Sea fog forecasting index extraction method and sea fog forecasting method thereof

Abstract

The invention discloses a sea fog forecasting index extraction method and a sea fog forecasting method thereof. The method comprises the steps of screening sea fog historical examples of a target sea area; determining an initially set forecasting index variable; determining a forecast index variable and a threshold thereof; on the basis of obtaining the forecasting indexes of the target sea area and the threshold value of the forecasting indexes, the numerical forecasting product of the target sea area at a certain time in the future is obtained, the determined forecasting index variable is extracted from the product, and if more than 80% of the forecasting indexes meet the threshold value within the forecasting time, the target sea area is forecasted to be foggy. According to the invention, a sea fog forecasting index extraction method under different complex weather situations is innovatively provided by utilizing reanalysis data, visibility data of a meteorological observation station in a target sea area and a percentile method, index variable values are read or calculated based on a real-time numerical model forecasting product in business, and sea fog forecasting of the target sea area is carried out according to whether the index variable values accord with a threshold value or not; a new index forecasting method is provided for sea fog forecasting, so that the sea fog forecasting level is further improved.

Description

Technical Field

[0001] The invention relates to a method for extracting sea fog forecast indicators and a forecasting method thereof, and belongs to the technical field of marine meteorology. Background Art

[0002] Sea fog is one of the important marine meteorological disasters. Sea fog is a weather phenomenon that occurs at sea, islands or target sea areas under the influence of the ocean, making the atmospheric horizontal visibility less than 1 km. When sea fog occurs, the horizontal visibility at sea or in the target sea area is reduced, which seriously affects the transportation, military activities, fishery production and agricultural production activities at sea and in the target sea area, and is also an important cause of various accidents at sea and in the target sea area. There is a lot of sea fog along the coastal areas of my country. With the increasing frequency of marine activities such as shipping, fishing, and production, sea fog forecasting is particularly important.

[0003] Sea fog often appears in the rear of high pressure at sea, in the situation of low in the west and high in the east, or in the situation of equal pressure field. In daily business forecasting, forecasters have a certain ability to forecast this type of sea fog. In addition to the common types of sea fog, sea fog may also occur when conditions are right in complex weather situations such as typhoons, cyclones, southerly gales, and the edge of the subtropical high. At present, forecasters' ability to forecast sea fog processes under complex weather situations still needs to be improved. Extracting sea fog forecast indicators under complex weather backgrounds is conducive to understanding the formation mechanism of sea fog, thereby improving the sea fog forecast service capabilities under complex weather backgrounds and facilitating disaster prevention and mitigation work. Summary of the invention

[0004] The purpose of the present invention is to provide a sea fog forecast index extraction method and a sea fog forecast method thereof, so as to improve the sea fog forecast level of target sea area under complex weather background.

[0005] A method for extracting sea fog forecast indicators is characterized by comprising the following steps: 1) Screening historical cases of sea fog in the target sea area; 2) Determine the initial forecast indicators: Obtain reanalysis data in the target sea area, and for each meteorological observation station, read the meteorological elements on the grid point closest to the observation station, including ground meteorological elements and high-altitude meteorological elements; The ground meteorological elements include: air-sea temperature difference (℃), 2m temperature dew point difference (℃), 2m relative humidity (%), 10m wind speed (m / s), wind direction (°), 10m gust wind speed (m / s), 10m east-west wind speed (m / s), north-south wind speed (m / s), north-south turbulent wind stress (N / m 2 ), east-west turbulent wind stress (N / m 2 ), sensible heat flux (W / m 2 ), latent heat flux (W / m 2 ); High-altitude meteorological elements include: 1000hPa, 925hPa, 850hPa relative humidity (%), specific humidity (g / kg), water vapor flux divergence (g•hPa -1 •cm -2 •s -1 ), wind direction (°) and wind speed (m / s), east-west wind speed (m / s), north-south wind speed (m / s), vertical velocity (Pa / s) and temperature advection (℃ / s), 975hPa, 950hPa, 925hPa, 850hPa air temperature and the difference between the temperature of each layer and the surface temperature (℃); and calculate the VIS index: VIS= (T s -T d ) / RH 1.67 , where T s is the surface temperature (K), T d is the dew point temperature 2 m above the ground surface (K), RH is the relative humidity 2 m above the ground surface (%); 3) Determine forecast indicators: The standard deviation and mean of each of the above-mentioned preliminary forecast indicators are calculated respectively, and the dispersion of the samples is measured by the ratio of the standard deviation to the mean, i.e., the coefficient of variation. The smaller the coefficient of variation, the more concentrated the values ​​of the preliminary forecast indicators are. The preliminary forecast indicators with a coefficient of variation less than 30% are used as sea fog forecast indicators. 4) Further determine the threshold of the forecast indicator: 4.1) Draw a box plot of the above sea fog forecast indicators. The corresponding data from top to bottom of the box plot are arranged from large to small, namely, maximum value, 75% quantile, 50% quantile (i.e. median), 25% quantile, and minimum value; 4.2) The threshold of the forecast indicator is determined based on the 75% quantile or 25% quantile of the box plot of the forecast indicator. If the value range of the samples less than 25% in the box plot is larger than the value range of the samples above 75%, the quantile greater than 25% is selected as the threshold of the forecast indicator; conversely, if the value range of the samples less than 25% in the box plot is smaller than the value range of the samples above 75%, the quantile less than 75% is selected as the threshold of the forecast indicator.

[0006] Preferably, the step 1 is to obtain at least five years of historical data from meteorological observation stations in the target sea area and extract visibility data therefrom, and classify foggy days based on visibility <1 km.

[0007] After step 1, the process further includes screening four types of complex weather situation sea fog processes: determining the weather situation of the foggy day based on a conventional ground weather map, and screening out four types of complex weather situation sea fog processes: cyclone type, typhoon type, southerly gale type, and subtropical high edge type; Among them, cyclone-type sea fog is defined as heavy fog with cyclonic activity in the target sea area (isobars are spaced 2.5hPa apart, with at least one closed isobar) and visibility <1km; typhoon-type sea fog is heavy fog that occurs in the first and fourth quadrants of the typhoon when the typhoon center is within 1000km of the target sea area; southerly gale-type sea fog is heavy fog that occurs when the south wind speed in the coastal area of ​​the target sea area reaches 6m / s or more; subtropical high edge-type sea fog is heavy fog that occurs when the target sea area is within 200km of the 500hPa 588 line.

[0008] In step 2, the reanalysis data may be ERA5 reanalysis data or FNL reanalysis data.

[0009] The Yellow Sea fog forecast indicators and their thresholds under four types of complex weather backgrounds determined by the above method are as follows: For cyclonic weather background, there are 7 forecast indicators, namely latent heat flux, sensible heat flux, 10-meter wind speed, north-south turbulent wind stress, east-west turbulent wind stress, 925hPa temperature, and the temperature difference between the surface temperature and 850hPa. The thresholds of the 7 indicators are latent heat flux ~ less than 75% percentile, sensible heat flux ~ less than 75% percentile, 10-meter wind speed ~ less than 75% percentile, north-south turbulent wind stress ~ greater than 25% percentile, east-west turbulent wind stress ~ less than 75% percentile, 925hPa temperature ~ greater than 25% percentile, and the temperature difference between the surface temperature and 850hPa ~ greater than 25% percentile. For typhoon-type weather background, there are 6 forecast indicators, namely latent heat flux, 10-meter wind speed, 2-meter relative humidity, air-sea temperature difference, 925hPa temperature advection, and surface temperature and 850hPa temperature difference; the thresholds of the 6 indicators are latent heat flux~greater than 25% percentile, 10-meter wind speed~greater than 25% percentile, 2-meter relative humidity~greater than 25% percentile, air-sea temperature difference~greater than 25% percentile, 925hPa temperature advection~greater than 25% percentile, and surface temperature and 850hPa temperature difference~less than 75% percentile; For the subtropical high edge weather background, there are four forecast indicators, namely, sensible heat flux, 2m relative humidity, 1000hPa relative humidity, and the temperature difference between the surface temperature and 850hPa; the thresholds of the four indicators are sensible heat flux ~ less than 75% percentile, 2m relative humidity ~ greater than 25% percentile, 1000hPa relative humidity ~ greater than 25% percentile, and the temperature difference between the surface temperature and 850hPa ~ greater than 25% percentile; For the weather background of strong south wind type, there are four forecast indicators, namely, north-south turbulent wind stress, 10-meter wind direction, 850hPa wind speed, and 850hPa temperature advection; the thresholds of the four indicators are north-south turbulent wind stress ~ greater than 25% quantile, 10-meter wind direction ~ greater than 25% quantile, 850hPa wind speed ~ greater than 25% quantile, and 850hPa temperature advection ~ greater than 25% quantile.

[0010] It can be seen that under the four weather backgrounds, a total of 14 preliminary forecast indicators are included in the forecast indicators, and their statistics are shown in Table 1: Table 1 Sea fog forecast indicators and their thresholds under four types of complex weather backgrounds

[0011] A sea fog forecasting method using the sea fog forecast index extraction method is characterized by comprising the following steps: 1) Obtain the forecast indicators and thresholds of the target sea area according to the above method; 2) Obtain the numerical forecast product of the target sea area at a certain time in the future, and extract the forecast indicators determined in the previous step from it; if the numerical forecast product does not provide sea surface heat flux, north-south turbulent wind stress, and east-west turbulent wind stress, these three indicators are calculated separately; 3) If more than 80% of the forecast indicators meet the threshold within the forecast period, it is predicted that there will be fog in the target sea area.

[0012] The sea surface heat flux is calculated by using the COARE 3.0 heat flux algorithm to calculate the sensible and latent heat fluxes of the sea-air interface at a specified point; this algorithm is based on the block method. , in H s is the sensible heat flux, H l is the latent heat flux; C is the heat transfer coefficient, which is taken as 0.00115; S is the sea surface wind speed, replaced by the 10m wind speed; T s is the sea surface temperature; q s is the sea surface specific humidity, which is replaced by the 2m specific humidity; and q For air specific humidity, it is replaced by 1000hPa specific humidity; c pa is the specific heat capacity of air, take 1004.67; θ is the potential temperature, which is defined by the air temperature T , air pressure p and air humidity q calculate: , ρ a is the air density, using the sea surface pressure p and 2m air temperature t Perform the calculation: L e is the sea surface Monin-Obukhov length, using the sea surface temperature T s calculate: .

[0013] The calculation of the north-south turbulent wind stress and the east-west turbulent wind stress is as follows: Using the 10m wind field as the approximate surface wind, the turbulent wind stress is calculated using the approximate formula: , in, is the surface wind stress, ρ is the air density, is the surface wind, C s is the surface resistance coefficient, and the empirical formula is: , in, V is the wind speed; The zonal wind component is used to calculate the east-west turbulent wind stress, and the meridional wind component is used to calculate the north-south turbulent wind stress.

[0014] The present invention uses the data of observation stations in the target sea area and its coastal areas and conventional ground weather maps to screen out historical cases of sea fog. Based on the reanalysis grid data, according to the preliminary preset indicators of the physical quantities that affect fog formation, the percentile method is used to extract the sea fog forecast indicators of the target sea area under four types of complex weather conditions; and according to the real-time numerical forecast product reading indicators, the sea fog in the target sea area is forecasted in real time according to the threshold, so as to be applied in business.

[0015] The percentile method is often used in the study of physical quantity indicators of severe convective weather environments such as short-term heavy precipitation, thunderstorms, gale, and hail. This method is simple and intuitive, and is an effective method for extracting forecast indicators. Based on the percentile method, sea fog forecast indicators are extracted under various complex weather conditions including typhoons, cyclones, southerly gales, and the edge of the subtropical high, aiming to provide business personnel with sea fog forecast support technology under complex weather backgrounds.

[0016] The present invention innovatively proposes a method for extracting sea fog forecast indicators under different complex weather conditions by using reanalysis data, visibility data of meteorological observation stations in the target sea area, and percentile methods. The index value is read or calculated based on the real-time numerical model forecast product in the business, and the sea fog forecast for the target sea area is performed according to whether the threshold is met. This provides a new index forecasting method for sea fog forecasting, so as to further improve the level of sea fog forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flow chart of the method for extracting sea fog forecast indicators.

[0018] Figure 2 These are the box plots of some forecast indicators, where (a) is the box plot of the air-sea temperature difference, and (b) is the box plot of 10 wind directions. The percentages in the figures are the coefficients of variation.

[0019] Figure 3 The present invention is a flow chart of the sea fog forecasting method. DETAILED DESCRIPTION

[0020] The implementation of the present invention is described in detail below in conjunction with the accompanying drawings.

[0021] The purpose of the present invention is to provide a method for extracting sea fog forecast indicators and a forecasting method thereof, so as to improve the level of sea fog forecasting under complex weather backgrounds.

[0022] The data used in the present invention are the observation data of automatic stations and buoy stations in the Yellow Sea from 2016 to 2022, ERA5 reanalysis data, and conventional ground weather maps. Four types of sea fog processes occurring under complex weather conditions are screened out based on the ground weather maps. Based on the reanalysis grid data, the forecast index is preliminarily determined according to the physical quantities affecting fog formation, and the percentile method is used to determine the sea fog forecast index of the target sea area under four types of complex weather conditions, and the sea fog forecast of the target sea area is realized based on the real-time numerical forecast product.

[0023] A method for extracting sea fog forecast indicators, such as Figure 1 As shown, Step 1: Based on the visibility data obtained from the meteorological observation station in the target sea area, foggy days are classified as those with visibility less than 1 km.

[0024] Based on the conventional ground weather map, the weather situation of the above-mentioned foggy day is determined, and four types of complex weather situation sea fog processes are further screened out, namely, cyclone type, typhoon type, south gale type and subtropical high edge type. Among them, cyclone type sea fog is defined as heavy fog with visibility less than 1km when there is cyclonic activity in the target sea area (isobars are spaced 2.5hPa, at least one closed isobar); typhoon type sea fog is heavy fog that occurs in the first and second quadrants of the typhoon when the typhoon center is within 1000km from the target sea area; south gale type sea fog is heavy fog that occurs when the south wind speed in the coastal area of ​​the target sea area reaches 6m / s or more; subtropical high edge type sea fog is heavy fog that occurs when the target sea area is within 200km of the 500hPa 588 line. Taking the above four types of complex weather situations as examples, the sea fog forecast index extraction method of the present invention is first described.

[0025] Step 2: For each observation station, read the meteorological elements in the reanalysis data at the nearest grid point in the target sea area as the initial forecast indicators. The ground meteorological elements include: air-sea temperature difference (℃), 2m temperature dew point difference (℃), 2m relative humidity (%), 10m wind speed above the surface (m / s), wind direction (°), 10m gust wind speed (m / s), 10m east-west wind speed (m / s), north-south wind speed (m / s), north-south turbulent wind stress (N / m 2 ), east-west turbulent wind stress (N / m 2 ), sensible heat flux (W / m 2 ), latent heat flux (W / m 2 ); high-altitude meteorological elements include: relative humidity (%), specific humidity (g / kg), water vapor flux divergence (g•hPa) on the 1000hPa, 925hPa, and 850hPa isobaric surfaces -1 •cm -2 •s -1 ), wind direction (°) and wind speed (m / s), east-west wind speed (m / s), north-south wind speed (m / s), vertical velocity (Pa / s) and temperature advection (℃ / s), the temperature on the 975hPa, 950hPa, 925hPa, 850hPa isobaric surfaces and the difference between the temperature of each layer and the surface temperature (℃); and calculate the VIS index: VIS= (T s -T d ) / RH 1.67 , where T s is the surface temperature (K), T d is the dew point temperature 2 m above the ground surface (K), and RH is the relative humidity 2 m above the ground surface (%).

[0026] Step 3: The selection of forecast indicators is based on the dispersion of the initial forecast indicators. The more concentrated the variable data values ​​are, the more reliable the variable is as a forecast indicator. The standard deviation and mean of each initial forecast indicator are calculated respectively, and the dispersion of the sample is measured by the ratio of the standard deviation to the mean, i.e., the coefficient of variation, to eliminate the influence of different variable dimensions. The smaller the coefficient of variation, the more concentrated the values ​​of the initial forecast indicator are. Variables with a coefficient of variation of less than 30% are used as forecast indicators for sea fog.

[0027] Step 4.1: For the above sea fog forecast indicators, four types of box plots are drawn, namely cyclone, typhoon, periphery of subtropical high and southerly gale type. The box plots represent all the data in the sample arranged from large to small from top to bottom, namely maximum value, 75% quantile, 50% quantile (median), 25% quantile and minimum value.

[0028] Step 4.2: The threshold of the forecast indicator is determined based on the 75% quantile or 25% quantile of the box plot of the forecast indicator. If the value range of the samples less than 25% in the box plot is larger than the value range of the samples above 75%, the quantile greater than 25% is selected as the threshold of the forecast indicator; conversely, if the value range of the samples less than 25% in the box plot is smaller than the value range of the samples above 75%, the quantile less than 75% is selected as the threshold of the forecast indicator.

[0029] by Figure 2 Taking (a) and (b) in the figure as an example, the minimum coefficient of variation of the air-sea temperature difference in typhoon-type sea fog is 25%, which is used as the forecast indicator of typhoon-type sea fog. The minimum coefficient of variation of the 10-meter wind direction in the southerly gale-type sea fog is 7%, which is used as the forecast indicator of southerly gale-type sea fog. The threshold of the forecast indicator is determined based on the 75% quantile or 25% quantile of the initial forecast indicator box plot. If the range of sample values ​​less than 25% in the box plot is greater than the range of sample values ​​above 75%, the quantile value greater than 25% is selected as the threshold of the forecast indicator. Conversely, if the range of sample values ​​less than 25% in the box plot is less than the range of sample values ​​above 75%, the quantile value less than 75% is selected as the threshold of the forecast indicator. Figure 2 As shown in (a) and (b), the threshold of the forecast index of typhoon sea fog for the difference between air and sea temperature is taken as greater than the 25th percentile value, and the threshold of the forecast index of southerly gale sea fog for the 10-meter wind direction is also taken as greater than the 25th percentile value. The threshold range includes 75% of the samples. The forecast indexes and thresholds of four complex sea fog are shown in Table 1.

[0030] Table 1 Sea fog forecast indicators and their thresholds under four types of complex weather backgrounds

[0031] like Figure 3 ,After obtaining the sea forecast indicators under four types of complex weather situations, we further carry out sea fog forecast.

[0032] Step 5: Use the numerical forecast product of the target sea area at a certain time in the future, and read the relevant meteorological elements of the designated point or calculate the indicator variable value of the current grid point according to the forecast indicators under different types of complex weather conditions determined in step 4.2. If 80% of the indicator values ​​calculated within the forecast period meet the threshold determined in step 3.3, it is predicted that there will be fog in the target sea area.

[0033] The sea surface heat flux in step 5 is calculated using the COARE 3.0 heat flux algorithm to calculate the sensible and latent heat fluxes at the sea-air interface at a specified point. This algorithm is based on the block method. , in H s is the sensible heat flux, H l is the latent heat flux; C is the heat transfer coefficient, which is taken as 0.00115; S is the sea surface wind speed, replaced by the 10m wind speed; T s is the sea surface temperature; q s is the sea surface specific humidity, which is replaced by the 2m specific humidity; and q For air specific humidity, it is replaced by 1000hPa specific humidity; c pa is the specific heat capacity of air, take 1004.67; θ is the potential temperature, which is defined by the air temperature T , air pressure p and air humidity q calculate: , ρ a is the air density, using the sea surface pressure p and 2m air temperature t Perform the calculation: L e is the sea surface Monin-Obukhov length, using the sea surface temperature T s calculate: .

[0034] In step 4, the 10m wind field is used to approximate the surface wind. The zonal wind component is used for the east-west wind stress, and the meridional wind component is used for the north-south surface wind stress. The surface wind stress is calculated using the approximate formula: , in, is the surface wind stress, ρ is the air density, is the surface wind, C s is the surface resistance coefficient, and the empirical formula is: , in, V is the wind speed;

[0035] The present invention innovatively proposes a sea fog forecast index by using reanalysis data, visibility data of meteorological observation stations in the target sea area and the percentile method. In the business, the index variable value is extracted or calculated based on the numerical forecast product, and the sea fog forecast is made according to whether the threshold is met. This provides a new index forecast method for sea fog forecasting in the complex weather background of the target sea area, further improving the level of sea fog forecasting.

Claims

1. A method for extracting sea fog forecast indicators, characterized in that: The following steps are involved: 1) Screening historical cases of sea fog in the target sea area; 2) Determine the initial forecast indicators: Obtain reanalysis data in the target sea area, and for each meteorological observation station, read the meteorological elements on the grid point closest to the station, including ground meteorological elements and high-altitude meteorological elements; 3) Determine forecast indicators: The standard deviation and mean of each preliminary forecast indicator were calculated respectively, and the dispersion of the samples was measured by the ratio of the standard deviation to the mean, i.e. the coefficient of variation. The smaller the coefficient of variation, the more concentrated the values ​​of the preliminary forecast indicator. The preliminary forecast indicators with a coefficient of variation less than 30% were used as sea fog forecast indicators. 4) Further determine the threshold of the forecast indicator: 4.1) Draw a box plot of the above sea fog forecast indicators. The corresponding data from top to bottom of the box plot are arranged from large to small, namely, maximum value, 75% quantile, 50% quantile, 25% quantile, and minimum value; 4.2) The threshold of the forecast indicator is determined based on the 75% quantile or 25% quantile of the box plot of the forecast indicator. If the value range of the samples less than 25% in the box plot is larger than the value range of the samples above 75%, the quantile greater than 25% is selected as the threshold of the forecast indicator; conversely, if the value range of the samples less than 25% in the box plot is smaller than the value range of the samples above 75%, the quantile less than 75% is selected as the threshold of the forecast indicator.

2. The method for extracting sea fog forecast indicators as claimed in claim 1, characterized in that: When screening historical cases of sea fog in the target sea area in step 1, at least five years of historical data of meteorological observation stations in the target sea area are obtained and visibility data is extracted therefrom, and foggy days are divided according to the standard of visibility <1 km.

3. The method for extracting sea fog forecast indicators as claimed in claim 1, characterized in that: In step 2, the reanalysis data includes: ERA5 reanalysis data or FNL reanalysis data.

4. The method for extracting sea fog forecast indicators as claimed in claim 1, characterized in that: In step 2, the ground meteorological elements include: air-sea temperature difference; 2m temperature dew point difference; 2m relative humidity; 10m wind speed above the ground surface; wind direction; 10m gust wind speed; 10m east-west wind speed; north-south wind speed; north-south turbulent wind stress; east-west turbulent wind stress; sensible heat flux; latent heat flux; High-altitude meteorological elements include: 1000hPa, 925hPa, 850hPa relative humidity; specific humidity; water vapor flux divergence; wind direction; wind speed; east-west wind speed; north-south wind speed; vertical velocity; temperature advection; 975hPa, 950hPa, 925hPa, 850hPa air temperature and the difference between the temperature of each layer and the surface temperature; and calculate the VIS index: VIS = (T s -T d ) / RH 1.67 , where T s is the surface temperature; T d is the dew point temperature 2m above the ground surface; RH is the relative humidity 2m above the ground surface.

5. The method for extracting sea fog forecast indicators as claimed in claim 4, characterized in that: After step 1, the process further includes screening four types of complex weather situation sea fog processes: determining the weather situation of the foggy day based on a conventional ground weather map, and screening out four types of complex weather situation sea fog processes: cyclone type, typhoon type, southerly gale type, and subtropical high edge type; Among them, cyclone-type sea fog is defined as cyclonic activity in the target sea area - the isobars are 2.5hPa apart, there is at least one closed isobar, and heavy fog with visibility less than 1km occurs; typhoon-type sea fog is heavy fog that occurs in the first and fourth quadrants of the typhoon when the center of the typhoon is within 1000km of the target sea area; southerly gale-type sea fog is heavy fog that occurs when the south wind speed in the coastal area of ​​the target sea area reaches more than 6m / s; subtropical high edge-type sea fog is heavy fog that occurs when the target sea area is within 200km of the 500hPa 588 line.

6. The method for extracting sea fog forecast indicators as claimed in claim 5, characterized in that: The forecast indicators and thresholds of Yellow Sea fog under four types of complex weather backgrounds determined by this method are as follows: For cyclonic weather background, there are 7 forecast indicators, namely latent heat flux, sensible heat flux, 10-meter wind speed, north-south turbulent wind stress, east-west turbulent wind stress, 925hPa temperature, and the temperature difference between the surface temperature and 850hPa. The thresholds of the 7 indicators are latent heat flux ~ less than 75% percentile, sensible heat flux ~ less than 75% percentile, 10-meter wind speed ~ less than 75% percentile, north-south turbulent wind stress ~ greater than 25% percentile, east-west turbulent wind stress ~ less than 75% percentile, 925hPa temperature ~ greater than 25% percentile, and the temperature difference between the surface temperature and 850hPa ~ greater than 25% percentile. For typhoon-type weather background, there are 6 forecast indicators, namely latent heat flux, 10-meter wind speed, 2-meter relative humidity, air-sea temperature difference, 925hPa temperature advection, and surface temperature and 850hPa temperature difference; the thresholds of the 6 indicators are latent heat flux~greater than 25% percentile, 10-meter wind speed~greater than 25% percentile, 2-meter relative humidity~greater than 25% percentile, air-sea temperature difference~greater than 25% percentile, 925hPa temperature advection~greater than 25% percentile, and surface temperature and 850hPa temperature difference~less than 75% percentile; For the subtropical high edge weather background, there are four forecast indicators, namely, sensible heat flux, 2m relative humidity, 1000hPa relative humidity, and the temperature difference between the surface temperature and 850hPa; the thresholds of the four indicators are sensible heat flux ~ less than 75% percentile, 2m relative humidity ~ greater than 25% percentile, 1000hPa relative humidity ~ greater than 25% percentile, and the temperature difference between the surface temperature and 850hPa ~ greater than 25% percentile; For the weather background of strong south wind type, there are four forecast indicators, namely, north-south turbulent wind stress, 10-meter wind direction, 850hPa wind speed, and 850hPa temperature advection; the thresholds of the four indicators are north-south turbulent wind stress ~ greater than 25% quantile, 10-meter wind direction ~ greater than 25% quantile, 850hPa wind speed ~ greater than 25% quantile, and 850hPa temperature advection ~ greater than 25% quantile.

7. A sea fog forecasting method, characterized in that: The following steps are involved: 1) Obtaining the forecast index and threshold value of the target sea area according to the method of claim 1; 2) Obtain the numerical forecast product of the target sea area at a certain time in the future, and extract the forecast indicators determined in the previous step from it; if the numerical forecast product does not provide sea surface heat flux, north-south turbulent wind stress, and east-west turbulent wind stress, these three indicators are calculated separately; 3) If more than 80% of the forecast indicators meet the threshold within the forecast period, it is predicted that there will be fog in the target sea area.

8. The sea fog forecasting method according to claim 7, characterized in that: The sea surface heat flux is calculated by using the COARE 3.0 heat flux algorithm to calculate the sensible and latent heat fluxes of the sea-air interface at a specified point; this algorithm is based on the block method. ， ， in is the sensible heat flux, is the latent heat flux; is the heat transfer coefficient, which is taken as 0.00115; S is the sea surface wind speed, replaced by the 10m wind speed; is the sea surface temperature; is the sea surface specific humidity, replaced by 2m specific humidity; as well as q For air specific humidity, it is replaced by 1000hPa specific humidity; is the specific heat capacity of air, take 1004.67; is the potential temperature, which is defined by the air temperature , air pressure and air humidity calculate: ， is the air density, using the sea surface pressure and 2m air temperature Perform the calculation: ， is the sea surface Monin-Obukhov length, using the sea surface temperature calculate: 。 9. The sea fog forecasting method according to claim 7, characterized in that: The calculation of the north-south turbulent wind stress and the east-west turbulent wind stress is as follows: Using the 10m wind field as the approximate surface wind, the turbulent wind stress is calculated using the approximate formula: ， in, is the surface wind stress, is the air density, is the surface wind, is the surface drag coefficient, The empirical formula is: ， in, V is the wind speed; The zonal wind component is used to calculate the east-west turbulent wind stress, and the meridional wind component is used to calculate the north-south turbulent wind stress.

Images