and the off-diagonal terms are functions of the distance between different grids and the spatial correlation scale.

[0087] Next, this embodiment obtains the Kalman gain (i.e., the weight coefficient matrix) by algebraic calculation based on the observation error covariance matrix and the model error covariance matrix, where the Kalman gain is:

[0088] G=S a K T (KS a K T +S ε ) -1 , K is the observation operator, T is the transposed matrix, K T is the transposed matrix of the observation operator. Furthermore, the observation operator K will also transform the pattern variables from the pattern space to the observation position, including spatial transformation (such as interpolation operator S) and physical transformation (such as unit conversion), which is specifically expressed as:

[0089] Kx a =x ap +A(S(x a )-x ap ), x a is the model prediction value, S(x a ) is to perform spatial transformation on the model prediction value. Finally, this embodiment determines the optimal estimate value according to the observation error, model error and Kalman gain:

[0090] x′ p is the observation value to be assimilated at the current moment. And, the optimal estimate Error use Describe, I n At this time, the present embodiment can obtain multi-source assimilation data according to the optimal estimation value, and complete the ozone simulation according to the multi-source assimilation data to obtain the ozone concentration analysis field.

[0091] In one implementation, after performing ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field, the present embodiment includes:

[0092] Step S304: determining the frequency of the data assimilation, and determining a first assimilation time and a second assimilation time based on the frequency;

[0093] Step S305: Update the forecast field at the second assimilation time to the ozone concentration analysis field at the first assimilation time.

[0094] like Figure 6 As shown, after performing ozone simulation based on multi-source assimilation data and obtaining the ozone concentration analysis field, this embodiment will also update the forecast field based on the ozone concentration analysis field. Specifically, this embodiment first needs to determine the frequency of data assimilation. For example, in the ozone observation data of the GEMS satellite whose target observation time is 01-08 UTC (i.e. 09-16 Beijing time), the frequency of data assimilation is once an hour. At this time, the first assimilation moment can be determined as the current assimilation moment, and the second assimilation moment is the moment one hour after the current assimilation moment. In other words, after this embodiment determines the ozone concentration analysis field at the first assimilation moment, it will also promptly update the forecast field at the next assimilation moment based on this ozone concentration analysis field, thereby improving the real-time performance and reliability of data in the data assimilation process.

[0095] In one implementation, Figure 2As shown, this embodiment will also quantitatively evaluate the improvement of ozone simulation in the target monitoring area after obtaining multi-source assimilation data, which is specifically manifested in the numerical simulation and evaluation of the control run and assimilation run using the GEOS-Chem model. For example, when the WRF-Chem model and the GEOS-Chem model are used to collect the distribution of ozone concentration in the horizontal and vertical directions during the same period (2019), the two models simulate different meteorological field data, chemical mechanisms and emission source lists, among which the real atmospheric observation data obtained by the WRF-Chem model and the assimilated simulation data obtained by the GEOS-Chem model can be obtained by comparing the data collected by the two different models: whether it is the ozone data of the surface layer, the ozone data at a surface pressure of 850hPa, the ozone data at a surface pressure of 700hPa, or the ozone data at a surface pressure of 500hPa, the experimental results show that there are obvious differences in the ozone concentration data under the WRF-Chem model and the GEOS-Chem model. It can be seen that this difference can ensure the reliability and rigor of the observation system simulation test method, and to a large extent can avoid the systematic errors that may be caused by the similarities between models.

[0096] In another implementation, this embodiment will also evaluate the numerical simulation of the WRF-Chem model simulation (true value test) and the GEOS-Chem model simulation (control test). For example, this embodiment compares the ozone concentrations simulated and calculated by the two groups of tests (control test and assimilation test) with the true value test, statistically analyzes the evaluation indicators such as the root mean square error, average error, relative error, and ozone pollution event misjudgment rate, and gives the magnitude of the improvement in the surface and tropospheric ozone simulation performance, thereby quantifying the improvement of multi-source stereo data on the simulation of surface and tropospheric ozone pollution in the Guangdong-Hong Kong-Macao Greater Bay Area.

[0097] In another implementation, Figure 7 As shown, this embodiment also compares the GEMS satellite ozone column content profile before and after assimilation. Figure 7 (1) is the ozone data before assimilation of GEMS satellite ozone column content profile, Figure 7 (2) is the ozone data after assimilation of the GEMS satellite ozone column content profile, Figure 7 (1) and Figure 7The black line in (2) represents the ozone profile inverted by the GEMS satellite, the red line and the blue line represent the GEOS-Chem ozone profile simulation values ​​from the GEMS satellite perspective before and after assimilation, respectively. The numerical value represents the total column content of ozone at an air pressure below 200 hPa, and the numerical value in brackets represents the deviation from the GEMS ozone profile inversion value. It can be seen that in the test experiment of assimilating the GEMS satellite ozone column content profile of a single grid (at different ozone pollution levels) in this embodiment, after data assimilation, the simulated deviations of ozone at the two stations are increased from -15.6% and -10.3% to -9.1% and -8.1%, respectively, and the simulated effect of the ozone profile and the near-ground ozone column content at an air pressure below 200 hPa (red line) is closer to the observed value (black line), which further verifies that the data assimilation method based on the GEOS-Chem model in this embodiment can improve the simulation effect of surface and tropospheric ozone pollution in the target monitoring area.

[0098] In summary, this embodiment provides an ozone simulation method based on multi-source data assimilation. When implemented specifically, this embodiment first obtains the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite. Then, based on the spatiotemporal distribution data, the ozone column content data of the geostationary orbit satellite is inverted, and the ozone column content data includes the distribution of ozone column content, observation value, inversion error, prior value and average kernel function matrix. Finally, the ground monitoring data, low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and ozone simulation is performed based on the multi-source assimilated data. It can be seen that this embodiment has developed a multi-source data assimilation system for ground monitoring, low-orbit satellite and geostationary orbit satellite data, and uses the ozone observation data of the geostationary orbit satellite as one of the reference factors for ozone simulation, thereby improving the accuracy of ozone simulation.

[0099] Based on the above embodiments, the present invention also provides an ozone simulation device based on multi-source data assimilation, such as Figure 8As shown in , the ozone simulation device based on multi-source data assimilation includes: a geostationary orbit satellite data acquisition module 10, an ozone column content data inversion module 20, and a multi-source data assimilation and ozone simulation module 30. Specifically, the geostationary orbit satellite data acquisition module 10 is used to obtain the spatiotemporal distribution data of the ozone column content of the geostationary orbit satellite. The ozone column content data inversion module 20 is used to invert the ozone column content data of the geostationary orbit satellite according to the spatiotemporal distribution data, and the ozone column content data includes the distribution of ozone column content, observation values, inversion errors, prior values ​​and average kernel function matrices. The multi-source data assimilation and ozone simulation module 30 is used to input ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation, obtain multi-source assimilated data, and perform ozone simulation based on the multi-source assimilated data.

[0100] In one implementation, the geostationary orbit satellite data acquisition module 10 includes:

[0101] A real atmosphere simulation unit is used to determine a target monitoring area of ​​a geostationary orbit satellite, and to simulate the atmosphere of the target monitoring area using an atmospheric chemistry online coupling mode to obtain a real atmosphere;

[0102] The spatiotemporal distribution data unit is used to determine the target observation time corresponding to the target monitoring area, perform satellite sampling on the real atmosphere based on the target observation time, and obtain the spatiotemporal distribution data of the ozone column content.

[0103] In one implementation, the ozone column content data inversion module 20 includes:

[0104] An ozone column content profile determination unit, used for inputting the spatiotemporal distribution data into a preset radiation transfer model to obtain an ozone column content profile of the real atmosphere, wherein the ozone column content profile of the real atmosphere reflects an observed value;

[0105] The ozone column content data determination unit is used to input the climatological ozone column content profile and the ozone column content profile of the real atmosphere into a preset inversion model to invert the ozone column content data of the geostationary orbit satellite, wherein the climatological ozone column content profile reflects a priori values.

[0106] In one implementation, the ozone column content data determination unit includes:

[0107] The ozone column content data determination subunit is used to input the priori value and the observed value into the inversion model for inversion calculation to obtain the ozone column content data, wherein the inversion calculation process is: p =x s +A(xap -x s )+ε,x′ p is the inversion profile of the pth layer surface, x s is the observed value, ε is the inversion error, x ap is the prior value, and A is the average kernel function matrix.

[0108] In one implementation, the multi-source data assimilation and ozone simulation module 30 includes:

[0109] an observation error determination unit, configured to determine an observation error of the observation data based on the ground monitoring data, the low-orbit satellite data, and the ozone column content data;

[0110] A model error determination unit, used for inputting the priori error of the forecast field and the observation error into a preset multi-source data assimilation system for data assimilation to obtain a model error;

[0111] The ozone simulation unit is used to determine an optimal estimate value according to the observation error and the model error, obtain multi-source assimilation data based on the optimal estimate value, and perform ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field.

[0112] In one implementation, determining the optimal estimate value according to the observation error and the pattern error includes:

[0113] A covariance matrix construction unit, used to construct an observation error covariance matrix and a pattern error covariance matrix;

[0114] A Kalman gain determination unit, configured to determine a Kalman gain according to the observation error covariance matrix and the pattern error covariance matrix;

[0115] The optimal estimation value determination module is used to determine the optimal estimation value according to the observation error, the pattern error and the Kalman gain.

[0116] In one implementation, after performing ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field, the method includes:

[0117] A data assimilation frequency determination unit, used to determine the frequency of the data assimilation, and determine a first assimilation time and a second assimilation time based on the frequency;

[0118] A forecast field updating unit is used to update the forecast field at the second assimilation time to the ozone concentration analysis field at the first assimilation time.

[0119] The working principles of each module in the ozone simulation device based on multi-source data assimilation in this embodiment are the same as the principles of each step in the above method embodiment, and will not be repeated here.

[0120] Based on the above embodiment, the present invention further provides a terminal, the principle block diagram of the terminal can be as follows: Fig. 9 The terminal may include one or more processors 100 ( Fig. 9 Only one is shown), a memory 101 and a computer program 102 stored in the memory 101 and executable on one or more processors 100, for example, an ozone simulation program based on multi-source data assimilation. When one or more processors 100 execute the computer program 102, the various steps in the embodiment of the ozone simulation method based on multi-source data assimilation can be implemented. Alternatively, when one or more processors 100 execute the computer program 102, the functions of the various modules / units in the embodiment of the ozone simulation device based on multi-source data assimilation can be implemented, which is not limited here.

[0121] In one embodiment, the processor 100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0122] In one embodiment, the memory 101 may be an internal storage unit of an electronic device, such as a hard disk or memory of the electronic device. The memory 101 may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device. Further, the memory 101 may also include both an internal storage unit of the electronic device and an external storage device. The memory 101 is used to store computer programs and other programs and data required by the terminal. The memory 101 may also be used to temporarily store data that has been output or is to be output.

[0123] Those skilled in the art will understand that Fig. 9The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the terminal to which the scheme of the present invention is applied. The specific terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0124] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, operating database or other media used in the embodiments provided by the present invention may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double operational data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0125] In summary, the present invention discloses an ozone simulation method, device, terminal and medium based on multi-source data assimilation, the method comprising: obtaining the spatiotemporal distribution data of the ozone column content of a geostationary orbit satellite; inverting the ozone column content data of the geostationary orbit satellite according to the spatiotemporal distribution data, the ozone column content data comprising the distribution of the ozone column content, observation values, inversion errors, prior values ​​and average kernel function matrix; inputting the ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and performing ozone simulation according to the multi-source assimilated data. The present invention has developed a multi-source data assimilation system for ground monitoring, low-orbit satellite and geostationary orbit satellite data, taking the ozone observation data of the geostationary orbit satellite as one of the reference factors for ozone simulation, thereby improving the accuracy of ozone simulation.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An ozone simulation method based on multi-source data assimilation, characterized in that: The method comprises: Obtain the temporal and spatial distribution data of ozone column content from geostationary orbit satellites; According to the spatiotemporal distribution data, inverting to obtain ozone column content data of a geostationary orbit satellite, the ozone column content data including ozone column content distribution, observation value, inversion error, prior value and average kernel function matrix; The ground monitoring data, low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilation data, and ozone simulation is performed based on the multi-source assimilation data.

2. The ozone simulation method based on multi-source data assimilation according to claim 1, characterized in that: The step of obtaining the temporal and spatial distribution data of the ozone column content of the geostationary orbit satellite comprises: Determine a target monitoring area of ​​a geostationary orbit satellite, and use an atmospheric chemistry online coupling model to simulate the atmosphere of the target monitoring area to obtain a real atmosphere; The target observation time corresponding to the target monitoring area is determined, and satellite sampling of the real atmosphere is performed based on the target observation time to obtain the spatiotemporal distribution data of the ozone column content.

3. The ozone simulation method based on multi-source data assimilation according to claim 2 is characterized in that: The ozone column content data of the geostationary orbit satellite is obtained by inversion according to the spatiotemporal distribution data, wherein the ozone column content data includes ozone column content distribution, observation value, inversion error, prior value and average kernel function matrix, including: Inputting the spatiotemporal distribution data into a preset radiation transfer model to obtain a real atmospheric ozone column content profile, wherein the real atmospheric ozone column content profile reflects an observed value; The climatological ozone column content profile and the real atmospheric ozone column content profile are input into a preset inversion model to invert the ozone column content data of the geostationary orbit satellite, wherein the climatological ozone column content profile reflects a priori values.

4. The ozone simulation method based on multi-source data assimilation according to claim 3 is characterized in that: The step of inputting the climatological ozone column content profile and the real atmospheric ozone column content profile into a preset inversion model to invert and obtain the ozone column content data of the geostationary orbit satellite includes: The priori value and the observed value are input into the inversion model for inversion calculation to obtain the ozone column content data, wherein the inversion calculation process is: p =x s +A(x ap -x s )+ε,x′ p is the inversion profile of the pth layer surface, x s is the observed value, ε is the inversion error, x ap is the prior value, and A is the average kernel function matrix.

5. The ozone simulation method based on multi-source data assimilation according to claim 4 is characterized in that: The ground monitoring data, the low-orbit satellite data and the ozone column content data are input into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and ozone simulation is performed according to the multi-source assimilated data, including: Determining an observation error of the observation data based on the ground monitoring data, the low-orbit satellite data and the ozone column content data; Inputting the priori error of the forecast field and the observation error into a preset multi-source data assimilation system for data assimilation to obtain a model error; An optimal estimate is determined according to the observation error and the model error, multi-source assimilation data is obtained based on the optimal estimate, and ozone simulation is performed based on the multi-source assimilation data to obtain an ozone concentration analysis field.

6. The ozone simulation method based on multi-source data assimilation according to claim 5, characterized in that: The determining of the optimal estimation value according to the observation error and the model error comprises: Construct observation error covariance matrix and model error covariance matrix; Determining a Kalman gain according to the observation error covariance matrix and the pattern error covariance matrix; An optimal estimation value is determined based on the observation error, the model error and the Kalman gain.

7. The ozone simulation method based on multi-source data assimilation according to claim 5, characterized in that: After performing ozone simulation according to the multi-source assimilation data to obtain an ozone concentration analysis field, the method includes: determining a frequency of the data assimilation, and determining a first assimilation time and a second assimilation time based on the frequency; The forecast field at the second assimilation time is updated to the ozone concentration analysis field at the first assimilation time.

8. An ozone simulation device based on multi-source data assimilation, characterized in that: The device comprises: A geostationary orbit satellite data acquisition module is used to obtain the temporal and spatial distribution data of the ozone column content of the geostationary orbit satellite; An ozone column content data inversion module is used to invert the ozone column content data of the geostationary orbit satellite according to the spatiotemporal distribution data, wherein the ozone column content data includes the distribution of ozone column content, observation value, inversion error, prior value and average kernel function matrix; The multi-source data assimilation and ozone simulation module is used to input ground monitoring data, low-orbit satellite data and the ozone column content data into a preset multi-source data assimilation system for data assimilation to obtain multi-source assimilated data, and perform ozone simulation based on the multi-source assimilated data.

9. A terminal, characterized in that: The terminal includes a memory, a processor, and an ozone simulation program based on multi-source data assimilation stored in the memory and executable on the processor. When the processor executes the ozone simulation program based on multi-source data assimilation, the steps of the ozone simulation method based on multi-source data assimilation as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an ozone simulation program based on multi-source data assimilation. When the ozone simulation program based on multi-source data assimilation is executed by a processor, the steps of the ozone simulation method based on multi-source data assimilation as described in any one of claims 1 to 7 are implemented.

Images