A method, system and related devices for predicting meteorological dispersion conditions

Abstract

The application discloses a meteorological diffusion condition prediction method, system and related device, which is used for predicting the meteorological diffusion condition in the extension period and prolonging the predictable duration of the meteorological diffusion condition. The method comprises the following steps: acquiring air quality monitoring data of a prediction area; determining an air quality index of the prediction area within a preset time according to the air quality monitoring data; acquiring ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time; analyzing the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and establishing a first prediction model according to the analysis result, wherein the first prediction model is used for predicting the air quality index in the extension period; and determining the meteorological diffusion condition grade in the extension period according to the prediction result of the first prediction model.

Description

TECHNICAL FIELD

[0001] The present application relates to the meteorological technical field, and in particular to a meteorological diffusion condition prediction method, system and related device. BACKGROUND

[0002] The rapid development of economy and society has led to a large amount of pollutant emissions, and air pollution has caused serious threats to human living environment and ecological health and safety. Meteorological diffusion conditions refer to meteorological factors that affect the diffusion and dilution of air pollutants. Studies have shown that under certain conditions of air pollution sources, pollutant concentration mainly depends on the diffusion capacity of the atmosphere, and meteorological conditions such as haze, calm, inversion, etc. will make the distribution of air pollutants more complex.

[0003] Meteorological diffusion conditions are usually evaluated and represented using diffusion indices or diffusion condition levels. These level divisions are usually based on comprehensive evaluation of meteorological factors such as wind speed, wind direction, atmospheric stability, and humidity. The higher the level of meteorological diffusion conditions, the worse the air quality, and the poorer the diffusion and dilution effect of pollutants; the lower the level of meteorological diffusion conditions, the better the air quality, and the better the diffusion and dilution effect of pollutants. By predicting meteorological diffusion conditions, we can better understand the spread and distribution of pollutants, and help develop more effective air quality management strategies. However, in the prior art, meteorological diffusion conditions are often predicted using subjective experience, and only 3-10 days of prediction can be achieved, and prediction for the extended period (11-30 days) is not possible, i.e. there is a certain limitation in the time range of meteorological diffusion condition prediction. SUMMARY

[0004] The present application provides a meteorological diffusion condition prediction method, system and related device for predicting meteorological diffusion conditions in the extended period, extending the forecast period of meteorological diffusion conditions.

[0005] The first aspect of the present application provides a meteorological diffusion condition prediction method, comprising:

[0006] obtaining air quality monitoring data of a prediction area;

[0007] determining an air quality index of the prediction area within a preset time according to the air quality monitoring data;

[0008] obtaining ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time;

[0009] analyze a correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and establish a first prediction model according to an analysis result, the first prediction model being used for predicting the air quality index in the extended period;

[0010] determine a meteorological diffusion condition level in the extended period according to a prediction result of the first prediction model.

[0011] Optionally, the analyzing the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and establishing the first prediction model according to the analysis result, comprises:

[0012] analyzing the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data to obtain a corresponding correlation coefficient;

[0013] filtering a target factor in the ground meteorological element data, the sounding data and the circulation characteristic data according to the correlation coefficient and a characteristic of a meteorological system in the prediction area;

[0014] establishing the first prediction model based on the air quality index and the target factor by a multivariate stepwise regression method.

[0015] Optionally, before the establishing the first prediction model based on the air quality index and the target factor by the multivariate stepwise regression method, the prediction method further comprises:

[0016] calculating a similarity coefficient between a current circulation field and a previous circulation field, and determining the previous circulation field with the highest similarity coefficient as a prediction field;

[0017] the establishing the first prediction model based on the air quality index and the target factor by the multivariate stepwise regression method comprises:

[0018] establishing the first prediction model based on the air quality index, the target factor and the prediction field by the multivariate stepwise regression method.

[0019] Optionally, the ground meteorological element data comprises air temperature, precipitation, wind speed, wind direction, relative humidity, visibility and sea level pressure, the sounding data comprises sounding data at 20:00 500hpa, 700hpa, 850hpa and 925hpa, and the circulation characteristic data comprises height field, temperature field, wind field and pressure field at 500hpa, 700hpa, 850hpa, 925hpa and sea level.

[0020] Optionally, after the air quality monitoring data of the prediction area is obtained, the prediction method further comprises:

[0021] determining the air pollution ratio of each month and each season according to the air quality monitoring data;

[0022] analyzing the correlation between the air pollution ratio of each month and each season and each of the circulation indices of the climate system circulation index, and establishing a second prediction model according to the analysis result, the second prediction model being used to predict the air pollution ratio of the monthly and seasonal scales;

[0023] determining the meteorological diffusion condition grade of the monthly and seasonal scales according to the prediction result of the second prediction model.

[0024] Optionally, the determining the air pollution ratio of each month and each season according to the air quality monitoring data comprises:

[0025] counting the number of typical persistent pollution weather processes of each month and each season according to the air quality monitoring data, the typical persistent pollution weather process being a pollution weather process with the air quality index reaching the light pollution or above and the duration being greater than or equal to 3 days;

[0026] determining the air pollution ratio of each month and each season according to the number of the typical persistent pollution weather processes of each month and each season.

[0027] Optionally, the analyzing the correlation between the air pollution ratio of each month and each season and each of the circulation indices of the climate system circulation index, and establishing a second prediction model according to the analysis result comprises:

[0028] analyzing the correlation between the air pollution ratio and each of the circulation indices of the climate system circulation index based on the monthly and seasonal scales, and screening out a target index affecting the air pollution ratio of each month and each season according to the analysis result;

[0029] establishing the second prediction model by the method of multivariate stepwise regression based on the air pollution ratio and the target index.

[0030] The second aspect of the present application provides a prediction system of meteorological diffusion condition, comprising:

[0031] a first acquisition unit configured to acquire air quality monitoring data of a prediction area;

[0032] a first determination unit configured to determine an air quality index of the prediction area within a preset time according to the air quality monitoring data;

[0033] a second obtaining unit, configured to obtain ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time;

[0034] a first analysis unit, configured to analyze a correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and establish a first prediction model according to an analysis result, the first prediction model being used to predict the air quality index in the extended period;

[0035] a first prediction unit, configured to determine a meteorological diffusion condition level in the extended period according to a prediction result of the first prediction model.

[0036] Optionally, the first analysis unit is specifically configured to:

[0037] analyze the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, to obtain a corresponding correlation coefficient;

[0038] filter target factors from the ground meteorological element data, the sounding data and the circulation characteristic data according to the correlation coefficient and characteristics of a meteorological system in the prediction area;

[0039] establish the first prediction model based on the air quality index and the target factors by using a multivariate stepwise regression method.

[0040] Optionally, the first analysis unit is specifically further configured to:

[0041] calculate a similarity coefficient between a current circulation field and a previous circulation field, and determine the previous circulation field with the highest similarity coefficient as a prediction field;

[0042] establish the first prediction model based on the air quality index, the target factors and the prediction field by using the multivariate stepwise regression method.

[0043] Optionally, the ground meteorological element data includes temperature, precipitation, wind speed, wind direction, relative humidity, visibility and sea level pressure, the sounding data includes sounding data at 20:00 500 hpa, 700 hpa, 850 hpa and 925 hpa, and the circulation characteristic data includes height fields, temperature fields, wind fields and pressure fields at 500 hpa, 700 hpa, 850 hpa, 925 hpa and sea level.

[0044] Optionally, the prediction system further includes:

[0045] a second determination unit, configured to determine an air pollution ratio of each month and each season according to the air quality monitoring data;

[0046] a second analysis unit configured to analyze the correlation between the air pollution ratio of each month or season and each of the circulation indices of the climate system circulation index, and to establish a second prediction model based on the analysis result, the second prediction model being configured to predict the air pollution ratio at the monthly or seasonal scale;

[0047] a second prediction unit configured to determine the meteorological diffusion condition level at the monthly or seasonal scale based on the prediction result of the second prediction model.

[0048] Optionally, the second determination unit is specifically configured to:

[0049] count the number of typical persistent pollution weather processes of each month or season based on the air quality monitoring data, the typical persistent pollution weather process being a pollution weather process in which the air quality index reaches the light pollution or above and the duration is greater than or equal to 3 days;

[0050] determine the air pollution ratio of each month or season based on the number of typical persistent pollution weather processes of each month or season.

[0051] Optionally, the second analysis unit is specifically configured to:

[0052] analyze the correlation between the air pollution ratio and each of the target indices at the monthly or seasonal scale, and to screen out target indices affecting the air pollution ratio of each month or season based on the analysis result;

[0053] establish the second prediction model based on the air pollution ratio and the target indices by using the multivariate stepwise regression method.

[0054] The third aspect of the present application provides a device for predicting meteorological diffusion conditions, the device comprising:

[0055] a processor, a memory, an input / output unit and a bus;

[0056] the processor being connected with the memory, the input / output unit and the bus;

[0057] the memory storing a program, and the processor invoking the program to execute the first aspect and any optional method for predicting meteorological diffusion conditions in the first aspect.

[0058] The fourth aspect of the present application provides a computer readable storage medium, the computer readable storage medium storing a program, the program being executed on a computer to execute the first aspect and any optional method for predicting meteorological diffusion conditions in the first aspect.

[0059] As can be seen from the above technical solutions, the present application has the following advantages:

[0060] By studying the correlation between the air quality index and the previous, current meteorological elements, and the circulation field, a first prediction model for predicting the air quality index in the extended period is constructed, and the air quality index predicted by the first prediction model is converted into the corresponding meteorological diffusion condition grade, thereby realizing the prediction of the meteorological diffusion condition in the extended period (11-30 days), prolonging the predictable time limit of the meteorological diffusion condition, and effectively improving the decision-making meteorological service capability of the environmental meteorology. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0062] Figure 1 An embodiment flowchart of the meteorological diffusion condition prediction method provided by the present application is shown in the figure.

[0063] Figure 2 Another embodiment flowchart of the meteorological diffusion condition prediction method provided by the present application is shown in the figure.

[0064] Figure 3 An embodiment flowchart of the meteorological diffusion condition prediction method provided by the present application is shown in the figure.

[0065] Figure 4 An embodiment structure diagram of the meteorological diffusion condition prediction system provided by the present application is shown in the figure.

[0066] Figure 5 An embodiment structure diagram of the meteorological diffusion condition prediction device provided by the present application is shown in the figure. DETAILED DESCRIPTION

[0067] The present application provides a meteorological diffusion condition prediction method, system and related device, which is used for predicting the meteorological diffusion condition in the extended period, and prolongs the predictable time limit of the meteorological diffusion condition.

[0068] It should be noted that the meteorological diffusion condition prediction method provided by the present application can be applied to a terminal, and can also be applied to a server, for example, the terminal can be a smart phone or a computer, a tablet computer, a smart television, a smart watch, a portable computer terminal, and a fixed terminal such as a desktop computer. For convenience of description, the system is taken as an example in the present application.

[0069] Please refer to Figure 1 , Figure 1An embodiment of the prediction method of meteorological dispersion conditions provided in the present application comprises:

[0070] 101. Obtain air quality monitoring data of the prediction area;

[0071] The air quality monitoring data is closely related to the meteorological dispersion conditions, which refers to the ability of the dispersion and dilution of pollutants in the atmosphere, and directly affects the formation and change of air quality. The level of meteorological dispersion conditions helps to more intuitively understand the relationship between air quality and meteorological conditions, and provides more understandable air quality information for the public.

[0072] The system first needs to determine the corresponding city or region according to the prediction range and research purposes, that is, to determine the prediction area, and then to obtain the air quality monitoring data of the prediction area within a certain time, which includes air quality index AQI, pollutant concentration and other related data.

[0073] 102. Determine the air quality index of the prediction area within a preset time according to the air quality monitoring data;

[0074] The air quality index AQI is a comprehensive evaluation of air quality, which is usually calculated from multiple pollutant concentrations. In the same pollutant concentration, the good or bad of meteorological dispersion conditions will directly affect the level of AQI. In the case of good meteorological dispersion conditions, pollutants can quickly disperse and dilute, resulting in lower AQI and better air quality; while in the case of poor meteorological dispersion conditions, pollutants are easy to accumulate, resulting in higher AQI and deteriorating air quality.

[0075] In order to analyze the meteorological dispersion conditions of the prediction area in the past period of time, the system determines the air quality index of the prediction area within a preset time according to the air quality monitoring data. To realize the prediction of meteorological dispersion conditions in the extended period, the preset time needs to be greater than the time of the extended period (11-30 days), and the preferred preset time is 45 days, that is, the system needs to determine the air quality index of the prediction area within 45 days according to the air quality monitoring data.

[0076] 103. Obtain ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time;

[0077] Meteorological dispersion conditions are directly affected by ground meteorological elements, sounding data and circulation characteristics, etc. The changes of these meteorological elements directly affect the transmission and dispersion process of atmospheric pollutants, thus having an important impact on air quality. Therefore, in the prediction of meteorological dispersion conditions, these meteorological elements need to be considered comprehensively to more accurately evaluate and predict meteorological dispersion conditions.

[0078] Specifically, ground meteorological element data such as temperature, humidity, wind speed, and wind direction directly affect the stability and mobility of the atmosphere. Higher temperature and lower humidity make the atmosphere more stable, inhibit the rise and diffusion of pollutants, and result in poor meteorological diffusion conditions. Larger wind speed and moderate humidity are conducive to the diffusion and dilution of pollutants, promote air quality improvement, and result in good meteorological diffusion conditions; sounding data provide meteorological element information of the atmospheric vertical profile. Through sounding data, atmospheric stratification and stability, and other meteorological characteristics can be analyzed to determine meteorological diffusion conditions. For example, the inversion phenomenon causes the temperature to gradually increase with height, hinders the rise and diffusion of pollutants, and results in poor meteorological diffusion conditions. When the atmospheric stratification is weak or unstable, pollutants are easy to rise and diffuse, and the meteorological diffusion condition is good; circulation characteristic data provide information of atmospheric circulation elements such as high pressure, low pressure, and wind field. These circulation characteristics affect the horizontal movement and wind direction of the atmosphere, thereby affecting the transport and diffusion of pollutants. For example, a stable high-pressure system causes calm weather, hinders the diffusion of pollutants, and results in poor meteorological diffusion conditions. A low-pressure system is usually accompanied by strong wind speed and air flow, which is conducive to the diffusion of pollutants, and the meteorological diffusion condition is good.

[0079] In short, there is a close relationship between ground meteorological element data, sounding data, and circulation characteristic data and meteorological diffusion conditions. The system needs to obtain ground meteorological element data, sounding data, and circulation characteristic data of the prediction area within a preset time and make further analysis and prediction.

[0080] 104. Analyze the correlation between the air quality index and the ground meteorological element data, sounding data, and circulation characteristic data, and establish a first prediction model based on the analysis results, which is used to predict the air quality index in the extended period;

[0081] The system analyzes the correlation between the air quality index and the ground meteorological element data, sounding data, and circulation characteristic data within a preset time. Specifically, correlation analysis and other statistical methods can be used to measure the linear correlation between them. Based on the results of the correlation analysis, the system selects the ground meteorological elements, sounding data, and circulation characteristics that are significantly correlated with the air quality index as the input features of the first prediction model, and the air quality index as the output feature of the first prediction model, so as to obtain the first prediction model for predicting the air quality index in the extended period.

[0082] 105. Determine the meteorological diffusion condition level in the extended period according to the prediction results of the first prediction model.

[0083] The system can predict the air quality index in the extension period through the first prediction model, and then convert the air quality index into different levels of meteorological diffusion conditions according to the level division, so as to obtain the prediction result of the meteorological diffusion level in the extension period, and realize the prediction of the meteorological diffusion condition in the extension period (11-30 days). In some specific embodiments, the division standards of the air quality index and the meteorological diffusion condition are as follows:

[0084] 0-50: excellent (excellent meteorological diffusion condition);

[0085] 51-100: good (good meteorological diffusion condition);

[0086] 101-150: light pollution (general meteorological diffusion condition);

[0087] 151-200: moderate pollution (poor meteorological diffusion condition);

[0088] 201-300: heavy pollution (poor meteorological diffusion condition);

[0089] 301 and above: serious pollution (very poor meteorological diffusion condition).

[0090] It should be noted that for different countries or regions, the level of meteorological diffusion condition may be different, and the actual division standard of meteorological diffusion condition may be different, which is not limited here.

[0091] In this embodiment, by studying the correlation between the air quality index and the meteorological elements of the previous period, the same period and the circulation field, a first prediction model for predicting the air quality index of the prediction area in the extension period is constructed, and the air quality index predicted by the first prediction model is converted into the corresponding meteorological diffusion condition level, thereby realizing the prediction of the meteorological diffusion condition in the extension period (11-30 days), prolonging the predictable time of the meteorological diffusion condition, and effectively improving the decision-making meteorological service ability of the environmental meteorology to the local.

[0092] The prediction method of the meteorological diffusion condition provided in the present application will be described in detail below. Please refer to Figure 2 , Figure 2 Another embodiment of the prediction method of the meteorological diffusion condition provided in the present application, the prediction method comprises:

[0093] 201, obtaining air quality monitoring data of a prediction area;

[0094] 202, determining the air quality index of the prediction area in a preset time according to the air quality monitoring data;

[0095] 203, obtaining ground meteorological element data, sounding data and circulation characteristic data of the prediction area in the preset time;

[0096] In the embodiment, steps 201 to 203 are similar to the aforementioned steps 101 to 103, and thus are not described herein.

[0097] In some specific embodiments, the ground meteorological element data obtained by the system includes air temperature, precipitation, wind speed, wind direction, relative humidity, visibility, and sea level pressure, the sounding data includes sounding data at 20 hours, 500hpa, 700hpa, 850hpa, and 925hpa, and the circulation feature data includes height field, temperature field, wind field, and pressure field at 500hpa, 700hpa, 850hpa, 925hpa, and sea level.

[0098] 204. Analyze the correlation between the air quality index and the ground meteorological element data, the sounding data, and the circulation feature data to obtain a corresponding correlation coefficient;

[0099] In the embodiment, the system analyzes the correlation between the air quality index and the ground meteorological element data, the sounding data, and the circulation feature data, and measures the linear correlation degree between them by calculating the correlation coefficient. The value range of the correlation coefficient is generally between -1 and 1, 0 indicates no correlation, a positive value indicates positive correlation, a negative value indicates negative correlation, and the closer the absolute value is to 1, the stronger the correlation.

[0100] It should be noted that when analyzing the correlation between the air quality index and the ground meteorological element data, the sounding data, and the circulation feature data, the data of the same period and the data of the previous period should be considered respectively. In some specific embodiments, the correlation between the air quality index and the ground and sounding data of the previous 1-45 days is found by calculating the correlation between the air quality index and the ground and sounding data of the previous 1-45 days. It is found that the pressure, air temperature, wind speed, relative humidity, and sunshine of the previous period still have a significant correlation with the air quality index, among which the pressure has a positive correlation, the remaining elements have a negative correlation, the sunshine of the previous 1-3 days has a positive correlation, and the air quality index and the sunshine also have a negative correlation as the time increases. The correlation coefficients of the temperature, temperature dew point, and wind speed of the sounding data at each layer increase with the decrease of the height, the correlation coefficients of the temperature, temperature dew point, and wind speed at 925hpa 20 points are the highest, the correlation coefficient of the highest pressure is higher than that of the average pressure and the lowest pressure, the correlation coefficient of the lowest temperature is higher than that of the average temperature and the highest temperature, and the correlation coefficient of the precipitation from 08 to 08 is higher than that of other statistical periods.

[0101] It should be noted that for the relationship between pollution weather and the same period circulation in the circulation characteristic data, the sea level pressure causing the prediction area regional pollution can be divided into several types by using the objective weather typing method, and then the influence of each type on the air quality index is analyzed to obtain the key circulation area affecting the air quality index. For the relationship between pollution weather and the previous circulation in the circulation characteristic data, the rhythm and cycle method can be used. Practice shows that this method plays an important role in the short-term climate prediction business in the northwest region, especially in the prediction of important weather processes in the extended period. The principle of the 150-day rhythm method is roughly from the frequency analysis of weather processes. On average, a weather process occurs every 3-5 days. From the perspective of meteorological data statistics, meteorological elements are usually counted in daily, candidate, decade, month, season, and year periods. Combining the two, the 500 hPa average geopotential height field at the candidate scale is used to represent the key area, which will not have too much "noise" and will not smooth out too much information.

[0102] 205、According to the correlation coefficient and the characteristics of the meteorological system in the prediction area, the target factor is screened from the ground meteorological element data, sounding data and circulation characteristic data;

[0103] The system screens the characteristics significantly related to the air quality index from all data as the input characteristics of the first prediction model according to the results of the correlation coefficient. In addition, when screening the target factor, the system also needs to consider the characteristics and influencing factors of the meteorological system in the prediction area, such as geographical location, climate type, sea-land distribution, terrain and other factors, which may have important influence on meteorological diffusion conditions, so as to ensure that the selected target factor can better reflect the characteristics and variation rules of the local meteorological system, that is, the system needs to select meteorological elements and circulation characteristics with clear physical meaning as target factors, which can directly or indirectly affect the diffusion and change of atmospheric pollutants, thereby affecting the air quality.

[0104] It should be noted that the screening process of the target factor may be adjusted according to different prediction areas and prediction needs, so when screening the target factor, the specific situation and related analysis results should be combined, and the specific place is not limited here.

[0105] 206、Based on the air quality index and the target factor, a first prediction model is established by a multivariate stepwise regression method;

[0106] The system adopts a multiple stepwise regression method to establish the first prediction model. The multiple stepwise regression method can gradually introduce target factors. In each step, the performance of the model after adding a factor is compared to determine whether to add the factor and the order of addition. The multiple stepwise regression method can be used to select an optimal combination of prediction factors, avoid excessive or unnecessary features, and thus further filter out the most important target factors and remove redundant or insignificant factors to construct an optimal first prediction model.

[0107] In some specific embodiments, the correlation analysis result indicates that the daily air quality index in autumn and winter in a certain place has a significant correlation with the average temperature, average pressure, maximum temperature, maximum pressure, minimum temperature, minimum pressure, daily precipitation, sunshine, average relative humidity, 2m average wind speed, 10m average wind speed, minimum visibility, maximum wind speed, temperature at each height layer, and dew point temperature in the same period and the previous period, especially the temperature at 925hpa, the maximum ground pressure, the minimum temperature, the wind speed, and the relative humidity.

[0108] Further, the circulation similarity method can be used to select the previous circulation as a prediction field to further improve the accuracy of the prediction model when establishing the first prediction model. The circulation similarity method is a prediction method based on historical meteorological circulation fields. It uses similar historical meteorological circulation conditions as prediction scenarios to predict future changes in air quality index. Specifically, the system quantitatively determines the similarity between two circulation fields by calculating the similarity coefficient between the same period circulation field and the previous circulation field. When establishing the prediction model, the prediction field is used as one of the input data to provide estimates of meteorological elements in the future period. By combining the prediction field with other target factors, the first prediction model can more accurately predict the future trend of air quality index in the extended period.

[0109] 207. Determine the meteorological diffusion condition level in the extended period according to the prediction result of the first prediction model.

[0110] In this embodiment, step 207 is similar to step 105 of the previous embodiment, which will not be described here.

[0111] It should be noted that in actual application, the NCEP reanalysis real-time circulation field can be used as the input of the first prediction model. The NCEP reanalysis real-time circulation field data can provide high temporal and spatial resolution real-time meteorological data globally. The system can filter out the corresponding target factors to run the first prediction model to predict the air quality index in the extended period.

[0112] In the embodiment, by studying the correlation between the air quality index and the previous and current meteorological elements and the circulation field, the meteorological elements and the circulation features with clear physical meaning are selected as the target factors according to the correlation, the previous circulation is selected as the prediction field by using the circulation similarity method, the first prediction model of the air quality index in the extended period is constructed by using the multivariate stepwise regression method, and the air quality index predicted by the first prediction model is converted into the corresponding meteorological diffusion condition grade, so that the prediction of the meteorological diffusion condition in the extended period (11-30 days) is realized, the predictable time of the meteorological diffusion condition is prolonged, and the decision-making meteorological service capability of the environmental meteorology is effectively improved.

[0113] In the application, in addition to the prediction of the meteorological diffusion condition in the extended period, the prediction of the meteorological diffusion condition based on the month and season scales can be further performed, which will be specifically described below. Please refer to Figure 3 , Figure 3 Another embodiment of the prediction method of the meteorological diffusion condition provided in the application includes:

[0114] 301. Obtain air quality monitoring data of a prediction area;

[0115] 302. Determine an air quality index of the prediction area in a preset time according to the air quality monitoring data;

[0116] 303. Obtain ground meteorological element data, sounding data and circulation feature data of the prediction area in the preset time;

[0117] 304. Analyze the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation feature data, and establish a first prediction model according to the analysis result, the first prediction model being used for predicting the air quality index in the extended period;

[0118] 305. Determine a meteorological diffusion condition grade in the extended period according to the prediction result of the first prediction model;

[0119] In the embodiment, the steps 301 to 305 are similar to the steps 101 to 105 of the foregoing embodiment, and will not be described herein.

[0120] 306. Determine an air pollution ratio of each month and each season according to the air quality monitoring data;

[0121] In the embodiment, the meteorological diffusion condition of the month and season scales is evaluated according to the air pollution ratio, and the air pollution ratio specifically refers to the ratio of the number of days with the air quality index reaching the light pollution or above to the total number of days of each month and each season. The system needs to determine the air pollution ratio of each month and each season of the prediction area in recent years according to the air quality monitoring data of the prediction area.

[0122] Further, the system can divide the monthly and seasonal diffusion conditions into three categories of good, medium and poor according to the value of the air pollution ratio. For example, if the monthly / seasonal air pollution ratio reaches 50% or above, the meteorological diffusion condition of the month / season is determined to be poor; if the monthly / seasonal air pollution ratio is between 30% and 50%, the meteorological diffusion condition of the month / season is determined to be medium; and if the monthly / seasonal air pollution ratio is 30% or below, the meteorological diffusion condition of the month / season is determined to be good. It should be noted that in the above classification of the meteorological diffusion condition on the monthly / seasonal scale, other levels can be further refined according to actual application requirements, which are not limited here.

[0123] In some specific embodiments, in order to exclude some occasional influences, the system can count the number of typical persistent pollution weather processes in each month and each season, and calculate the air pollution ratio of each month and each season based on this, which is described in detail below:

[0124] A1. Count the number of typical persistent pollution weather processes in each month and each season according to air quality monitoring data.

[0125] The system counts the number of typical persistent pollution weather processes in each month and each season in the prediction area according to the air quality monitoring data of the prediction area. The typical persistent pollution weather process specifically refers to a pollution weather process in which the air quality index reaches light pollution or above and the duration is greater than or equal to 3 days. By analyzing the typical persistent pollution weather process, the situation of air pollution for multiple consecutive days can be considered comprehensively to obtain more comprehensive and long-term data samples, and thus the contingency and changes of a single weather condition can be eliminated and the regularity of the meteorological diffusion condition can be more accurately reflected.

[0126] A2. Determine the air pollution ratio of each month and each season according to the number of typical persistent pollution weather processes in each month and each season.

[0127] The system counts the number of typical persistent pollution weather processes in each month and each season, and then multiplies the corresponding number of days as the number of days in which pollution is caused by meteorological conditions in the month or season, and then calculates the ratio of the number of days to the total number of days in the month or season, and determines the calculation result as the air pollution ratio of the month or season. The air pollution ratio calculated in this way can reflect the influence of long-term meteorological conditions, so as to more comprehensively and accurately reflect the relationship between the air pollution ratio and the meteorological diffusion condition.

[0128] 307、analyzing the correlation between the air pollution ratio of each month and each season and each of the circulation indices in the climate system circulation index, and establishing a second prediction model according to the analysis result, the second prediction model being used for predicting the air pollution ratio at the monthly and seasonal scales;

[0129] In this embodiment, since the meteorological diffusion condition is predicted at the monthly and seasonal scales, more attention needs to be paid to the influence of long-term climate characteristics and the climate system circulation index, and the influence of short-time meteorological elements and sounding data is no longer considered. This is because, at the long-term scale, climate change and the movement of the climate system will have a relatively persistent and stable influence on the meteorological diffusion condition, while short-time meteorological elements and sounding data may be affected by daily climate fluctuations and weather changes, and their influence is relatively weak at the long-term scale. By analyzing the correlation between the air pollution ratio and the climate system circulation index at the monthly and seasonal scales, the general rules of the meteorological diffusion condition in different months and seasons can be captured.

[0130] It should be noted that the climate system circulation index in the present application refers to the circulation index set issued by the National Climate Center, which includes 88 atmospheric circulation indices and 26 sea temperature circulation indices. Among them, the atmospheric circulation index is an index describing the motion state and characteristics of the atmospheric circulation system. These indices can be used to measure the strength, position and change of the atmospheric circulation system, for example, the Southern Oscillation Index (SOI), which can be used to describe the atmospheric circulation oscillation state in the southern hemisphere of the Pacific Ocean. The sea temperature circulation index is an index describing the distribution of sea surface temperature, for example, the El Niño-Southern Oscillation (ENSO) index, which can be used to describe the abnormal change of sea temperature in the equatorial region of the Pacific Ocean. These atmospheric circulation indices and sea temperature circulation indices are important indicators of the motion state of the atmosphere and ocean, and they play a key role in climate and weather change, and can reflect large-scale meteorological and climate change. Since the formation and transmission of air pollution are closely related to meteorological conditions and climate state, the atmospheric circulation index and the sea temperature circulation index are of great significance for the prediction of the air pollution ratio. Based on this, in this embodiment, the atmospheric circulation index and the sea temperature circulation index are selected as the prediction factor of the air pollution ratio, the correlation between the air pollution ratio of each month and each season and each of the circulation indices is analyzed, and then a second prediction model of the air pollution ratio at the monthly and seasonal scales is established based on the analysis result.

[0131] The establishment process of the second prediction model includes the following steps:

[0132] B1, analyze the correlation between the air pollution ratio and each of the circulation indices in the climate system on a monthly or seasonal scale, and select target indices that affect the air pollution ratio in each month or each season according to the analysis results;

[0133] The correlation between the air pollution ratio and each of the circulation indices is analyzed on a monthly or seasonal scale, and target indices that have a significant correlation with the air pollution ratio in each month or each season are selected according to the analysis results. In actual application, the six circulation indices with the highest correlation coefficients can be directly selected as the target indices.

[0134] Specifically, for the monthly scale, the system needs to analyze the monthly correlation between the air pollution ratio in each month and each of the circulation indices in the climate system in the same period and the previous three months; for the seasonal scale, the system needs to analyze the seasonal correlation between the air pollution ratio in each season and each of the circulation indices in the climate system in the same period and the previous two seasons.

[0135] B2, based on the air pollution ratio and the target indices, a second prediction model is established by a multivariate stepwise regression method.

[0136] The system performs multivariate stepwise regression based on the air pollution ratio in each month and the corresponding target indices, as well as the key circulation area in the same period, to obtain the second prediction model of the air pollution ratio on a monthly scale.

[0137] 308, determine the grade of the meteorological diffusion condition on a monthly or seasonal scale according to the prediction result of the second prediction model.

[0138] The system can predict the air pollution ratio in each month or each season by the second prediction model, and then convert the air pollution ratio into different grades of the meteorological diffusion condition according to the grade division, so as to obtain the prediction result of the meteorological diffusion grade in each month or each season, and realize the prediction of the meteorological diffusion condition on a monthly or seasonal scale.

[0139] In this embodiment, in addition to predicting the meteorological diffusion condition in the extended period, the meteorological diffusion condition can also be predicted on a monthly or seasonal scale, the air pollution ratio in the monthly or seasonal scale is used to quantify the meteorological diffusion condition, and quantitative prediction is realized, which has a good prediction effect. That is, the prediction method of the meteorological diffusion condition provided in this embodiment can cover the extended period, month and season as three different time scales, and further extend the predictable time limit of the meteorological diffusion condition.

[0140] Please refer to Figure 4 , Figure 4 An embodiment of the prediction system of the meteorological diffusion condition provided in this application, the prediction system comprises:

[0141] The first obtaining unit 401 is configured to obtain air quality monitoring data of a prediction area.

[0142] The first determining unit 402 is configured to determine an air quality index of the prediction area within a preset time according to the air quality monitoring data.

[0143] The second obtaining unit 403 is configured to obtain ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time.

[0144] The first analyzing unit 404 is configured to analyze a correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and establish a first prediction model according to an analysis result, the first prediction model being used to predict the air quality index within an extended period.

[0145] The first prediction unit 405 is configured to determine a meteorological diffusion condition grade within the extended period according to a prediction result of the first prediction model.

[0146] Optionally, the first analyzing unit 404 is specifically configured to:

[0147] analyze the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and obtain a corresponding correlation coefficient;

[0148] screen a target factor in the ground meteorological element data, the sounding data and the circulation characteristic data according to the correlation coefficient and a characteristic of a meteorological system within the prediction area;

[0149] establish the first prediction model based on the air quality index and the target factor by using a multivariate stepwise regression method.

[0150] Optionally, the first analyzing unit 404 is specifically configured to:

[0151] calculate a similarity coefficient between a current circulation field and a previous circulation field, and determine the previous circulation field with the highest similarity coefficient as a prediction field;

[0152] establish the first prediction model based on the air quality index, the target factor and the prediction field by using the multivariate stepwise regression method.

[0153] Optionally, the ground meteorological element data includes air temperature, precipitation, wind speed, wind direction, relative humidity, visibility and sea level pressure, the sounding data includes sounding data at 20 o'clock 500 hpa, 700 hpa, 850 hpa and 925 hpa, and the circulation characteristic data includes height fields, temperature fields, wind fields and pressure fields at 500 hpa, 700 hpa, 850 hpa, 925 hpa and sea level.

[0154] Optionally, the prediction system further comprises:

[0155] The second determining unit 406 is configured to determine the air pollution ratio of each month and each season according to the air quality monitoring data.

[0156] The second analyzing unit 407 is configured to analyze the correlation between the air pollution ratio of each month and each season and each of the circulation indexes in the climate system circulation index, and establish a second prediction model according to the analysis result, the second prediction model being used to predict the air pollution ratio of the month and season scale.

[0157] The second prediction unit 408 is configured to determine the weather diffusion condition grade of the month and season scale according to the prediction result of the second prediction model.

[0158] Optionally, the second determining unit 406 is specifically configured to:

[0159] count the number of typical persistent pollution weather processes of each month and each season according to the air quality monitoring data, the typical persistent pollution weather process being a pollution weather process with the air quality index reaching light pollution or above and the duration being greater than or equal to 3 days;

[0160] determine the air pollution ratio of each month and each season according to the number of the typical persistent pollution weather processes of each month and each season.

[0161] Optionally, the second analyzing unit 407 is specifically configured to:

[0162] analyze the correlation between the air pollution ratio and each of the circulation indexes in the climate system circulation index based on the month and season scale, and screen out a target index affecting the air pollution ratio of each month and each season according to the analysis result;

[0163] establish the second prediction model by the method of multivariate stepwise regression based on the air pollution ratio and the target index.

[0164] In the system of the embodiment, the functions of each unit correspond to the steps in the method embodiments shown in the foregoing Figure 1 、 2 , 3, and thus will not be described here again.

[0165] The present application also provides a prediction device of weather diffusion condition, please refer to Figure 5 , Figure 5 An embodiment of the prediction device of weather diffusion condition provided by the present application, the device comprises:

[0166] The processor 501, the memory 502, the input and output unit 503, and the bus 504;

[0167] The processor 501 is connected with the memory 502, the input and output unit 503, and the bus 504.

[0168] The memory 502 stores a program, and the processor 501 invokes the program to perform the prediction method of any one of the above meteorological dispersion conditions.

[0169] The application also relates to a computer readable storage medium, which stores a program, and the program causes a computer to perform the prediction method of any one of the above meteorological dispersion conditions when the program runs on the computer.

[0170] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0171] In the several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. There can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0172] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, can be located in one place, or can be distributed on a plurality of network units. According to actual needs, part or all of the units can be selected to achieve the purpose of the embodiment scheme.

[0173] In addition, each functional unit in the embodiments of the present application can be integrated in one processing unit, or each unit can be a physically independent unit, or two or more units can be integrated in one unit. The integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0174] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

Claims

1. A method of predicting meteorological dispersion conditions, characterized by, The prediction method comprises: obtaining air quality monitoring data of a prediction area; determining an air quality index of the prediction area within a preset time according to the air quality monitoring data; obtaining ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time; analyzing a correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data to obtain a corresponding correlation coefficient; screening a target factor from the ground meteorological element data, the sounding data and the circulation characteristic data according to the correlation coefficient and a characteristic of a meteorological system in the prediction area; establishing a first prediction model by a multivariate stepwise regression method based on the air quality index and the target factor, the first prediction model being used to predict the air quality index within an extended period; determining a meteorological diffusion condition grade within the extended period according to a prediction result of the first prediction model.

2. The prediction method of claim 1, wherein, Before the first prediction model is established by the multivariate stepwise regression method based on the air quality index and the target factor, the prediction method further comprises: calculating a similarity coefficient between a current circulation field and a previous circulation field, and determining the previous circulation field with the highest similarity coefficient as a prediction field; the first prediction model is established by the multivariate stepwise regression method based on the air quality index, the target factor and the prediction field. The ground meteorological element data comprises temperature, precipitation, wind speed, wind direction, relative humidity, visibility and sea level pressure, the sounding data comprises sounding data at 500hpa, 700hpa, 850hpa and 925hpa at 20:00, and the circulation characteristic data comprises height field, temperature field, wind field and pressure field at 500hpa, 700hpa, 850hpa, 925hpa and sea level.

3. The prediction method of claim 1, wherein, After the air quality monitoring data of the prediction area is obtained, the prediction method further comprises:

4. The prediction method of any one of claims 1 to 3, characterized in that, determining an air pollution ratio of each month and each season according to the air quality monitoring data; analyzing a correlation between the air pollution ratio of each month and each season and each circulation index in a climate system circulation index, and establishing a second prediction model according to an analysis result, the second prediction model being used to predict the air pollution ratio of a monthly and seasonal scale; determining a meteorological diffusion condition grade of a monthly and seasonal scale according to a prediction result of the second prediction model. The air pollution ratio of each month and each season is determined according to the air quality monitoring data, which comprises:

5. The prediction method of claim 4, wherein, counting a number of typical persistent pollution weather processes of each month and each season according to the air quality monitoring data, the typical persistent pollution weather process being a pollution weather process with an air quality index reaching light pollution or above and a duration of more than or equal to 3 days; determining the air pollution ratio of each month and each season according to the number of the typical persistent pollution weather processes of each month and each season. ​ 6. The prediction method of claim 4, wherein, The correlation between the air pollution ratio of each month and each season and each of the circulation indexes in the climate system circulation index is analyzed, and a second prediction model is established according to the analysis result, including: The correlation between the air pollution ratio and each of the circulation indexes in the climate system circulation index is analyzed on a monthly and seasonal scale, and a target index affecting the air pollution ratio of each month and each season is screened out according to the analysis result; A second prediction model is established by a multivariate stepwise regression method based on the air pollution ratio and the target index.

7. A system for predicting meteorological dispersion conditions, characterized in that, The prediction system includes: A first acquisition unit configured to acquire air quality monitoring data of a prediction area; A first determination unit configured to determine an air quality index of the prediction area within a preset time according to the air quality monitoring data; A second acquisition unit configured to acquire ground meteorological element data, sounding data and circulation characteristic data of the prediction area within the preset time; A first analysis unit configured to analyze the correlation between the air quality index and the ground meteorological element data, the sounding data and the circulation characteristic data, and obtain a corresponding correlation coefficient; screen a target factor from the ground meteorological element data, the sounding data and the circulation characteristic data according to the correlation coefficient and the characteristics of a meteorological system in the prediction area; and establish a first prediction model by a multivariate stepwise regression method based on the air quality index and the target factor, the first prediction model being used to predict an air quality index in an extended period; A first prediction unit configured to determine a meteorological diffusion condition level in the extended period according to the prediction result of the first prediction model.

8. An apparatus for predicting meteorological dispersion conditions, characterized by, The prediction device includes: A processor, a memory, an input / output unit and a bus; The processor is connected with the memory, the input / output unit and the bus; The memory stores a program, and the processor invokes the program to execute the prediction method in any one of claims 1 to 6. 9.A computer readable storage medium, which stores a program, and the program performs the prediction method in any one of claims 1 to 6 when executed on a computer.

Images