A method and system for forecasting and assessing atmospheric fine particulate matter pollution processes

Abstract

This invention belongs to the field of ambient air quality forecasting technology, and relates to a method and system for forecasting and evaluating atmospheric fine particulate matter pollution processes. The method includes: 1) data acquisition and processing; 2) calculating level accuracy, pollution hit rate, trend consistency rate, and peak accuracy; 3) calculating a comprehensive score of the forecast effect based on level accuracy, pollution hit rate, trend consistency rate, and peak accuracy; 4) analyzing all PM2.5 concentrations within the target time period in the city. 2.5 The forecast effectiveness of each pollution process is calculated into a comprehensive score, and these scores are categorized to allow for horizontal comparison of the forecast effectiveness of different pollution processes. By constructing a comprehensive scoring method based on multi-dimensional evaluation indicators, it can achieve the assessment of PM2.5 levels. 2.5 A quantitative assessment of the overall forecasting effectiveness of pollution processes, filling the gap in PM2.5 prediction. 2.5 There is a lack of comprehensive evaluation methods for the effectiveness of pollution process forecasting.

Description

Technical Field

[0001] This invention belongs to the field of ambient air quality forecasting technology, and relates to a method and system for forecasting and assessing pollution processes, particularly a method and system for forecasting and assessing atmospheric fine particulate matter pollution processes. Background Technology

[0002] Affected by unfavorable meteorological conditions, fine particulate matter (PM2.5) in the atmosphere during autumn and winter... 2.5 Frequent pollution. Taking a typical city as an example, PM2.5 levels... 2.5 Pollution episodes often exhibit characteristics such as multiple consecutive days, escalating pollution levels, and high peak concentrations. Air quality forecasting is a crucial tool for supporting effective responses to heavy pollution weather and helping to "shorten and reduce the peak" of heavy pollution episodes. The simulation results of air quality forecasting models are an important reference for forecasting and judging pollution episodes. (Autumn and winter PM2.5 concentrations) 2.5 The effectiveness of pollution forecasting directly impacts the precision of pollution control measures and their implementation. Therefore, a systematic assessment of PM2.5 is necessary. 2.5 The effectiveness of pollution process forecasts, especially the forecasting effectiveness of air quality forecasting models, should be evaluated to assess their ability to support pollution process control, analyze the causes of deviations, and promote continuous improvement in forecast quality.

[0003] Currently, for air quality forecasting models PM 2.5 Forecast assessments of single pollutant concentrations typically employ internationally used statistical indicators such as standardized mean deviation (NMB), root mean square error (RMSE), and correlation coefficient (r), as proposed in the "Technical Specification for Numerical Forecasting of Ambient Air Quality" (HJ 1130-2020). These indicators primarily evaluate the degree of deviation and trend consistency between forecast and measured values. Alternatively, classification test indicators can be used, with the exceedance limit as the boundary, to separately count the number of days where the forecast and measured values ​​exceed the limit, thereby assessing the hit rate, false alarm rate, and false negative rate of polluted days. Meanwhile, numerous studies also evaluate PM2.5 concentration models according to the "Technical Regulations for Ambient Air Quality Index (AQI) (Trial)" (HJ 633—2012). 2.5 The accuracy rate of IAQI level forecasts or IAQI range forecasts corresponding to the concentration. However, the above assessment method has the following main problems:

[0004] (1) The statistical indicators are out of sync with the management needs.

[0005] Domestic and international scientific research fields often use statistical indicators to evaluate the effectiveness of model forecasts, which can reflect the degree of deviation and trend correlation between forecast and measured values. However, environmental management departments consider more the degree of consistency between forecast and measured values ​​at different air quality levels, especially for PM2.5. 2.5In forecasting pollution events, the focus is more on whether polluted days and their corresponding pollution levels are predicted. Therefore, relying solely on statistical indicators is insufficient to meet the needs of environmental management departments in understanding the effectiveness of pollution forecasts.

[0006] (2) The difference in forecasting capabilities for different levels of pollution is not adequately reflected.

[0007] The classification test indicators mainly assess the ability of forecast results to determine whether or not a polluted day will occur, such as PM2.5. 2.5 The concentration exceeding the limit is the boundary, which affects PM2.5. 2.5 Treating light, moderate, heavy, and severe pollution as the same "polluted day" makes it difficult to reflect the differences in forecasting capabilities for different levels of pollution events. While the accuracy of level-based forecasts can be used to assess the forecasting effectiveness for different levels of pollution days within a specific time period, this is equivalent to breaking down and classifying pollution days of different levels for multiple pollution events within a specific time period, and cannot reflect the forecasting effectiveness for different levels of pollution days within the same pollution event.

[0008] (3) There is a lack of methods for evaluating the overall forecasting effect of pollution processes.

[0009] For a single typical PM 2.5 In evaluating the effectiveness of pollution process forecasts, research institutions and management departments often use the accuracy rate of IAQI level forecasts. However, the accuracy rate of level forecasts is calculated based on a single city and a single natural day. It is only a basic assessment of the consistency of daily air quality levels during a pollution process and cannot reflect the PM2.5 levels over multiple consecutive days of pollution. 2.5 Even grasping the trend of concentration fluctuations cannot accurately assess PM2.5 levels. 2.5 The predictive effectiveness of peak concentration levels.

[0010] Therefore, there is currently a lack of comprehensive evaluation methods that can take into account multiple aspects of the forecast effect, such as the duration, trend, and peak level of the pollution process. Summary of the Invention

[0011] To solve the current PM 2.5 The current invention addresses the problem that relying on a single indicator for evaluating the effectiveness of pollution process forecasts makes it difficult to comprehensively assess the overall forecasting effectiveness. It proposes a method and system for forecasting and evaluating atmospheric fine particulate matter (PM2.5) pollution processes. By constructing a comprehensive scoring method based on multi-dimensional evaluation indicators, it can achieve a comprehensive assessment of PM2.5 levels. 2.5 A quantitative assessment of the overall forecasting effectiveness of pollution processes, filling the gap in PM2.5 prediction. 2.5 There is a lack of comprehensive evaluation methods for the effectiveness of pollution process forecasting.

[0012] To achieve the above objectives, the present invention provides the following technical solution:

[0013] A method for forecasting and assessing atmospheric fine particulate matter pollution processes, characterized by comprising the following steps:

[0014] 1) The actual PM levels measured in the city 2.5 Light or higher pollution levels lasting for two consecutive days or more are considered one PM event. 2.5 The pollution process involves marking all PM2.5 particles within the target time period. 2.5 The pollution process, and the processing of each PM2.5. 2.5 PM2.5 concentrations on the day before the pollution event begins, on each day during the event, and on the day following the event's conclusion. 2.5 Measured and forecasted concentration values, IAQI measured and forecasted levels, and PM2.5 concentrations based on both measured and forecasted values. 2.5 Hourly concentration data, statistical analysis of PM2.5 concentrations for each period. 2.5 PM during pollution processes 2.5 Measured and predicted peak concentrations;

[0015] 2) For each PM 2.5 For each pollution process, the accuracy rate of the pollution level forecast, the pollution hit rate, the trend consistency rate, and the peak accuracy are calculated. If the IAQI forecast level matches the actual measured level on a given day, the forecast is considered accurate. Each PM2.5 level is then recorded as an example of accurate forecasting. 2.5 The percentage of days with accurate PM2.5 level forecasts during a pollution episode is [percentage missing]. 2.5 The percentage of total days of a pollution episode is used as the level accuracy; the pollution hit rate refers to PM2.5. 2.5 Accurately predict the percentage of polluted days out of the actual number of polluted days during the pollution process; for each PM... 2.5 The pollution event was divided into groups: the day before the start of the pollution event, each day during the event, and the day after the event ended. Two days prior to and following each group were considered as one group, and the PM2.5 concentrations for each group were compared between the two days prior to and following each group. 2.5 The consistency between the rise and fall of predicted and measured concentration values ​​yields the trend consistency rate; based on PM... 2.5 PM during pollution process 2.5 The peak concentration accuracy is obtained by comparing the predicted and measured values.

[0016] 3) For each PM 2.5 For pollution processes, a comprehensive score for forecast effectiveness is calculated based on level accuracy, pollution hit rate, trend consistency rate, and peak accuracy.

[0017] 4) For all PM2.5 concentrations within the target time period in the city 2.5 The comprehensive score of the forecast effect is calculated for each pollution process, and the comprehensive score of the forecast effect is divided into different levels to make a horizontal comparison of the forecast effects of different pollution processes.

[0018] Preferably, in step 2), the formula for calculating the level accuracy is as follows:

[0019]

[0020] In the formula, A G Here, n represents the level accuracy, and n represents PM. 2.5 The number of days with accurate level forecasts during a pollution episode, where N represents the PM2.5 concentration during that episode. 2.5 The total number of days of the pollution process.

[0021] Preferably, in step 2), the formula for calculating the contamination hit rate is as follows:

[0022]

[0023] In the formula, D POD A represents the pollution hit rate, where A is the PM2.5 concentration. 2.5 The number of correct forecasts for polluted days during the pollution process, where C represents PM2.5. 2.5 The number of days that were predicted not to be polluted during the pollution process but actually occurred.

[0024] Preferably, in step 2), the formula for calculating the trend consistency rate is as follows:

[0025]

[0026] In the formula, R CT For trend consistency rate, I i0 and I fi0 The PM of the previous day in group i are respectively 2.5 Measured and predicted concentration values, I i and I fi P represents the measured and forecast values ​​of PM2.5 concentration for the following day in group i, respectively. i Let N be the trend change forecast score for group i, and N be the PM value for that period. 2.5 The total number of groups in the pollution process.

[0027] Preferably, in step 2), the formula for calculating peak accuracy is as follows:

[0028] ① 0≤|Δpeak|≤20, P P =100;

[0029] ② 20<∣Δpeak∣≤40, P P =80;

[0030] ③ 40<∣Δpeak∣≤60, P P =60;

[0031] ④ 60<∣Δpeak∣≤80, P P =30;

[0032] ⑤ 80<|Δpeak|,P P=0;

[0033] In the formula, P P For peak accuracy, Δpeak is PM 2.5 PM during pollution process 2.5 The difference between the predicted and measured peak concentration.

[0034] Preferably, in step 3), the comprehensive score of the forecast effect is ST, and its calculation formula is as follows:

[0035] S T =0.3(A) G ×100)+0.2(D POD ×100)+0.3(R CT (×100)+0.2P P .

[0036] Preferably, step 4) specifically involves: calculating the comprehensive forecast score ST for all PM2.5 pollution processes within the target time period in the city, and classifying the comprehensive forecast score ST into four levels: excellent, good, passable, and poor, according to the following rules, to facilitate horizontal comparison of the forecast effects of different pollution processes and support forecasters in analyzing the reasons for forecast deviations.

[0037] ①90≤S T ≤100, excellent;

[0038] ②75≤S T <90, good;

[0039] ③60≤S T <75, passing grade;

[0040] ④S T <60, poor.

[0041] Furthermore, the present invention also provides an atmospheric fine particulate matter pollution process forecasting and assessment system, characterized in that it includes:

[0042] The data acquisition and processing module is used to process the measured PM levels in the city. 2.5 Light or higher pollution levels lasting for two consecutive days or more are considered one PM event. 2.5 The pollution process involves marking all PM2.5 particles within the target time period. 2.5 The pollution process, and the processing of each PM2.5. 2.5 PM2.5 concentrations on the day before the pollution event begins, on each day during the event, and on the day following the event's conclusion. 2.5 Measured and forecasted concentration values, IAQI measured and forecasted levels, and PM2.5 concentrations based on both measured and forecasted values. 2.5 Hourly concentration data, statistical analysis of PM2.5 concentrations for each period. 2.5 PM during pollution processes 2.5Measured and predicted peak concentrations;

[0043] The single-process, multi-angle evaluation index calculation module is used to calculate the PM for each process. 2.5 For each pollution process, the accuracy rate of the pollution level forecast, the pollution hit rate, the trend consistency rate, and the peak accuracy are calculated. If the IAQI forecast level matches the actual measured level on a given day, the forecast is considered accurate. Each PM2.5 level is then recorded as an example of accurate forecasting. 2.5 The percentage of days with accurate PM2.5 level forecasts during a pollution episode is [percentage missing]. 2.5 The percentage of total days of a pollution episode is used as the level accuracy; the pollution hit rate refers to PM2.5. 2.5 Accurately predict the percentage of polluted days out of the actual number of polluted days during the pollution process; for each PM... 2.5 The pollution event was divided into groups: the day before the start of the pollution event, each day during the event, and the day after the event ended. Two days prior to and following each group were considered as one group, and the PM2.5 concentrations for each group were compared between the two days prior to and following each group. 2.5 The consistency between the rise and fall of predicted and measured concentration values ​​yields the trend consistency rate; based on PM... 2.5 PM during pollution process 2.5 The peak concentration accuracy is obtained by comparing the predicted and measured values.

[0044] The single-process comprehensive scoring calculation module is used to calculate the PM score for each process. 2.5 For pollution processes, a comprehensive score for forecast effectiveness is calculated based on level accuracy, pollution hit rate, trend consistency rate, and peak accuracy.

[0045] The multi-process forecast quality classification module is used to classify all PM2.5 levels within a city's target time period. 2.5 The comprehensive score of the forecast effect is calculated for each pollution process, and the comprehensive score of the forecast effect is divided into different levels to make a horizontal comparison of the forecast effects of different pollution processes.

[0046] Furthermore, the present invention also provides an atmospheric fine particulate matter pollution process forecasting and assessment device, characterized in that it includes:

[0047] One or more processors;

[0048] Memory, used to store one or more programs;

[0049] When the one or more programs are executed by the one or more processors, the one or more processors implement the atmospheric fine particulate matter pollution process forecasting and assessment method as described above.

[0050] Finally, the present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the program is executed by a processor, it implements the steps of the atmospheric fine particulate matter pollution process forecasting and assessment method as described above.

[0051] Compared with the prior art, the atmospheric fine particulate matter pollution process forecasting and assessment method and system of the present invention has one or more of the following beneficial technical effects:

[0052] 1. This invention constructs a PM that comprehensively considers multiple perspectives. 2.5 The evaluation method for the forecasting effectiveness of pollution processes can achieve a quantitative assessment of the overall forecasting effectiveness of pollution processes. It is applicable to the reliability evaluation of pollution process model forecasting results and can provide technical support for improving the forecasting capability of pollution processes.

[0053] 2. This invention is applicable to individual PM in cities. 2.5 The overall forecast effectiveness assessment of pollution processes sets forecast evaluation indicators to reflect the ability of forecast results to grasp four aspects of the pollution process: basic characteristics, duration, trend of change, and peak level. Then, different weight coefficients are set for the four evaluation indicators by assigning scores to each indicator. Finally, a comprehensive score of the forecast effectiveness of the pollution process is obtained. The comprehensive score can then be used to compare and analyze the performance of multiple pollution processes in a specific period. Attached Figure Description

[0054] Figure 1 This is a flowchart of the atmospheric fine particulate matter pollution process forecasting and assessment method of the present invention.

[0055] Figure 2 This is an exemplary PM of the present invention. 2.5 A schematic diagram of the pollution process.

[0056] Figure 3 This is a schematic diagram of the atmospheric fine particulate matter pollution process forecasting and assessment system of the present invention. Detailed Implementation

[0057] Before detailing any embodiment of the invention, it should be understood that the invention, in its application, is not limited to the details of the construction and arrangement of the components set forth in the following description or illustrated in the following figures. The invention can have other embodiments and can be practiced or carried out in various ways. Furthermore, it should be understood that the wording and terminology used herein are for descriptive purposes and should not be considered limiting. The use of “comprising” or “having” and variations thereof is intended to cover the items set forth below and their equivalents, as well as any additional items. Unless otherwise specified or limited, the terms “installation,” “connection,” “support,” and “linkage,” and variations thereof are used broadly and cover both direct and indirect installation, connection, support, and linking. Moreover, “connection” and “linkage” are not limited to physical or mechanical connections or links.

[0058] Furthermore, firstly, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention. Secondly, the term "a" should be understood as "at least one" or "one or more," that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple. The term "a" should not be construed as a limitation on the quantity.

[0059] Figure 1 A flowchart of the atmospheric fine particulate matter pollution process forecasting and assessment method of the present invention is shown. For example... Figure 1 As shown, the atmospheric fine particulate matter pollution process forecasting and assessment method of the present invention includes the following steps:

[0060] I. Data Acquisition and Processing.

[0061] In this invention, the PM2.5 measured in urban areas will be... 2.5 Light or higher pollution levels lasting for two consecutive days or more are considered one PM event. 2.5 Pollution process (e.g., light pollution on day 1, light pollution on day 2, moderate pollution on day 3, this three consecutive days is marked as one pollution process).

[0062] Based on real-time observation data, all PM2.5 concentrations within the target time period (e.g., 1 month or 3 months) are marked. 2.5 Pollution process, sorting out each PM 2.5 PM2.5 concentrations on the day before the pollution event begins, on each day during the event, and on the day following the event's conclusion. 2.5 The measured concentration values ​​and IAQI levels are then compiled and matched with the PM2.5 levels predicted for each day the previous day. 2.5 Forecast values ​​for PM2.5 concentrations and IAQI forecast levels, based on both measured and forecasted PM2.5 concentrations. 2.5 Hourly concentration data, statistical analysis of PM2.5 concentrations for each period. 2.5 PM during pollution processes 2.5 Measured and predicted values ​​of peak concentration.

[0063] II. Calculation of multi-dimensional evaluation indicators for a single process.

[0064] For each PM 2.5The pollution process is evaluated by calculating the results of four indicators: air quality level accuracy, pollution hit rate, trend consistency rate, and peak accuracy. These indicators reflect the forecast performance of the pollution process from different perspectives, including its basic characteristics, duration, trend, and peak level.

[0065] 1. Accuracy of air quality levels.

[0066] If the IAQI forecast level matches the actual IAQI level on a given day, the forecast is considered accurate. PM 2.5 The percentage of days with accurate PM2.5 level forecasts during a pollution episode is [percentage missing]. 2.5 The percentage of the total number of days in a pollution episode is the accuracy rate of the pollution level, and its calculation formula is as follows:

[0067]

[0068] In the formula: A G The accuracy rate of air quality levels is given by n, where n is the PM2.5 concentration at a given time. 2.5 The number of days with accurate level forecasts during a pollution episode, where N represents the PM2.5 concentration during that episode. 2.5 The total number of days of the pollution process.

[0069] 2. Contamination hit rate.

[0070] PM 2.5 The limit for light pollution concentration is the boundary, marking a certain PM2.5 concentration level. 2.5 The daily forecast and measured values ​​during the pollution process were divided into four categories based on whether they exceeded the light pollution limit. The number of occurrences for each of the four categories was counted, as shown in the table below:

[0071]

[0072] Pollution hit rate refers to the percentage of PM2.5 concentrations during a given PM2.5 concentration period. 2.5 The ratio of correctly predicted polluted days out of the actual polluted days during a pollution event can, to some extent, reflect the accuracy of the forecast in predicting the duration of the pollution event. The calculation formula is as follows:

[0073]

[0074] In the formula: D POD The pollution hit rate is represented by A, which is the number of correctly predicted polluted days; and C, which is the number of days that were predicted not to be polluted but actually occurred, i.e., the number of missed polluted days.

[0075] 3. Trend consistency rate.

[0076] Trend consistency rate is used to reflect the impact of forecast results on PM2.5. 2.5The degree of understanding of the overall evolution trend of a pollution process from "occurrence-development-dissipation". In this invention, to facilitate the calculation of trend consistency rate, each pollution process is grouped into three groups: the day before the occurrence, each day during the process, and the day after the end of the pollution process. The two days before and after each group form a single group, and the PM2.5 concentrations of the two days within each group are compared. 2.5 The consistency between predicted and measured concentration fluctuations is considered accurate if they rise, fall, or remain unchanged. The specific formula is as follows:

[0077]

[0078] In the formula: R CT The trend consistency rate, i.e., the rate at which predicted concentration changes and measured concentration changes move in the same direction; I i0 and I fi0 These are the measured and forecast values ​​of PM2.5 concentration for the previous day in group i, respectively; I i and I fi P represents the measured and forecast values ​​of PM2.5 concentration for the following day in group i, respectively; i The trend change forecast score for group i is given; N is the total number of groups, where group 1 is the day before the PM2.5 pollution process and the first day of the pollution process, reflecting the initial pollution accumulation; group 2 is the first day of the PM2.5 pollution process and the second day of the pollution process, reflecting the intermediate changes such as rapid accumulation, continuous aggravation, stable maintenance or slight improvement of pollution; ...; group N is the last day of the pollution process and the day after the end of the pollution process, reflecting the significant relief of pollution at the end of the pollution process.

[0079] 4. Peak accuracy.

[0080] Peak accuracy is used to assess how well the forecast results grasp the peak PM2.5 concentration during a pollution process. Specifically, it involves comparing the peak PM2.5 concentration during the pollution process. 2.5 The highest measured concentration of each hour and the highest predicted concentration are compared. The absolute value of the difference between the two is calculated, and scores are assigned according to the range. PM 2.5 The difference between the predicted peak concentration and the measured peak concentration is denoted as Δpeak, and the peak accuracy is denoted as P. P Based on the magnitude of the absolute value of Δpeak, for P P Scoring is based on tiers, and the scoring rules are as follows:

[0081] ① 0≤|Δpeak|≤20, P P =100;

[0082] ② 20<∣Δpeak∣≤40, P P =80;

[0083] ③ 40<∣Δpeak∣≤60, P P =60;

[0084] ④ 60<∣Δpeak∣≤80, P P =30;

[0085] ⑤ 80<|Δpeak|,P P =0.

[0086] III. Calculation of Comprehensive Score for a Single Process

[0087] Based on the evaluation results of the four indicators in step two, weights are assigned to each indicator according to the degree of concern, and the comprehensive score S of the pollution process forecast effectiveness is calculated. T The rating is expressed on a 100-point scale and is calculated using the following formula:

[0088] S T =0.3(A) G ×100)+0.2(D POD ×100)+0.3(R CT (×100)+0.2P P .

[0089] IV. Quality Classification of Multi-Process Forecasts

[0090] For all PM during the target time period in the city 2.5 The comprehensive score S of the prediction effect of each pollution process was calculated. T S is treated according to the following rules T The grades are divided into four levels: Excellent, Good, Pass, and Poor.

[0091] ①90≤S T ≤100, excellent;

[0092] ②75≤S T <90, good;

[0093] ③60≤S T <75, passing grade;

[0094] ④S T <60, poor.

[0095] This allows for a horizontal comparison of the forecasting effects of different pollution processes, supporting forecasters in analyzing the causes of forecast deviations and thus correcting them in a targeted manner to continuously improve forecast quality.

[0096] The following is an example of a three-day PM2.5 concentration in a certain city. 2.5 Taking a pollution process as an example, this paper details how to use the atmospheric fine particulate matter pollution process forecasting and assessment method of the present invention to evaluate the forecasting effect.

[0097] This PM 2.5 During the pollution process, PM 2.5The forecast and measured values ​​of daily average and hourly values ​​are shown in the table below. Figure 2 As shown. Among them, PM 2.5 The measured and predicted peak concentrations were 202 μg / m³. 3 and 182 μg / m 3 .

[0098]

[0099] Based on the above table and Figure 2 Based on the data, the evaluation indicators were calculated as follows:

[0100] Air quality level accuracy A G =2 / 3 × 100% = 66.7% (During the pollution process, the forecast level and the actual measured level matched for two days);

[0101] Pollution hit rate D POD =2 / 3 × 100% = 66.7% (two days of pollution were correctly predicted during the pollution process);

[0102] Trend Consistency R CT =3 / 4×100=75% (two days before and after are grouped together, with a total of four groups, of which three groups show the same trend between the forecast and the actual change).

[0103] Peak accuracy P P =100 (The absolute value of the difference between the predicted and measured peak concentrations is 20 μg / m³) 3 The corresponding P P =100).

[0104] Therefore, the overall forecasting effectiveness of the pollution process is:

[0105] S T =0.3×66.7%×100+0.2×66.7%×100+0.3×75%×100+0.2×100=75.9, corresponding to a good level.

[0106] Figure 3 A schematic diagram of the atmospheric fine particulate matter pollution process forecasting and assessment system of the present invention is shown. Figure 3 As shown, the atmospheric fine particulate matter pollution process forecasting and assessment system of the present invention includes:

[0107] 1. Data Acquisition and Processing Module.

[0108] The data acquisition and processing module is used to process the measured PM levels in the city. 2.5 Light or higher pollution levels lasting for two consecutive days or more are considered one PM event. 2.5 The pollution process involves marking all PM2.5 particles within the target time period. 2.5 The pollution process, and the processing of each PM2.5.2.5 PM2.5 concentrations on the day before the pollution event begins, on each day during the event, and on the day following the event's conclusion. 2.5 Measured and forecasted concentration values, IAQI measured and forecasted levels, and PM2.5 concentrations based on both measured and forecasted values. 2.5 Hourly concentration data, statistical analysis of PM2.5 concentrations for each period. 2.5 PM during pollution processes 2.5 Measured and predicted values ​​of peak concentration.

[0109] 2. Single-process multi-angle evaluation index calculation module.

[0110] The single-process multi-angle evaluation index calculation module is used to calculate each PM. 2.5 For each pollution process, the accuracy rate of the pollution level forecast, the pollution hit rate, the trend consistency rate, and the peak accuracy are calculated. If the IAQI forecast level matches the actual measured level on a given day, the forecast is considered accurate. Each PM2.5 level is then recorded as an example of accurate forecasting. 2.5 The percentage of days with accurate PM2.5 level forecasts during a pollution episode is [percentage missing]. 2.5 The percentage of total days of a pollution episode is used as the level accuracy; the pollution hit rate refers to PM2.5. 2.5 Accurately predict the percentage of polluted days out of the actual number of polluted days during the pollution process; for each PM... 2.5 The pollution event was divided into groups: the day before the start of the pollution event, each day during the event, and the day after the event ended. Two days prior to and following each group were considered as one group, and the PM2.5 concentrations for each group were compared between the two days prior to and following each group. 2.5 The consistency between the rise and fall of predicted and measured concentration values ​​yields the trend consistency rate; based on PM... 2.5 PM during pollution process 2.5 The peak concentration accuracy is obtained by calculating the difference between the predicted and measured values.

[0111] 3. Single-process comprehensive score calculation module.

[0112] The single-process comprehensive scoring calculation module is used to calculate the PM score for each process. 2.5 For pollution processes, a comprehensive score is calculated based on level accuracy, pollution hit rate, trend consistency rate, and peak accuracy to determine the forecast effectiveness.

[0113] 4. Multi-process forecast quality classification module.

[0114] The multi-process forecast quality classification module is used to classify all PM2.5 levels within the target time period for the city. 2.5 The comprehensive score of the forecast effect is calculated for each pollution process, and the comprehensive score of the forecast effect is divided into different levels to make a horizontal comparison of the forecast effects of different pollution processes.

[0115] Furthermore, the present invention also provides an atmospheric fine particulate matter pollution process forecasting and assessment device, which includes: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the atmospheric fine particulate matter pollution process forecasting and assessment method as described above.

[0116] Finally, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the atmospheric fine particulate matter pollution process forecasting and assessment method as described above.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Those skilled in the art can modify or make equivalent substitutions to the technical solutions of the present invention based on the concept of the present invention, without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for atmospheric fine particulate pollution process forecast assessment, characterized in that, The method comprises the following steps: 1) The city measured PM 2.5 pollution and above for two consecutive days and above, as 1 PM 2.5 pollution process, all PM in the target period are marked one by one 2.5 pollution process, and arrange each PM 2.5 pollution process before the day, each day during the process and the day after the process 2.5 concentration measured value and forecast value, IAQI measured level and forecast level, based on measured and forecast PM 2.5 concentration hourly data, statistics of each PM 2.5 PM of pollution process 2.5 measured value and forecast value of peak concentration; 2) For each PM 2.5 For each pollution process, the accuracy rate of the pollution level forecast, the pollution hit rate, the trend consistency rate, and the peak accuracy are calculated. If the IAQI forecast level matches the actual measured level on a given day, the forecast is considered accurate. Each PM2.5 level is then recorded as an example of accurate forecasting. 2.5 The percentage of days with accurate PM2.5 level forecasts during a pollution episode is [percentage missing]. 2.5 The percentage of total days of a pollution episode is used as the level accuracy; the pollution hit rate refers to PM2.

5. 2.5 Accurately predict the percentage of polluted days out of the actual number of polluted days during the pollution process; for each PM... 2.5 The pollution event was divided into groups: the day before the start of the pollution event, each day during the event, and the day after the event ended. Two days prior to and following each group were considered as one group, and the PM2.5 concentrations for each group were compared between the two days prior to and following each group. 2.5 The consistency between the rise and fall of predicted and measured concentration values ​​yields the trend consistency rate; based on PM... 2.5 PM during pollution process 2.5 The peak concentration accuracy is obtained by comparing the predicted and measured values. 3) For each PM 2.5 pollution episode, the overall score of the prediction is calculated based on the level accuracy, pollution hit rate, trend consistency and peak accuracy. 4) All PM 2.5 The overall scores of the prediction effects of the pollution processes are calculated respectively, and the overall scores of the prediction effects are graded to compare the prediction effects of different pollution processes horizontally.

2. The method of claim 1, wherein the method further comprises: In the step 2), the calculation formula of the level accuracy is as follows: In the formula, A G is the level accuracy, n is the PM 2.5 the number of days with accurate level prediction during the pollution process, N is the PM 2.5 total number of days of the pollution process.

3. The method of claim 2, wherein the method further comprises: In the step 2), the calculation formula of the pollution hit rate is as follows: where D POD is the pollution hit rate, A is the PM 2.5 is the number of correct predictions of pollution days during pollution episodes, C is the PM 2.5 is the number of pollution days predicted not to occur but which actually did occur during pollution episodes.

4. The method of claim 3, wherein the method further comprises: In the step 2), the calculation formula of the trend consistency rate is as follows: where R CT is the consistency rate of trend, I i0 and I fi0 are the measured and predicted values of PM 2.5 concentration of the previous day in the i-th group, I i and I fi are the measured and predicted values of PM2.5concentration of the following day in the i-th group, P i is the predicted score of trend change in the i-th group, and N is the total number of groups in the PM 2.5 pollution process.

5. The method of claim 4, wherein the method further comprises: In the step 2), the calculation formula of the peak precision is as follows: P = 100;​ P = 80;​ P = 60;​ (4) 60 < |Δpeak| < 80, P P = 30; 5 80 < | Δpeak |, P P = 0; where P P is the peak precision, Δpeakis the PM 2.5 PM 2.5 the difference between the predicted and measured values of the peak concentration.

6. The method of atmospheric fine particulate pollution process forecast evaluation of claim 5, wherein, In step 3), the comprehensive score of the predicted effect is S T The calculation formula is as follows: S T = 0.3(A G x 100) + 0.2(D POD x 100) + 0.3(R CT x 100) + 0.2P P .

7. The method of atmospheric fine particulate pollution process forecast evaluation according to any one of claims 1-6, characterized in that, The step 4) is specifically: calculating the comprehensive score S of the prediction effect of all PM 2.5 pollution processes in the urban target period T The comprehensive score S of the prediction effect is graded according to the following rules: excellent, good, pass, and poor, to compare the prediction effects of different pollution processes horizontally, and support the forecasters to analyze the reasons for the prediction deviation: T ​ T ≤ 100, excellent;​ ii) 75 < S < 90, good; and T <90, good; T <75, pass;​ (4) S T <60, poor.

8. An atmospheric fine particulate pollution process forecast evaluation system characterized by, The method comprises the following steps: Data acquisition and arrangement module for the city real-time PM 2.5 Light and above pollution and two consecutive days and above, considered 1 PM 2.5 Pollution process, mark all the PM in the target period one by one 2.5 Pollution process, and arrange each PM 2.5 Pollution process before the beginning of the day, each day and the end of the next day of PM 2.5 Concentration of real-time and forecast values, IAQI real-time and forecast levels, based on real-time and forecast PM 2.5 Concentration of hourly data, statistics of each PM 2.5 PM of pollution process 2.5 The measured value and the forecast value of the peak concentration; a single-process multi-angle evaluation index calculation module, which is used for calculating the level accuracy, pollution hit rate, trend consistency rate and peak precision for each PM 2.5 pollution process, wherein if the IAQI prediction level of a day is consistent with the measured level, it is recorded as level prediction accuracy, the number of days of level prediction accuracy in each PM 2.5 pollution process accounts for the total number of days of the PM 2.5 pollution process, the percentage of the total number of days as the level accuracy; the pollution hit rate refers to the ratio of the number of days of the PM 2.5 pollution process to the actual number of pollution days; the number of days of the PM 2.5 pollution process, the number of days of the PM 2.5 pollution process, the number of days of the PM 2.5 pollution process, the number of days of the PM 2.5 pollution process, the number of days of the PM a single-process comprehensive score calculation module for calculating a comprehensive score of the prediction effect of each PM 2.5 pollution process based on the level accuracy, pollution hit rate, trend consistency rate and peak accuracy. a multi-process forecast quality grading module for grading all PM 2.5 The integrated score of the forecast effect of each pollution process is calculated and graded to compare the forecast effects of different pollution processes horizontally.

9. An atmospheric fine particulate pollution process prediction evaluation device, characterized by, The method comprises the following steps: One or more processors; Memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the atmospheric fine particulate pollution process prediction evaluation method according to any one of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the steps of the atmospheric fine particulate pollution process prediction evaluation method according to any one of claims 1-7.

Images