CN113156395B - Aerosol laser radar data fusion method and system - Google Patents
Aerosol laser radar data fusion method and system Download PDFInfo
- Publication number
- CN113156395B CN113156395B CN202110376079.8A CN202110376079A CN113156395B CN 113156395 B CN113156395 B CN 113156395B CN 202110376079 A CN202110376079 A CN 202110376079A CN 113156395 B CN113156395 B CN 113156395B
- Authority
- CN
- China
- Prior art keywords
- data
- groups
- aerosol
- subsystem
- laser radar
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000000443 aerosol Substances 0.000 title claims abstract description 63
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 51
- 230000004927 fusion Effects 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims description 32
- 238000000034 method Methods 0.000 claims description 29
- 238000009792 diffusion process Methods 0.000 claims description 20
- 239000000126 substance Substances 0.000 claims description 19
- 239000012071 phase Substances 0.000 claims description 14
- 238000001914 filtration Methods 0.000 claims description 12
- 238000004088 simulation Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 11
- 238000003908 quality control method Methods 0.000 claims description 10
- 230000005484 gravity Effects 0.000 claims description 8
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000009826 distribution Methods 0.000 claims description 7
- 239000007791 liquid phase Substances 0.000 claims description 7
- 230000004807 localization Effects 0.000 claims description 7
- 238000004891 communication Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000013076 uncertainty analysis Methods 0.000 claims description 5
- 230000008878 coupling Effects 0.000 claims description 4
- 238000010168 coupling process Methods 0.000 claims description 4
- 238000005859 coupling reaction Methods 0.000 claims description 4
- 238000007726 management method Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000012937 correction Methods 0.000 abstract description 2
- 230000008033 biological extinction Effects 0.000 description 22
- 238000006243 chemical reaction Methods 0.000 description 21
- 239000011159 matrix material Substances 0.000 description 16
- 239000013618 particulate matter Substances 0.000 description 13
- 239000007789 gas Substances 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 8
- 239000003344 environmental pollutant Substances 0.000 description 8
- 239000002245 particle Substances 0.000 description 8
- 231100000719 pollutant Toxicity 0.000 description 8
- 238000011160 research Methods 0.000 description 8
- 238000004062 sedimentation Methods 0.000 description 8
- 238000012795 verification Methods 0.000 description 7
- 239000000203 mixture Substances 0.000 description 5
- 241000894007 species Species 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000002776 aggregation Effects 0.000 description 3
- 238000004220 aggregation Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 3
- 239000000356 contaminant Substances 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- MWUXSHHQAYIFBG-UHFFFAOYSA-N nitrogen oxide Inorganic materials O=[N] MWUXSHHQAYIFBG-UHFFFAOYSA-N 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- RRHGJUQNOFWUDK-UHFFFAOYSA-N Isoprene Chemical compound CC(=C)C=C RRHGJUQNOFWUDK-UHFFFAOYSA-N 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 239000012798 spherical particle Substances 0.000 description 2
- 238000012876 topography Methods 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000001856 aerosol method Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 150000001335 aliphatic alkanes Chemical class 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 150000004945 aromatic hydrocarbons Chemical class 0.000 description 1
- 239000000926 atmospheric chemistry Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000001311 chemical methods and process Methods 0.000 description 1
- 239000013626 chemical specie Substances 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000004090 dissolution Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 238000007716 flux method Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 239000013067 intermediate product Substances 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003094 perturbing effect Effects 0.000 description 1
- 238000006552 photochemical reaction Methods 0.000 description 1
- 239000002243 precursor Substances 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 230000002000 scavenging effect Effects 0.000 description 1
- 238000009991 scouring Methods 0.000 description 1
- 238000001179 sorption measurement Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XTQHKBHJIVJGKJ-UHFFFAOYSA-N sulfur monoxide Chemical class S=O XTQHKBHJIVJGKJ-UHFFFAOYSA-N 0.000 description 1
- 229910052815 sulfur oxide Inorganic materials 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000005436 troposphere Substances 0.000 description 1
- 238000004800 variational method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/48—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00
- G01S7/4802—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S17/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Radar Systems Or Details Thereof (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The invention discloses an aerosol laser radar data fusion method, which comprises the following steps: s1, taking n groups of mode initial values as input values of the nested grid air quality prediction mode subsystems respectively, and then obtaining n groups of prediction field results; s2, taking n groups of forecast field results and radar observation data as input values of a parallel data assimilation subsystem, and then obtaining n groups of analysis field results; and S3, taking the n groups of analysis field results as n groups of mode initial values of the next moment to carry out forecast and data fusion of the next moment. Correspondingly, the invention also discloses an aerosol laser radar data fusion system. The invention can perform fusion correction on the aerosol laser radar data according to different conditions, thereby obtaining more accurate laser radar data with pertinence.
Description
Technical Field
The invention relates to the technical field of aerosol laser radars, in particular to an aerosol laser radar data fusion method and system.
Background
Data assimilation combines external field observation with a numerical mode according to uncertainty to obtain more accurate state variables and corresponding uncertainty (Lahoz et al, 2010), which is a general term based on data fusion methods such as variational and kalman filtering. The data fusion research of the aerosol laser radar in China just started. Zheng billows (2018) developed a modal-based second layer PM 2.5 An observation operator of mass concentration and extinction coefficient proportion firstly constructs a ground-based laser radar data fusion system based on NAQPMS (the New Air Quality prediction Model System) by taking OI (optimal interpolation) as a fusion algorithm, and assimilates 12 ground-based laser radar joint observations in 11, 10 and 11, 20 days in 2014. The results show that the assimilation improvement observed with the addition of lidar was significant (RMSE improvement was about 24%), as well as a longer duration of mode effect. The term derivation (2018) utilizes 3D-Var algorithm of GSI (Gridpoint Statistical interference), and combines 12 ground-based laser radars and 160-minute conjugate walkthrough laser radar observation from 29 days to 5 days in 11 months in 2017 based on WRF-chem (weather Research and formulation Model with the chemistry) through the fitting relation of aerosol mass concentration and extinction coefficient. The average value of the correlation of the extinction coefficient fitting to the mass concentration reaches 0.86, which shows the feasibility of the constructed observation operator. The assimilation of the lidar resulted in an average increase in correlation of 42.9% and an average reduction in RMSE of 50.9%. In addition, the study also clarified PM at the early stage of this heavy pollution process, at the time of pollution and at the later stage of pollution 2.5 And (3) evolution process of the three-dimensional mass concentration field. Cheng et al (2019a) uses the CRTM (the Community radiated Transfer model) as an observation operator, and combines WRF-Chem and four sets of Foundation laser radar observations of 12:00-18:00 in 13.3.13.2018 in Beijing to obtain an improved aerosol initial field. The result shows that the assimilation laser radar extinction coefficient vertical profile can obviously improve PM 2.5 Prediction of underestimation, RMSE drop 25 μ g/m 3 The correlation coefficient increased by 0.45. Cheng et al (2019b) fused the extinction coefficient vertical profile of CALIPA in MODES and the AOD product of MODIS in the year 2016 for 11 months using the 4-D LETKF (a four-Dimensional Local energy Transform Kalman Filter) algorithm with the Atmospheric chemical Transport mode NICAM-SPRINTARS (Non-hydraulic Icosahedral atomic Model-Spectral Radiation Transport Model for Aerosol specs). The results show that assimilation of both calep and MODIS can improve the model-to-aerosol simulation. Assimilation of CALIP was superior to MODIS for incorrect aerosol vertical distribution, and the model improvement on mode AOT was superior. Because of the relatively sparse observation space-time range of CALIP, assimilation CALIP has little effect on mode AOT simulation. Liang et al (2020) developed a set of 3D-Var data assimilation systems for the aerosol extinction coefficient observation operator and the aerosol mass concentration operator. The research applies an assimilation system to 5 foundation laser radars fusing WRF-Chem and 2018, 11, 13 and obtain better PM 2.5 And forecasting an initial value field. PM (particulate matter) 2.5 The error of the forecast initial value field is reduced by 10.5 mu g/m 3 (17.6%) and the forecast aging can last for more than 24 hours. Ma et al (2020) use WRF-Chem/DART (data optimization Research testbed) atmospheric chemical data Assimilation system to fuse ground PM 2.5 Observation, AOD and ground-based laser radar extinction coefficients study two pollution cases lasting for 7 days in 6 months and 11 months in 2018, and the study finds that negative increment occurs on the fusion result by assimilating the AOD and the aerosol extinction coefficient vertical profile. This is mainly due to the fact that WRF-Chem simulates lower humidity in the boundary layer, resulting in lower simulated extinction coefficient, and simulated ground PM 2.5 The two higher deviation trends are not consistent. This study reduced the problem of the variation tendency inconsistency by correction of variation, and made the problem to be simultaneously assimilatedFlour PM 2.5 The results of AOD and extinction coefficient vertical profile are optimal.
Lidar data fusion has three major sources of uncertainty: atmospheric chemical transmission mode, laser radar observation and fusion algorithm. The sources of uncertainty in atmospheric chemical transmission modes are mainly meteorological fields, aerosol initial values, emission sources, mode parameterization schemes, observation operators, and the like (Ma et al, 2019). As a bridge connecting a mode variable and an observation variable, a mode observation operator is an important factor influencing the data fusion of the aerosol laser radar. In current research, a single method is mostly adopted as an observation operator, such as rice scattering under spherical particles (Yumimoto et al, 2007), a ratio method (zheng billows, 2018), an observation-based approach of impulse Monitoring of Protected Visual environments (Liang et al, 2020), a fitting method (item derivation, 2018), and a simplified radiation transmission mode method (Cheng et al, 2019 a). However, since aerosols have different extinction performances in different particle size spectra, different compositions, different complex refractive indexes and different environmental humidities, it is necessary to consider the way of aggregation to deal with different aerosols, for example, Mie theory for spherical particles, TMM (T-Matrix Method) or igom (improved Geometric Optics methods) Method for ellipsoidal particles, and dda (diffraction Dipole application) Method for irregular particles (gastiger and Wiegner, 2018). The fusion algorithm is mainly divided into a variational method and an ensemble Kalman filtering method. The method is characterized in that a complex nonlinear observation operator is needed for the simulation observation of the integrated Kalman filtering energy processing mode, the flow dependence characteristic of a mode background field can be reflected, and the method is more suitable for laser radar data fusion research (Houtekamer and Zhang, 2016).
Disclosure of Invention
The invention aims to provide an aerosol laser radar data fusion method and system to obtain more accurate state variables and corresponding uncertainty of aerosol laser radar data.
In order to solve the technical problems, the technical scheme of the invention is as follows:
an aerosol laser radar data fusion method comprises the following steps:
s1, taking n groups of mode initial values as input values of the nested grid air quality prediction mode subsystems respectively, and then obtaining n groups of prediction field results;
s2, taking n groups of forecast field results and radar observation data as input values of a parallel data assimilation subsystem, and then obtaining n groups of analysis field results;
and S3, taking the n groups of analysis field results as n groups of mode initial values of the next moment to carry out forecast and data fusion of the next moment.
Preferably, the nested grid air quality prediction mode subsystem comprises one or more of a source disturbance analysis processing module, an advection transport analysis processing module, a turbulence diffusion analysis processing module, a liquid phase and heterogeneous phase analysis processing module, a gravity settling and dry settling analysis processing module, a gas phase chemical analysis processing module and a wet cleaning analysis processing module; and the n groups of mode initial values are converted into n groups of forecast field results after passing through one or any plurality of modules in the nested grid air quality forecast mode subsystem respectively.
Preferably, the parallel data assimilation subsystem comprises one or more of an assembly Kalman filtering module, a horizontal and vertical localization module, a vertical diffusion module, a REMOVE observation operator module, an MPI secondary parallel module and an online coupling module; and the n groups of forecast field results and radar data are converted into n groups of analysis field results after passing through one or more modules in the parallel data assimilation subsystem.
Preferably, the obtaining of radar observation data comprises the following steps:
s100, scanning aerosol distribution to obtain initial data of the foundation laser radar;
s200, performing quality control management on the polarity data of the initial data of the foundation laser radar to obtain quality control data of the foundation laser radar;
s300, respectively carrying out normalization processing and virtual observation on the quality control data of the foundation laser radar;
and S400, carrying out uncertainty analysis on the normalized data and the data after virtual observation so as to obtain radar observation data in the step S2.
Preferably, the advection transport analysis processing module adopts a high-precision positive fixed mass conservation difference format scheme for calculation.
Preferably, the turbulent diffusion analysis processing module adopts a gradient delivery theory.
Preferably, the dry sedimentation process in the gravity sedimentation and dry sedimentation analytical treatment module is based on a resistance model treatment, the gas part adopts a Wesely scheme, and the aerosol part adopts a Slinn scheme.
Preferably, the gas phase chemical mechanism in the gas phase chemical analysis processing module is CBM-Z.
The invention also discloses an aerosol laser radar data fusion system which comprises a nested grid air quality prediction mode subsystem and a data assimilation subsystem and is used for any one of the aerosol laser radar data fusion methods.
Compared with the prior art, the invention has the beneficial effects that:
the aerosol laser radar data is more accurate through the nested grid air quality prediction mode subsystem and the data assimilation subsystem; the nested grid air quality prediction mode subsystem corrects aerosol laser radar data by adopting a method comprising source disturbance, advection transportation, turbulent diffusion, liquid phase and heterogeneous phase, gravity sedimentation and dry sedimentation, gas phase chemistry and wet removal; the parallel data assimilation subsystem further IMPROVEs the accuracy of aerosol laser radar data by adopting a method including ensemble Kalman filtering, horizontal and vertical localization, vertical diffusion, IMPROVE observation operator, MPI two-stage parallel and online coupling nested grid air quality prediction mode subsystem output data, such as: the aerosol laser radar data fusion can obviously improve the simulation of extinction coefficient, and the mode overestimation and the mode underestimation under specific conditions can be corrected through the laser radar data fusion.
Drawings
FIG. 1 is a flow chart of the NAQPMS-PDAF data fusion system operation of the present invention;
FIG. 2 is a detailed diagram of the operation flow of the NAQPMS-PDAF data fusion system of the present invention;
FIG. 3 is a diagram of the application of the NAQPMS-PDAF data fusion system of the present invention;
FIG. 4 is a two-stage parallel schematic of the NAQPMS and PDAF of the present invention;
FIG. 5 is a graph of a prior RMSE and a prior total travel time series at 300m in accordance with the present invention;
FIG. 6 is a graph of prior RMSE and prior total travel time sequence at 470m in accordance with the present invention;
FIG. 7 is a prior RMSE and a prior total propagation time series plot at 680m in accordance with the present invention;
FIG. 8 is a graph of prior RMSE and prior total propagation time series at 930m according to the present invention;
FIG. 9 is a comparison graph of AOD verification results of Beijing PKU stations of AOD and AERONET converted by extinction coefficients before and after data fusion of the laser radar of the present invention;
FIG. 10 is a graph comparing AOD verification results of Xianghe station of AOD and AERONET converted by extinction coefficient before and after the laser radar data fusion;
FIG. 11 is a graph comparing extinction coefficients before and after the lidar data fusion of 12 days 4 months 4 years 2019 with CALIP verification results;
FIG. 12 is a graph comparing extinction coefficients before and after the lidar data fusion of 4, 13 and 2019 with CALIP verification results.
Detailed Description
The following further describes embodiments of the present invention with reference to the drawings. It should be noted that the description of the embodiments is provided to help understanding of the present invention, and is not intended to limit the present invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Example one
As shown in fig. 1, an aerosol lidar data fusion method includes the following steps: s1, taking n groups of mode initial values as input values of the nested grid air quality prediction mode subsystems respectively, and then obtaining n groups of prediction field results, so that an ensemble prediction result at the current moment can be obtained; s2, taking n groups of forecast field results and radar observation data as input values of a parallel data assimilation subsystem, and then obtaining n groups of analysis field results, so that a set analysis field at the current moment is obtained; and S3, taking the n groups of analysis field results as n groups of mode initial values of the next moment to carry out forecast and data fusion of the next moment.
In the embodiment, the 1 st group of mode initial values are used as input values of the 1 st nested grid air quality prediction mode subsystem, and then the 1 st group of prediction field results are obtained; taking the 1 st group of forecast field results and radar observation data as input values of a parallel data assimilation subsystem, and then obtaining 1 st group of analysis field results; the same principle applies to the initial values of the 2 nd group mode to the n nth group mode, and the results of the 2 nd group analysis field to the n nth group analysis field can be obtained respectively.
As shown in fig. 2, the nested grid air quality prediction mode subsystem includes one or any more of a source disturbance analysis processing module, an advection transport analysis processing module, a turbulence diffusion analysis processing module, a liquid phase and heterogeneous phase analysis processing module, a gravity settling and dry settling analysis processing module, a gas phase chemical analysis processing module, and a wet cleaning analysis processing module; and the n groups of mode initial values are converted into n groups of prediction field results after passing through one or any plurality of modules in the nested grid air quality prediction mode subsystem respectively.
In this embodiment, the nested grid Air quality Prediction mode subsystem is a two-way nested three-dimensional euler atmospheric chemical transmission mode, which is referred to as naqpms (nested Air quality Prediction Model system) for short. NAQPMS includes physical and chemical processes of pollutant discharge, advection transport, turbulent diffusion, wet and dry sedimentation, gas phase, liquid phase, and heterogeneous reactions. The terrain following coordinate is used as a vertical coordinate, the horizontal structure is nested in multiple grids, and sand dust and PM in regional and urban dimensions can be simulated simultaneously 2.5 、PM 10 、SO 2 、NO x 、CO、O 3 、NH 3 And the like.
The general form of the chemical composition euler transport equation is:
the Euler transport equation of the three-dimensional chemical composition under the height terrain following spherical coordinate system is as follows:
wherein C is i Is the concentration of the i-th pollutant, t is the time, theta,
Latitude and longitude, R is the radius of the earth, K θ 、
And K σ Respectively the diffusion coefficients of the longitudinal turbulence, the latitudinal turbulence and the vertical turbulence, u and v are horizontal wind speed, P is a chemical conversion term, S is the source discharge source rate, R is the source discharge source rate d For dry sedimentation term, W ash For the wet clearance item, topography has certain influence to pollutant transport, and sigma is the topography coordinate, satisfies:
wherein
For the height of the top of the pattern,
z is the elevation of the grid point, which is the terrain height. In the terrain following coordinate system, the lower horizontal plane is a terrain plane, which is very convenient for discussing the distribution of pollutants. W is the equivalent vertical motion velocity, and is obtained by an air mass conservation equation:
the advection conveying process in the NAQPMS adopts a high-precision positive constant mass conservation differential format scheme for calculation. The scheme is a modified and simplified version based on a flux method, can ensure the concentration of chemical species to be positive and the mass conservation, reduces the problems of calculating pseudo diffusion and the like, and is widely applied to a troposphere chemical mode. Turbulent diffusion adopts a gradient transport theory, namely the atmospheric diffusion problem is treated into a problem of solving a diffusion equation under a certain boundary condition. In NAQPMS, the horizontal turbulent diffusion coefficient takes a constant value, since horizontal diffusion is much smaller than the contribution of horizontal advection. The vertical diffusion was calculated using the Byun and Dennis protocol, which has been widely used in tropospheric chemistry models such as CMAQ and RAQM. The dry settling process in NAQPMS is based on a drag model treatment, with the Wesely protocol for the gas portion and Slinn et al for the aerosol portion. Dry sedimentation flux F ═ V d C, wherein V d For dry settling rate, C is aerosol mass concentration. In the drag model, the gas dry settling rate V d Is calculated as follows:
wherein r is a Representing aerodynamics, characterizing the migration of contaminants to the ground due to turbulent diffusion; r is b Represents the sub-boundary layer resistance, which characterizes the direct arrival at the ground through the thin layer of air due to molecular diffusion; r is c The surface resistance is mainly characterized by the capture and adsorption of pollutants by buildings, vegetation and the like. Gravity settling is also a concern for aerosol particles. In addition, it is generally assumed that the particles can adhere to the surface, i.e. the surface resistance r c When the value is 0, the calculation formula is as follows:
wherein V s The gravity settling rate is primarily related to the size and density of the particles. The wet cleaning module in NAQPMS mainly comprises subclouds (below 1600 m)) A scavenging process, which has been successfully used for east Asia NO x Wet clean simulation of (d). The process of under-cloud rain washout is defined as:
W ash ＝W a ·C i
wherein, W ash For the amount of scouring, C i Is the concentration of the contaminant, W a The wash-out factor is related to rain intensity, rain drop spectral distribution, concentration of various contaminants, aerosol particle spectrum, and aerosol chemical composition. In NAQPMS, cloud convection, liquid phase chemistry, and wet removal process under the cloud in the cloud are processed based on RADM2 mechanism, which has been widely applied to numerical simulation of atmospheric chemistry. The method comprises the steps of utilizing cloud water and rainwater data provided by a WRF mode and meteorological elements in grids, considering liquid phase processes of cloud and rainwater absorption, dissolution, ionization, chemical reaction and the like of pollutants, clearing in cloud and flushing under cloud under rainfall conditions, diagnosing whether accumulated cloud convection exists or not, and vertically redistributing the pollutants. The gas phase chemistry mechanism of NAQPMS is CBM-Z. CBM-Z is a new generalised chemical mechanism based on the structural classification developed for CBM-IV. The chemical reaction equation is increased from 81 to 176, and the reaction species is increased from 32 to 67, wherein the species involved in photochemical reaction is 20. Compared with the CBM-IV mechanism, CBM-Z has more comprehensive consideration in the aspects of chemical reaction of active long-life species and intermediate products, chemical reaction of inorganic substances, chemical reaction of active alkanes, paraffins and aromatic hydrocarbons, chemical reaction of a plurality of free radicals and isoprene, and the like. Meanwhile, four reaction scenes including background conditions, cities, suburbs, biological areas, oceans and the like are set in the mechanism, and the mechanism is applicable to research on the scale of the whole world, the region and the city. According to different reaction scenes, species participating in the reaction and reaction equations are different. Under background atmosphere, there are 74 chemical reactions; in urban atmosphere, the number of chemical reactions is 118; in urban atmosphere with biological sources, the number of chemical reactions is increased to 134; the chemical reaction is 109 in the air above the sea far away from the man-made source; in the offshore, the number of chemical reactions increased to 153; in offshore locations affected by biological sources, the chemical reactions reach 169. The aerosol process considered in NAQPMS mainly comprises inorganic aerosol thermodynamic process and secondary organic aerosolGel generation, heterogeneous chemical reactions on the surface of the aerosol, etc. Calculation of constituents of inorganic aerosols using the thermodynamic model of aerosol IsororopiA in NAQPMS
Equal to the thermodynamic equilibrium and distribution of the precursors. The heterogeneous chemical reaction module on the surface of the aerosol mainly focuses on the influence on sulfur oxides and nitrogen oxides. Under certain temperature and humidity, water vapor is condensed and grown on the surfaces of aerosol particles to form a thin water layer, and gaseous pollutants are diffused to the surfaces of the aerosol and can be absorbed and dissolved by surface solution of the aerosol.
With continued reference to fig. 2, the parallel data assimilation subsystem includes one or more of an ensemble kalman filtering module, a horizontal and vertical localization module, a vertical diffusion module, a REMOVE observation operator module, an MPI secondary parallel module, and an online coupling module; and the n groups of forecast field results and radar data are converted into n groups of analysis field results after passing through one or more modules in the parallel data assimilation subsystem.
In order to realize the fusion of mode Data, ground observation and space observation, an LESTKF (local Error Subspace ensemble transform Kalman Filter) method under PDAF (parallel Data optimization framework) is used for Data fusion, and an NAQPMS-PDAF online Assimilation system is built. The PDAF is free open source software aiming at simplifying the assimilation of the deployment aggregation, and provides a unified framework for implementing parallel aggregation filtering and smoothing, such as LETKF, NETF, ESTKF and the like. The PDAF is written based on Fortran and can be operated in parallel through MPI and OpenMP.
ESTKF as an EnKF using a set of N e Set of state variables of a dimension
To characterize the state variable x of the mode system k Sum error covariance matrix P k . Dimension N of state variables x Including mode variables and parameters. It is composed ofCharacterized by ensemble averaging:
the error covariance matrix is:
wherein,
is a set perturbation matrix.
The algorithm for ensemble assimilation is divided into a forecasting step and an analysis step, which are denoted by subscripts f and a, respectively. The forecasting step is to set the state variable and error covariance matrix of the mode system from t to t k-1 Propagation to the next observation time t ═ t k . During mode operation, numerical mode (M) k ) Propagation of state variables and error covariance for each set member:
at t ═ t k Size is N y Is observed vector
By passing
And truth system x f And establishing contact. Wherein an observation error ε is assumed k Obey covariance matrix R k H is an observation operator that projects the pattern space into the observation space.
The analyzing step provides an estimate of the new state by fusing the forecast field and the observation information. Since the analysis steps are performed at the same time, the time index is omitted from the following description. The ESTBF uses Kalman filtering to update equations to calculate analysis field state variables and an error covariance matrix from uncertainty information of a prediction field, an observation field and both:
wherein,
k is Kalman gain:
K＝P f H T (HP f H T +R) -1
for a set:
Y f ＝HX f
wherein,
is an aggregate matrix of the prediction fields,
a matrix of observation sets corresponding to the projection of the mode field into the observation space.
The dimension of the state variable is large, and the size is N e The overall approximated sample covariance matrix of (a) is only a low-order approximation of the true covariance matrix with a rank of at most N e -1. ESTKF uses this property to write an analysis field to the collection of tensed N e In a 1-dimensional subspace, this space is called the error subspace. The prediction error covariance matrix P is then f Is written as:
wherein,
comprises the following steps:
L＝X f Ω
matrix array
Is defined as:
will gather matrix X f Projected into the error subspace. After this, P a Can be written as:
P a ＝LAL T
wherein the transformation matrix
Is composed of
A -1 ＝ρ(N e -1)I+(Y f Ω) T R -1 YΩ
Wherein rho epsilon [0, 1] is a forgetting factor. Finally, the updating step converts the forecast set into an analysis set:
wherein,
in order to forecast the ensemble averaging matrix,
each column of (a) is:
w is:
where C is the symmetric square root of A obtained by singular value decomposition. Finally, the set X is analyzed a As the next aggregate preAnd the initial values are reported and are continued in a rolling mode until the assimilation window is finished.
As shown in fig. 3, the obtaining of radar observation data includes the following steps: s100, scanning aerosol distribution to obtain initial data of the foundation laser radar; s200, performing quality control management on the polarity data of the initial data of the foundation laser radar to obtain quality control data of the foundation laser radar; s300, respectively carrying out normalization processing and virtual observation on the quality control data of the foundation laser radar; and S400, carrying out uncertainty analysis on the normalized data and the data after virtual observation so as to obtain radar observation data in the step S2.
In this embodiment, the radar observation data may be processed before being input to the PDAF, and the processing method includes performing quality control management, normalization, virtual observation, and uncertainty analysis on the initial data of the ground-based laser radar. The aerosol laser radar data fusion system outputs a set analysis field after being coupled on line by PDAF and NAQPMS, and the more accurate aerosol laser radar data can be applied to the following technical fields: data fusion system uncertainty analysis and long-time PM analysis 2.5 Vertical structure change characterization and evaluation of NAQPMS vs PM 2.5 Simulation/prediction improves. The invention can evaluate whether the processing of the initial data of the foundation laser radar is appropriate or not and whether the adjustment is needed or not according to the effect of the fusion application. Meanwhile, the invention can also improve the scheme of NAQPMS according to the effect of the fusion application.
As shown in fig. 4, the NAQPMS mode employs a "one-level parallelism" scheme, that is, a communication domain World is set to implement MPI distributed computation. As shown in fig. 1, in order to couple NAQPMS and PDAF online, the communication domain World is split into three sub-communication domains: model, filter and couple, which respectively function as model set simulation, set Kalman filtering and data exchange. In addition, in order to fuse vertical observation, the original two-dimensional data frame in the PDAF is expanded, the vertical dimension is increased, and the vertical localization radius parameter of the observed data is increased to facilitate subsequent research.
Based on the observation of 7 micro-pulse laser radars in 2019, 4, 10, 07, 00-4, 14, 10, 00, NAQPMS-PDAF data are obtainedAnd testing the fusion system. 7 laser radars are respectively arranged on Beijing, Baoding (2), Shijiazhuang, Cangzhou, Hengshui and the chenchen platform. There are a total of four paths of calip observation during this period. The space-time range of this study contained 2 effective AERONET sites: beijing _ PKU station and Xianghe station, and also contains 72 PMs 2.5 And a state control monitoring station. The fusion algorithm of the present study uses the leskf, the localization radius is set to 40 lattice points (about 200km), and the localization weight function is a 5-th polynomial. Generating 20 sets of sets, SO, by perturbing the initial emission source 2 And the emission source uncertainty for the VOC and the remaining species were set at 20%, 50% and 30%, respectively (Ma et al, 2019; Zhang et al, 2006). The observation error of the extinction coefficient of the aerosol laser radar is set as 0.02km -1 。
In the ensemble Kalman filtering algorithm, an ensemble is used for characterizing the evolution characteristics of a mode background field, and prior RMSE and prior total propagation are used for verifying whether the generated ensemble propagation is sufficient. The prior RMSE is formulated as
A priori total propagation of
(Houtekamer et al, 2009). Wherein,
an observed value representing the ith observation,
representing the average of a prior set of observation spaces,
a method of indicating an observation error is shown,
representing the observation space prior observation error variance. 5-8, the prior RMSE and the prior total propagation at heights of 300m, 470m, 680m, and 930m are substantially balanced in magnitude and trend of change, tabulationsThe clear set simulation propagation is more sufficient.
According to the results of an analysis field and a forecast field of the Kyoto Ji 7 laser radars data fusion, the RMSE of the extinction coefficients of the 7 laser radars is 0.25km after the data fusion -1 Reduced to 0.11km -1 BIAS from 0.13km -1 Reduced to 0.05km -1 MAE from 0.17km -1 Reduced to 0.09km -1 However, the correlation coefficient is increased to 0.81 from 0.28, and BIAS is also concentrated more in the vicinity of 0. Meanwhile, the RMSE of a prediction field with an extinction coefficient of 1 hour is from 0.25km -1 Down to 0.18km -1 BIAS from 0.13km -1 Reduced to 0.09km -1 MAE (mean absolute error) from 0.17km -1 Down to 0.14km -1 The correlation coefficient was changed from 0.28 to 0.41. It can be seen that the analysis field results are better than the 1 hour forecast field results. Internal verification shows that the aerosol laser radar data fusion can obviously improve the simulation of the extinction coefficient. Fig. 9 and 10 are AOD verification results of Beijing _ PKU station and Xianghe station of AOD and AERONET converted by extinction coefficients before and after the lidar data fusion, respectively. As shown in fig. 9, after lidar data fusion, although a deviation from the observation occurred at 18:00, 12 d 4, overall, the AOD time series was closer to the observation, RMSE decreased by 0.14, and the correlation coefficient increased by 0.1. Fig. 10 shows that AOD based on extinction coefficient scaling does not change much (RMSE is reduced by 0.02) from AOD to AERONET after lidar data fusion, since AOD simulation and observation at this site are closer compared to Beijing _ PKU site in NAQPMS mode. FIGS. 11 and 12 are verification graphs of laser radar data fusion and CALIPE results before and after 4/month 12/day and 4/month 13/day, respectively. As shown, the mode overestimation at about 1000-.
Example two
An aerosol laser radar data fusion system comprises a nested grid air quality prediction mode subsystem and a data assimilation subsystem, and is used for implementing the aerosol laser radar data fusion method in any one of the embodiments.
It should be understood that any system or device that can be used to implement the aerosol lidar data fusion method of any of the above embodiments of the present invention using the basic principles of the present invention is considered to be within the scope of the present invention.
The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. It will be apparent to those skilled in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, and the scope of protection is still within the scope of the invention.
Claims (7)
1. An aerosol laser radar data fusion method is characterized by comprising the following steps:
s1, taking n groups of mode initial values as input values of the nested grid air quality prediction mode subsystems respectively, and then obtaining n groups of prediction field results;
s2, taking n groups of forecast field results and radar observation data as input values of a parallel data assimilation subsystem, and then obtaining n groups of analysis field results;
s3, taking the n groups of analysis field results as n groups of mode initial values of the next moment to carry out prediction and data fusion of the next moment;
the parallel data assimilation subsystem comprises an assembly Kalman filtering module, a horizontal and vertical localization module, a vertical depth development module, a REMOVE observation operator module, an MPI secondary parallel module and an online coupling module; in the parallel data assimilation subsystem, expanding an original two-dimensional data frame in the parallel data assimilation subsystem, increasing vertical dimension, performing data fusion by using an LESTKF method, and building an online assimilation system of the nested grid air quality prediction mode subsystem and the parallel data assimilation subsystem; after the n groups of forecast field results and radar data pass through a plurality of modules in the parallel data assimilation subsystem respectively, the n groups of forecast field results and the radar data are converted into n groups of analysis field results;
the nested grid air quality prediction mode subsystem comprises one or more of a source disturbance analysis processing module, an advection transport analysis processing module, a turbulence diffusion analysis processing module, a liquid phase and heterogeneous phase analysis processing module, a gravity settling and dry settling analysis processing module, a gas phase chemical analysis processing module and a wet removal analysis processing module; n groups of mode initial values are converted into n groups of forecast field results after passing through one or any plurality of modules in the nested grid air quality forecast mode subsystem respectively;
the method comprises the following steps that a communication domain World is arranged in the nested grid air quality prediction mode subsystem to realize MPI distributed computation, and the communication domain World is divided into three sub-communication domains: model, filter and couple, which respectively function as model set simulation, set Kalman filtering and data exchange.
2. The aerosol lidar data fusion method of claim 1, wherein obtaining radar observation data comprises:
s100, scanning aerosol distribution to obtain initial data of the foundation laser radar;
s200, performing quality control management on the polarity data of the initial data of the foundation laser radar to obtain the quality control data of the foundation laser radar;
s300, respectively carrying out normalization processing and virtual observation on the quality control data of the foundation laser radar;
and S400, carrying out uncertainty analysis on the normalized data and the data after virtual observation so as to obtain radar observation data in the step S2.
3. The aerosol lidar data fusion method of claim 1, wherein: the advection transportation analysis processing module adopts a high-precision positive definite mass conservation difference format scheme for calculation.
4. The aerosol lidar data fusion method of claim 1, wherein: the turbulent diffusion analysis processing module adopts a gradient conveying theory.
5. The aerosol lidar data fusion method of claim 1, wherein: the dry settling process in the gravity settling and dry settling analytical treatment module is based on resistance model treatment, the gas part adopts Wesely scheme, and the aerosol part adopts Slinn scheme.
6. The aerosol lidar data fusion method of claim 1, wherein: the gas phase chemical mechanism in the gas phase chemical analysis processing module is CBM-Z.
7. An aerosol laser radar data fusion system comprises a nested grid air quality prediction mode subsystem and a data assimilation subsystem, and is characterized in that: for implementing the aerosol lidar data fusion method of any of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110376079.8A CN113156395B (en) | 2021-04-07 | 2021-04-07 | Aerosol laser radar data fusion method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110376079.8A CN113156395B (en) | 2021-04-07 | 2021-04-07 | Aerosol laser radar data fusion method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113156395A CN113156395A (en) | 2021-07-23 |
| CN113156395B true CN113156395B (en) | 2022-08-12 |
Family
ID=76889030
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110376079.8A Active CN113156395B (en) | 2021-04-07 | 2021-04-07 | Aerosol laser radar data fusion method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113156395B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114112995B (en) * | 2021-12-01 | 2024-01-30 | 中国人民解放军国防科技大学 | Aerosol optical characteristic data assimilation method and device based on three-dimensional variation technology |
| CN115479872A (en) * | 2022-09-28 | 2022-12-16 | 西北核技术研究所 | Device and method for measuring wet sedimentation removal coefficient of small-particle-size aerosol particles |
| CN118335245B (en) * | 2024-04-30 | 2024-09-13 | 中国科学院大气物理研究所 | System for on-line coupling atmospheric chemical transmission mode and data assimilation |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2953021B1 (en) * | 2009-11-26 | 2011-12-09 | Tanguy Griffon | METHOD FOR MEASURING WEEKLY AND ANNUAL EMISSIONS OF A GREENHOUSE GAS ON A DATA SURFACE |
| CN110186820A (en) * | 2018-12-19 | 2019-08-30 | 河北中科遥感信息技术有限公司 | Multisource data fusion and environomental pollution source and pollutant distribution analysis method |
| CN109709577A (en) * | 2018-12-28 | 2019-05-03 | 南京雨后地软环境技术有限公司 | A kind of Three-dimensional Variational Data Assimilation method of the aerosol LIDAR inverting PM2.5 based on WRF-Chem mode |
| CN111048161B (en) * | 2019-10-22 | 2022-09-13 | 中国人民解放军国防科技大学 | Aerosol extinction coefficient three-dimensional variation assimilation method based on IMPROVE equation |
| CN110910963B (en) * | 2019-10-29 | 2022-03-29 | 暨南大学 | Three-dimensional variation assimilation method and system for optical thickness of aerosol and storage medium |
| CN110989038B (en) * | 2019-11-22 | 2022-03-04 | 北京中科锐景科技有限公司 | Near-surface particle concentration estimation method and system |
| CN110909297B (en) * | 2019-11-22 | 2020-11-27 | 清华大学 | Numerical prediction set coupling assimilation system and method |
-
2021
- 2021-04-07 CN CN202110376079.8A patent/CN113156395B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113156395A (en) | 2021-07-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113156395B (en) | Aerosol laser radar data fusion method and system | |
| Vignati et al. | Global scale emission and distribution of sea-spray aerosol: Sea-salt and organic enrichment | |
| Satoh et al. | The non-hydrostatic icosahedral atmospheric model: Description and development | |
| Banks et al. | Impact of WRF model PBL schemes on air quality simulations over Catalonia, Spain | |
| Huebert et al. | An overview of ACE‐Asia: Strategies for quantifying the relationships between Asian aerosols and their climatic impacts | |
| Tang et al. | Improvement of ozone forecast over Beijing based on ensemble Kalman filter with simultaneous adjustment of initial conditions and emissions | |
| Sofiev | A model for the evaluation of long-term airborne pollution transport at regional and continental scales | |
| CN111487216B (en) | Carbon dioxide flux inversion method and system | |
| Govardhan et al. | Performance of WRF-Chem over Indian region: Comparison with measurements | |
| Lauer et al. | Simulating aerosol microphysics with the ECHAM/MADE GCM–Part I: Model description and comparison with observations | |
| Han et al. | MICS-Asia II: Model intercomparison and evaluation of ozone and relevant species | |
| Goto et al. | Application of a global nonhydrostatic model with a stretched-grid system to regional aerosol simulations around Japan | |
| CN109858686A (en) | A kind of ground emission inventories inverting optimization method based on EnKF | |
| CN111353634B (en) | Sulfur dioxide emission source optimization method based on mode initial field assimilation | |
| Sharipov et al. | Numerical modeling method for short-term air quality forecast in industrial regions | |
| Xiang et al. | Using Lidar technology to assess regional air pollution and improve estimates of PM2. 5 transport in the North China Plain | |
| Wen et al. | Source apportionment of PM 2.5 in Tangshan, China—Hybrid approaches for primary and secondary species apportionment | |
| Lazaridis et al. | Dynamics of fine particles and photo-oxidants in the Eastern Mediterranean (SUB-AERO) | |
| Yu et al. | A fast forecasting method for PM2. 5 concentrations based on footprint modeling and emission optimization | |
| CN116611547A (en) | Atmospheric pollutant source forecasting method, device, equipment and storage medium | |
| Wu et al. | Model evaluation and uncertainty analysis of PM2. 5 components over Pearl River Delta Region using Monte Carlo Simulations | |
| Venkatram et al. | The development of a model to examine source-receptor relationships for visibility on the Colorado Plateau | |
| Wang et al. | Development of four-dimensional variational assimilation system based on the GRAPES–CUACE adjoint model (GRAPES–CUACE-4D-Var V1. 0) and its application in emission inversion | |
| Fu et al. | MICS-Asia II: Modeling gaseous pollutants and evaluating an advanced modeling system over East Asia | |
| Wang et al. | A city-scale inverse modelling method for air pollutant source determination |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |