Tropical convergence zone characteristic analysis method based on rainfall data

Abstract

The invention provides a tropical convergence zone feature analysis method based on rainfall data, and belongs to the field of data analysis. The rainfall analysis problem of the tropical convergence zone is solved; the specific method comprises the following steps: obtaining standardized data; extracting a core characteristic index of the ITCZ from the standardized data; according to the core characteristic index of the ITCZ, predicting the future precipitation of the central latitude region; calculating water vapor flux and divergence of a central latitude area, and judging a corresponding relation between rainfall and water vapor polymerization: if a rainfall peak value is consistent with a flux convergence peak, indicating that the system is controlled by water vapor conveying; if excursion exists, the influence of thermal driving or sea temperature abnormity is explained, then the contribution proportion of specific humidity and wind speed to flux is decomposed, and the cause of ITCZ is obtained; according to the invention, through analysis and digital modeling of the rainfall data of the ITCZ area, the cause of the ITCZ is disclosed, and meteorological prediction of the ITCZ area is facilitated.

Description

TECHNICAL FIELD

[0001] The application is based on a tropical convergence zone feature analysis method based on precipitation data, and relates to the field of data analysis. BACKGROUND

[0002] The existing part of the analysis method related to the rainfall characteristics of the tropical convergence zone has the following shortcomings:

[0003] The data processing result ignores the regional characteristics: the existing part of the analysis method can be automatically run when analyzing precipitation data, but the automatic process ignores the abnormal physical phenomena of specific regions or special years while efficiently processing a large amount of data, and the automatic script used by the existing part of the method is mostly based on threshold judgment or fixed algorithm logic of half-width extraction, but these fixed conditions may not be applicable to all climate backgrounds.

[0004] Susceptible to noise: the existing part of the analysis method uses a multi-peak recognition algorithm when dealing with observation noise, insufficient spatial resolution, or asymmetric precipitation field; the accuracy of this algorithm depends on the smoothness of the precipitation profile and the stability of the peak value, but the actual precipitation field is often affected by local convection, tropical cyclones, and monsoon circulation disturbances, resulting in multiple secondary peaks, making it difficult to determine the physical attribution of the main peak and the secondary peak, and affecting the recognition result of the algorithm.

[0005] High risk of interpretation: the existing part of the analysis method uses The graph or time series graph is susceptible to smoothing parameters, and the differences in time resolution and accuracy of different data sources can cause false continuity or offset in the image; in addition, image interpretation often relies on visual impressions, lacks statistical significance test support, and reduces the credibility of the analysis result. SUMMARY

[0006] In view of the deficiencies of the prior art, the application aims to provide a tropical convergence zone feature analysis method based on precipitation data, which aims to solve the problem of rainfall analysis of the tropical convergence zone.

[0007] In order to achieve the above-mentioned purpose, the application is implemented by the following technical solution: the tropical convergence zone feature analysis method based on precipitation data comprises:

[0008] Obtaining the original data, and supplementing and filtering the missing values of the original data to obtain standardized data; extracting the core feature index of the ITCZ from the standardized data; predicting the future precipitation in the central latitude region according to the core feature index of the ITCZ;

[0009] The water vapor flux and divergence of the central latitude region are calculated to determine the corresponding relationship between precipitation and water vapor aggregation: if the peak value of precipitation coincides with the peak of flux convergence, it indicates that the precipitation in the target area is controlled by water vapor transport; if there is a deviation, it indicates that the precipitation in the target area is affected by thermal driving or sea surface temperature anomaly. The contribution proportion of specific humidity and wind speed to the flux is decomposed to obtain the cause of ITCZ;

[0010] The actual precipitation data of the central latitude region in the future is obtained and compared with the future precipitation; if there is no difference between the two, the cause of ITCZ is fed back; otherwise, the cause of ITCZ is reanalyzed.

[0011] Further, the steps of obtaining standardized data are as follows:

[0012] The grid size, time step and physical unit of all data in the original data are unified, and it is checked whether there are missing values or abnormal values in the original data;

[0013] If there are, the spatial interpolation and time linear interpolation method is used to correct the original data;

[0014] If not, it is not processed;

[0015] The corrected original data is smoothed and filtered;

[0016] Through Gaussian smoothing, different weight average of adjacent latitude points is carried out, and small scale noise in 1°-2° latitude range is suppressed;

[0017] Spectral domain filtering is used to remove short wave components with a wavelength less than 1000 km, and the seasonal and climate change rules of the original data are extracted to obtain standardized data;

[0018] According to the standardized data, three core feature indexes of ITCZ are extracted: central latitude, ITCZ width and maximum precipitation of ITCZ.

[0019] Further, the steps of extracting feature indexes are as follows:

[0020] Extract the central latitude:

[0021] Peak value method: find the latitude point corresponding to the maximum precipitation on the monthly latitudinal precipitation distribution curve, record it as the central latitude, which is suitable for single peak structure; record the specific latitude and time of each month's central latitude;

[0022] When there are double ITCZ phenomena, the centroid method is used to calculate the weighted average of the latitude of precipitation according to the peak ITCZ maximum precipitation or the symmetry principle:

[0023] Obtain the latitude range of the target area: la (sta) ~ la (end) ; longitude range: lo(sta) ~ lo(end) ;

[0024] In the same month, the latitude is la (sta) , and the longitude is lo (sta) ~ lo (end) The sum of the total precipitation of the corresponding area is pr (sta) ;

[0025] Similarly, the latitude is la (end) , and the longitude is lo (sta) ~ lo (end) The sum of the total precipitation of the corresponding area is pr (end) ;

[0026] Calculate the product of each latitude and the sum of the total precipitation at the same latitude, i.e. (la (sta) × pr (sta) ) to (la (end) × pr (end) ) ;

[0027] Calculate the sum of (la (sta) × pr (sta) ) to (la (end) × pr (end) ) as the dividend, and the sum of pr (sta) to pr (end) as the divisor to obtain the central latitude;

[0028] Record the specific latitude and time of the central latitude obtained by the centroid method each month;

[0029] Extract the ITCZ width: find the two latitude positions in the north-south direction on the monthly precipitation curve where the maximum precipitation of the ITCZ reaches half of the maximum value, and the difference between the two latitudes is the width of the ITCZ; Record the ITCZ width and time each month;

[0030] If the precipitation of each region in the target area is close, use the precipitation weighted variance method to calculate the second moment of the precipitation distribution of the target area to define the ITCZ width.

[0031] Further, the subsequent steps of the feature index extraction are as follows:

[0032] In the same month, obtain the latitude interval Δla and the longitude interval Δlo of each grid space in the standardized data of that month;

[0033] Obtain the central latitude lo (con) of that month; let the latitude be i and the longitude be j, and the precipitation of the region is pr(i,j) ;

[0034] Calculate the average value pr (ave) of the total precipitation in the target area:

[0035]

[0036] Calculate the second moment var:

[0037]

[0038] Calculate (equivalent) ITCZ ​​width W (lo) :

[0039]

[0040] The maximum precipitation value within the corresponding latitude range of the target area is taken as the maximum precipitation in the ITCZ.

[0041] After obtaining the three core characteristic indicators of ITCZ, time series analysis was performed on the standardized data corresponding to the central latitude region to predict the temporal pattern of precipitation in the central latitude region and precipitation in other regions.

[0042] Furthermore, the steps for predicting precipitation in the central latitude region are as follows:

[0043] Extract the maximum precipitation data corresponding to the central latitude region from the standard data in chronological order, and use it as the target data;

[0044] Using the autocorrelation function and the partial autocorrelation function, the difference order, autoregressive parameter and moving average parameter of the target data are calculated in sequence, as well as the seasonal difference order, seasonal autoregressive parameter, seasonal moving average parameter and seasonal cycle.

[0045] Calculate the autocorrelation coefficient and moving average coefficient of the central latitude region based on the difference parameter, autoregression parameter, and moving average parameter.

[0046] Calculate the seasonal autocorrelation coefficient and seasonal moving average coefficient of the central latitude region based on the seasonal difference parameter, seasonal autoregressive parameter, and seasonal moving average parameter.

[0047] Calculate the average value aX of the target data, and subtract aX from each data point in the target data in turn to obtain the white noise for the first time series prediction.

[0048] The lag factors and seasonal lag factors of the target data are calculated based on the difference order and seasonal difference order. The autocorrelation coefficient and moving average coefficient, along with the seasonal autocorrelation coefficient and seasonal moving average coefficient, are then substituted into the SARIMA model to calculate the precipitation (qpr) at the central latitude region after one seasonal cycle. (1) ;

[0049] QPR (1)Incorporate it into the existing target data as the target data for the first time series prediction; repeat qpr (1) The calculation process predicts the precipitation (qpr) in the central latitude region after the second seasonal cycle. (2) ;

[0050] Similarly, predict the subsequent precipitation in the central latitude region after other seasonal cycles.

[0051] Furthermore, the steps for predicting precipitation in other regions are as follows:

[0052] Based on the specific latitude value of the center latitude and the specific time of each center latitude, construct a latitude function φc(t) that represents the change of the latitude value of the center latitude with time using a sine or cosine function; where t represents time.

[0053] Calculate the standard deviation of all longitudes and latitudes in the target area to obtain σφ and σλ;

[0054] Let the maximum rainfall at time t be Ra. (t) ;

[0055] Let latitude be φ and longitude be λ; the range of values ​​for φ is: la (sta) ~la (end) The range of values ​​for λ is: lo (sta) ~lo (end) ;

[0056] Let P be the rainfall in the region with latitude φ and longitude λ at time t. (φ，λ，t) Construct a distribution model:

[0057]

[0058] The predicted rainfall is calculated based on the distribution model at different times, at different center latitudes, and in other regions affected by the maximum rainfall.

[0059] Furthermore, the steps to determine the correspondence between precipitation and water vapor aggregation are as follows:

[0060] Obtain specific humidity and wind field data for the target region each month;

[0061] The water vapor flux in the isobaric layer is weighted and summed to obtain the water vapor flux Q in the meridional and zonal directions.

[0062] Calculate water vapor flux divergence: Calculate the horizontal divergence of the integrated flux field.

[0063] The difference between precipitation (P) and evaporation (E) is calculated using the water vapor budget equation and taken as the effective precipitation.

[0064] Comparison in terms of spatial distribution and temporal variation With (P-E);

[0065] If the two are consistent, it indicates that precipitation is mainly driven by water vapor convergence, and precipitation migration is mainly controlled by changes in the low-level wind field and water vapor transport.

[0066] If there is a deviation, it indicates that the sea surface temperature or thermal convection center in the target area is changing asynchronously. The contribution ratios of specific humidity and wind speed to flux can be decomposed to further analyze the causes of ITCZ.

[0067] Furthermore, in comparison The steps for (P-E) are as follows:

[0068] Comparison in terms of spatial distribution:

[0069] At each time span, calculate (P-E) along the latitudinal direction. The distribution curves are used to find the latitude positions corresponding to the maximum values ​​in each curve, which are denoted as the latitude of the precipitation peak lap and the latitude of the water vapor convergence peak laq.

[0070] Calculate the absolute value of the difference between lap and laq;

[0071] If the absolute value is less than 1°, it indicates that precipitation in the target area is mainly formed by the transport of water vapor in the lower atmosphere.

[0072] If the offset is in the medium range, it indicates that the target area is affected by unstable precipitation and thermal conditions.

[0073] If the deviation exceeds 5° and remains stable over a long period, it indicates that thermal drive or sea surface temperature anomalies are the core factors driving precipitation in the target area.

[0074] Comparison over time:

[0075] Arrange the target region's monthly (P-E) data in chronological order. Organize into a time series;

[0076] At the same latitude, the cross-correlation function corr is used to calculate (P-E) and The curve of the maximum correlation coefficient of the corresponding time series versus time lag τ;

[0077] If the maximum correlation coefficient appears near τ = 0, it indicates that precipitation in the target area is formed by the transport of water vapor in the lower atmosphere.

[0078] If the maximum correlation coefficient appears in the region where τ > 0, it indicates that the precipitation in the target area is still affected by the transport of water vapor in the lower layer, but the magnitude of the influence is weakened.

[0079] If the maximum correlation coefficient appears in the region where τ < 0, it indicates that thermal driving or sea surface temperature anomalies are the core factors that dominate precipitation in the target area.

[0080] judge Whether the comparison results with (P-E) in terms of spatial distribution and temporal variation are consistent, and whether the comparison results show that precipitation in the target area is mainly formed by the transport of low-level water vapor;

[0081] If so, then do not process it;

[0082] If not, then the contribution ratios of specific humidity and wind speed to flux are decomposed to further analyze the causes of ITCZ.

[0083] Further, the steps to analyze the causes of ITCZ ​​are as follows:

[0084] By comparing the specific humidity (q) and horizontal wind speed (v) over the time dimension, the ground state values ​​of specific humidity and horizontal wind speed are obtained. and

[0085] Calculate the outliers q′ and v′; where,

[0086] In the vertical direction, the integral flux Qs is calculated using discrete numerical integration of the isobaric layer:

[0087] Where y represents the thickness of the cloud cover;

[0088] Expand the integral flux Qs into three terms:

[0089] This represents the change in water vapor flux caused by humidity anomalies under average wind conditions; This represents the change in water vapor flux caused by wind anomalies under average humidity.

[0090] Qs (3) =∫(q / ×v / d(y) represents the change in water vapor flux caused by simultaneous anomalies in humidity and wind field;

[0091] Calculate the divergence of the above three terms, and then perform area integration or latitude integration on the divergence of each term to obtain the time series corresponding to each term.

[0092] Normalize the time series corresponding to each item and calculate the relative contribution percentage of each item;

[0093] If Qs (1) If the contribution rate is the highest, it indicates that specific humidity is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area.

[0094] If Qs(2) If the contribution is the highest, it indicates that wind speed is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area.

[0095] If Qs (3) If the contribution ratio is the highest, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area.

[0096] If Qs (1) Qs (2) Qs (3) If at least two of the three factors have the highest contribution percentage, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area.

[0097] Compared with the prior art, the beneficial effects of the present invention are:

[0098] This invention enables quantitative analysis of the spatiotemporal evolution of the ITCZ: By performing spatial interpolation and time averaging on precipitation data, the comparability and continuity of the data are improved. Combined with Gaussian filtering and low-pass processing, the short-term impact of local weather disturbances is eliminated, making the extracted latitude-time profile smoother and its physical meaning clearer.

[0099] Automation is achievable: This invention forms a standardized process from data preprocessing and feature extraction to physical diagnosis, which facilitates batch calculation and automated execution. In actual operation, users or relevant technical personnel can use tools such as Python, MATLAB or R to write scripts to achieve automated processing from data reading, meridional averaging, peak identification to water vapor flux divergence analysis, avoiding the bias caused by human subjective judgment.

[0100] Achieving causal diagnosis: This invention introduces the water vapor budget equation into the diagnostic framework. By calculating the vertically integrated water vapor flux and its divergence, the contribution of water vapor convergence to precipitation formation can be directly quantified. By decomposing flux terms (such as humidity change terms and wind speed change terms), it is possible to identify whether the ITCZ ​​changes are mainly driven by "thermal effects" or "dynamic effects", providing a basis for climate simulation in the ITCZ ​​(Intertropical Convergence Zone) region. Attached Figure Description

[0101] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0102] Figure 1 This is a schematic diagram of the method of the present invention;

[0103] Figure 2 This is a schematic diagram of precipitation in the Intertropical Convergence Zone according to the present invention;

[0104] Figure 3 This is a schematic diagram of the central latitude region of the present invention. Detailed Implementation

[0105] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0106] Please see Figure 1 and Figure 2 Methods for analyzing the characteristics of the Intertropical Convergence Zone (ITCZ) based on precipitation data include:

[0107] Step S1: Obtain the raw data and perform missing value supplementation and filtering on the raw data to obtain standardized data; extract the three core characteristic indicators of the ITCZ ​​(Intertropical Convergence Zone): center latitude, ITCZ ​​width and maximum precipitation in the ITCZ; predict the future precipitation in the center latitude region based on the three core characteristic indicators of the ITCZ ​​(Intertropical Convergence Zone).

[0108] Data source selection, data preprocessing, spatial averaging and smoothing.

[0109] Global precipitation data (such as data provided by GPCP (Global Precipitation Climatology Project) or TRMM (Tropical Rainfall Measuring Mission) that includes the target area, has long-term time series, comprehensive spatial coverage, and has undergone quality control, are selected as the raw data for studying the precipitation characteristics of the ITCZ ​​(Intertropical Convergence Zone).

[0110] In this invention, "target area" refers to the area where precipitation analysis is performed using this invention (a method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data).

[0111] Along the longitude and latitude directions, unify the (spatial) grid size, time step and physical unit of all data in the original data, and then check whether there are missing or outlier values ​​in the original data.

[0112] The outlier detection method is the three-fold detection method: if the precipitation data of a certain region on a certain day is not within the range of "the monthly average precipitation data of the corresponding month on that day ± 3 times the standard deviation of the precipitation data of the corresponding month on that day", then the precipitation data is an outlier.

[0113] If missing or outlier values ​​exist, spatial interpolation and temporal linear interpolation methods are used to correct the original data.

[0114] If the value is missing (or outlier), no action is taken.

[0115] The corrected original data is then smoothed and filtered using Gaussian smoothing.

[0116] Gaussian smoothing is used to perform weighted averaging of adjacent latitude points with different weights, which suppresses small-scale noise and preserves large-scale structural features within the latitude range of 1° to 2°.

[0117] The specific steps for Gaussian smoothing are as follows:

[0118] Calculate the average precipitation value (or rainfall) for each grid space; take the weighted average of the precipitation value for each grid and the precipitation values ​​of its neighboring grids. The weights follow a Gaussian distribution, that is, the closer the adjacent grid points are, the higher their weights, and the farther away they are, the lower their weights.

[0119] Spectral domain filtering (such as wavelet or Fourier filtering) is used to remove short-wave components with wavelengths less than 1000km, extract seasonal and climate change patterns from the original data, and obtain standardized data.

[0120] The original data after Gaussian smoothing is treated as one-dimensional (discrete) data (i.e., only the magnitude of precipitation data is considered, and the location of precipitation data is not considered for the time being), and a Fast Fourier Transform (FFT) is performed on it to obtain the frequency domain result in the complex form of the one-dimensional (discrete) data; the frequency domain result contains real and imaginary parts;

[0121] The frequency domain result of each one-dimensional data is moduloed to obtain the amplitude of each frequency component; the amplitude is plotted as a function of frequency to obtain the spectral distribution of the one-dimensional data.

[0122] (In the frequency domain) construct a low-pass filter (usually using a Gaussian transfer function), multiply the spectral distribution of the one-dimensional data with the filter, and then perform an inverse fast Fourier transform (IFFT) to obtain precipitation data (mainly low-frequency components).

[0123] Standardized data is obtained by Z-score standardization of precipitation data (mainly low-frequency components);

[0124] Three core characteristic indicators of the ITCZ ​​(Intertropical Convergence Zone) were extracted from standardized data: center latitude, ITCZ ​​width, and maximum precipitation in the ITCZ.

[0125] The center latitude is determined using the peak method or the centroid method.

[0126] The width of the ITCZ, defined by the variance method, marks the zonal extent of the convective activity zone within the ITCZ ​​(Intertropical Convergence Zone).

[0127] The maximum precipitation in the ITCZ ​​is obtained by extracting the average or maximum precipitation in the ITCZ ​​(Intertropical Convergence Zone) region.

[0128] Extract center latitude:

[0129] Peak method: On the monthly latitudinal precipitation distribution curve, find the latitude point corresponding to the maximum precipitation, and record it as the center latitude. This method is suitable for single-peak structures.

[0130] Record the specific latitude and time of the center latitude each month;

[0131] When a double ITCZ ​​(Intertropical Convergence Zone) phenomenon exists (such as the bimodal distribution of the Pacific Ocean in spring and autumn), the centroid method is used to calculate the latitudinal weighted average of precipitation based on the maximum precipitation at the peak ITCZ ​​or the principle of symmetry.

[0132] Get the latitude range of the target area: la (sta) (Indicates the starting latitude of the target region in the Northern Hemisphere) ~la (end) (Indicates the ending latitude of the target region in the Southern (or Northern) Hemisphere);

[0133] Get the longitude range of the target area: lo (sta) (Indicates the starting longitude of the target region in the Eastern Hemisphere) ~lo (end) (Indicates the ending longitude of the target region in the Western (or Eastern) Hemisphere);

[0134] Calculate the latitude (la) within the target region in the same month. (sta) Longitude is lo (sta) ~lo (end) The sum of all precipitation in the corresponding region (pr) (sta) ;

[0135] Calculate the latitude (la) in the target region. (sta+1) Longitude is lo (sta) ~lo (end) The sum of all precipitation in the corresponding region (pr) (sta+1) ;

[0136] Similarly, calculate the latitude (la) in the target region. (end) Longitude is lo (sta) ~lo ( end ) The sum of all precipitation in the corresponding region (pr) (end) ;

[0137] Calculate the product of each latitude and the sum of all precipitation at the same latitude, i.e., (la (sta) ×pr (sta) (la) (sta+1) ×pr (sta+1) ) until (la (end) ×pr (end) );

[0138] Calculation (la) (sta) ×pr (sta) ) to (la(end) ×pr (end) Using the sum of 1 and 2 as the dividend, calculate pr. (sta) To pr (end) The sum of the terms is used as the divisor to obtain the central latitude;

[0139] Record the specific latitude and time of the monthly center latitude obtained using the centroid method;

[0140] Extract the ITCZ ​​width, which is the latitudinal extension range of the precipitation band (i.e., the range of influence of the central latitude);

[0141] The commonly used method for extracting the width of the ITCZ ​​is the "Full Width at Half Maximum" (FWHM) algorithm, which finds two latitudinal positions on the monthly precipitation curve where the maximum precipitation in the ITCZ ​​reaches half of the maximum value. The difference between the two latitudes is the width of the ITCZ ​​(Intertropical Convergence Zone).

[0142] Record the ITCZ ​​width and time each month;

[0143] If the precipitation amounts in different regions within the target area are similar (i.e., the precipitation curves fluctuate little and there are no obvious peaks or multiple peaks), then the second moment definition (equivalent) ITCZ ​​width of the precipitation distribution in the target area is calculated using the precipitation weighted variance method.

[0144] Within the same month, obtain the latitudinal interval Δla and longitude interval Δlo for each grid space in the standardized data for that month;

[0145] Get the center latitude lo of the month (con) Let pr(i,j) be the precipitation amount in the region with latitude i and longitude j.

[0146] Calculate the average precipitation (pr) of all precipitation within the target area. (ave) (Considering latitude and longitude):

[0147]

[0148] Calculate the second moment var:

[0149]

[0150] Calculate (equivalent) ITCZ ​​width W (lo) :

[0151]

[0152] The maximum precipitation in the ITCZ ​​(Intertropical Convergence Zone) can be defined either as the maximum precipitation value or as the average precipitation within the corresponding latitude range of the target area in the ITCZ.

[0153] After obtaining the three core characteristic indicators of the ITCZ ​​(Intertropical Convergence Zone) (i.e., center latitude, ITCZ ​​width, and maximum precipitation in the ITCZ), we conduct time series analysis on the standardized data corresponding to the center latitude region to predict the temporal pattern of precipitation in the center latitude region and precipitation in other regions.

[0154] Please see Figure 3 The steps for predicting the temporal patterns of precipitation in central latitude regions are as follows:

[0155] The central latitude region is defined as follows: with the central latitude as the midline, the horizontal range is defined by the latitude of the central latitude plus or minus half of wit, and the area within the horizontal range where the longitude of the maximum precipitation in the ITCZ ​​is twice Δlo.

[0156] Where wit represents the width of ITCZ;

[0157] Extract the maximum precipitation data corresponding to the central latitude region from the standard data in chronological order, and use it as the target data (without considering location information);

[0158] Using the autocorrelation function (ACF) and partial autocorrelation function (PACF), the difference order, autoregressive parameter, and moving average parameter of the target data are calculated sequentially, as well as the seasonal difference order, seasonal autoregressive parameter, seasonal moving average parameter, and seasonal cycle.

[0159] Calculate the autocorrelation coefficient and moving average coefficient of the central latitude region based on the difference parameter, autoregression parameter, and moving average parameter.

[0160] Calculate the seasonal autocorrelation coefficient and seasonal moving average coefficient of the central latitude region based on the seasonal difference parameter, seasonal autoregressive parameter, and seasonal moving average parameter.

[0161] Calculate the average value aX of the target data, and subtract aX from each data point in the target data in turn to obtain the white noise for the first time series prediction.

[0162] The lag factors and seasonal lag factors of the target data are calculated based on the difference order and seasonal difference order. The autocorrelation coefficient and moving average coefficient, along with the seasonal autocorrelation coefficient and seasonal moving average coefficient, are then substituted into the SARIMA model to calculate the precipitation (qpr) at the central latitude region after one seasonal cycle. (1) ;

[0163] QPR (1) Incorporate it into the existing target data as the target data for the first time series prediction; repeat qpr (1) The calculation process predicts the precipitation (qpr) in the central latitude region after the second seasonal cycle.(2) ;

[0164] Repeat qpr (2) The calculation steps are used to predict the subsequent precipitation in the central latitude region after other seasonal cycles;

[0165] Precipitation forecast for other regions:

[0166] Based on the specific latitude value of the center latitude and the specific time of each center latitude, construct a latitude function φc(t) that represents the change of the corresponding latitude value of the center latitude over time using a sine or cosine function; where t represents time (unit: month);

[0167] Calculate the standard deviation of all longitudes and latitudes in the target area to obtain σφ and σλ;

[0168] Let the maximum rainfall at time t be Ra. (t) ;

[0169] Let latitude be φ and longitude be λ; the range of values ​​for φ is: la (sta) ~la (end) The range of values ​​for λ is: lo (sta) ~lo (end) ;

[0170] Let P be the rainfall in the region with latitude φ and longitude λ at time t. (φ，λ，t) Construct a (two-dimensional Gaussian) distribution model:

[0171]

[0172] The predicted rainfall is calculated based on the (two-dimensional Gaussian) distribution model at different times and at different central latitudes (excluding the central latitude of the target area) after being affected by the maximum rainfall.

[0173] Step S2: Calculate the water vapor flux and divergence in the central latitude region (corresponding to the target region) to determine the correspondence between precipitation and water vapor convergence: if the precipitation peak coincides with the flux convergence peak, it indicates that the precipitation in the target region is controlled by water vapor transport; if there is a shift, it indicates that the precipitation in the target region is driven by thermal forces or affected by sea surface temperature anomalies. Then decompose the contribution ratio of specific humidity and wind speed to the flux to obtain the cause of the ITCZ ​​(Intertropical Convergence Zone).

[0174] Based on the water vapor budget equation, the formation of the ITCZ ​​(Intertropical Convergence Zone) is analyzed:

[0175] Calculate the water vapor flux and divergence in the central latitude region (corresponding to the target area) to determine the correspondence between precipitation and water vapor convergence:

[0176] If the peak precipitation coincides with the flux convergence peak, it indicates that the system is controlled by water vapor transport.

[0177] If there is a deviation, it indicates the influence of thermal drive or sea surface temperature anomaly. Further decomposition of the contribution ratio of specific humidity and wind speed to flux reveals the coupling effect of dynamic and thermal processes in the formation of ITCZ.

[0178] Obtain specific humidity and wind field data for the target region each month;

[0179] Calculate water vapor flux:

[0180] The water vapor flux in the isobaric layer is weighted and summed to obtain the water vapor flux Q in the meridional and zonal directions.

[0181] Calculate water vapor flux divergence: Calculate the horizontal divergence of the integrated flux field.

[0182] The difference between precipitation (P) and evaporation (E) is calculated using the water vapor budget equation and taken as the effective precipitation.

[0183] Comparison in terms of spatial distribution and temporal variation With (P-E);

[0184] If the two are consistent, it indicates that precipitation is mainly driven by water vapor convergence, and precipitation migration is mainly controlled by changes in the low-level wind field and water vapor transport.

[0185] If there is a deviation, it indicates that the sea surface temperature or the center of thermal convection in the target area is changing asynchronously. The contribution ratio of specific humidity and wind speed to flux can be decomposed to further analyze the causes of the ITCZ ​​(Intertropical Convergence Zone).

[0186] Comparison in terms of spatial distribution:

[0187] At each time span (e.g., month), calculate (P-E) along the latitudinal direction. The distribution curves are used to find the latitude positions corresponding to the maximum values ​​in each curve, which are denoted as the latitude of the precipitation peak lap and the latitude of the water vapor convergence peak laq.

[0188] Calculate the absolute value of the difference between lap and laq;

[0189] If the absolute value is less than 1°, it indicates that the target area is in a "moisture-dominated" state, meaning that precipitation in the target area is mainly formed by the transport of low-level water vapor.

[0190] If the offset is in the medium range (e.g., 1° to 5°), it indicates that the precipitation in the target area is affected by local convection development or thermal instability.

[0191] If the deviation exceeds 5° and remains stable over a long period, it indicates that precipitation in the target area is decoupled from water vapor convergence, and thermal drive or sea surface temperature anomaly is the core factor driving precipitation.

[0192] Comparison over time:

[0193] Arrange the target region's monthly (P-E) data in chronological order. Organize into a time series;

[0194] At the same latitude, the cross-correlation function corr is used to calculate (P-E) and The maximum correlation coefficient of the corresponding time series varies with ((P-E) and The curve corresponding to the time lag τ of the time series;

[0195] If the maximum correlation coefficient occurs near τ = 0 (i.e., approaching zero), the absolute value of the difference between the τ value corresponding to the maximum correlation coefficient and zero is within 10. -5 If the precipitation in the target area is within the range of water vapor convergence, it indicates that the precipitation is influenced by water vapor convergence and is a typical water vapor-driven system, meaning that the precipitation is mainly formed by the transport of water vapor in the lower atmosphere.

[0196] If the maximum correlation coefficient appears in the region where τ > 0, it indicates that water vapor convergence occurs before precipitation. The airflow in the target area rises and converges first, and then condenses after several hours or days, before precipitation. Precipitation is still affected by the transport of water vapor in the lower layer, but the influence is weakened.

[0197] If the maximum correlation coefficient appears in the region where τ < 0, it means that precipitation in the target area precedes the enhancement of water vapor convergence, indicating that thermal drive or sea surface temperature anomaly is the core factor dominating precipitation in the target area.

[0198] judge Whether the comparison results with (P-E) in terms of spatial distribution and temporal variation are consistent, and whether the comparison results show that precipitation in the target area is mainly formed by the transport of low-level water vapor;

[0199] If so, then do not process it;

[0200] If not, then the contribution ratios of specific humidity and wind speed to flux are decomposed to further analyze the causes of the ITCZ ​​(Intertropical Convergence Zone):

[0201] By comparing the specific humidity (q) and horizontal wind speed (v) over the time dimension, the ground state values ​​of specific humidity and horizontal wind speed are obtained. and

[0202] Calculate the outliers q′ and v′; where,

[0203] In the vertical direction, the integral flux Qs is (approximately) calculated using discrete numerical integration of the isobaric layer:

[0204] Where y represents the thickness of the cloud cover;

[0205] The discrete numerical integral of the isobaric layer is:

[0206] Among them, qy (k) vy represents the specific humidity of the k-th layer of atmosphere. (k) hy represents the horizontal wind speed in the k-th layer of the atmosphere. (k) This represents the thickness of the pressure layer of the k-th layer of the atmosphere;

[0207] g represents the acceleration due to gravity, and k represents the isobaric layer number (located above the target area).

[0208] Expand the integral flux Qs into three terms (actually four terms, the last term being the ground-state flux, which is not included in the final quantization):

[0209] This represents the change in water vapor flux caused by humidity anomalies under average wind conditions;

[0210] This represents the change in water vapor flux caused by wind anomalies under average humidity.

[0211] Qs (3) =∫(q / ×v / d(y) represents the change in water vapor flux caused by simultaneous anomalies in humidity and wind field;

[0212] This represents the water vapor flux under climatic mean conditions.

[0213] Calculate the divergence of the above three terms, and then perform area integration or latitude integration on the divergence of each term to obtain the time series corresponding to each term.

[0214] Normalize the time series corresponding to each item and calculate the relative contribution percentage of each item;

[0215] If Qs (1) If the contribution rate is the highest, it indicates that specific humidity is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area.

[0216] If Qs (2) If the contribution is the highest, it indicates that wind speed is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area.

[0217] If Qs (3) If the contribution ratio is the highest, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area.

[0218] If Qs (1) Qs (2) Qs (3)If at least two of the three factors have the highest contribution percentage, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area.

[0219] Step S3: Obtain the actual future precipitation data for the central latitude region and compare it with the future precipitation amount; if there is no difference, provide feedback (to the user or relevant technical personnel) on the cause of the ITCZ ​​(Intertropical Convergence Zone); otherwise, re-analyze the cause of the ITCZ ​​(Intertropical Convergence Zone).

[0220] The above formulas are all dimensionless calculations. The formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation. For example, there are weighting coefficients and proportional coefficients. The values ​​set are to quantify each parameter to obtain a specific value, which is convenient for subsequent comparison. The values ​​of the weighting coefficients and proportional coefficients are only required to not affect the proportional relationship between the parameters and the quantified values.

[0221] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data, characterized in that, The method includes: Obtain the raw data, and perform missing value supplementation and filtering on the raw data to obtain standardized data; extract the core feature indicators of ITCZ ​​from the standardized data; predict the future precipitation in the central latitude region based on the core feature indicators of ITCZ. Calculate the water vapor flux and divergence in the central latitude region to determine the correspondence between precipitation and water vapor convergence: if the precipitation peak coincides with the flux convergence peak, it is determined that the precipitation in the target area is controlled by water vapor transport; if there is a shift, it is determined that the precipitation in the target area is driven by thermal forces or affected by sea surface temperature anomalies. Then, the contribution ratios of specific humidity and wind speed to the flux are decomposed to obtain the cause of the ITCZ. Obtain the actual future precipitation data for the central latitude region and compare it with the future precipitation amount; if there is no difference, provide feedback on the cause of the ITCZ; otherwise, re-analyze the cause of the ITCZ.

2. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 1, characterized in that, The steps to obtain standardized data are as follows: Unify the grid size, time step and physical unit of all data in the original data, and check whether there are missing or outlier values ​​in the original data; If it exists, then spatial interpolation and temporal linear interpolation methods are used to correct the original data; If it does not exist, then do not process it; The corrected original data is then smoothed and filtered. Gaussian smoothing is used to perform weighted averaging of adjacent latitude points with different weights, which suppresses small-scale noise in the latitude range of 1° to 2°. Spectral domain filtering was used to remove short-wavelength components with wavelengths less than 1000km, extracting seasonal and climate change patterns from the original data to obtain standardized data. Based on standardized data, three core characteristic indicators of the ITCZ ​​were extracted: center latitude, ITCZ ​​width, and maximum precipitation in the ITCZ.

3. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 2, characterized in that, The steps for extracting feature indicators are as follows: Extract center latitude: Peak method: On the monthly latitudinal precipitation distribution curve, find the latitude point corresponding to the maximum precipitation, and record it as the center latitude. Record the specific latitude and time of the center latitude for each month. When the double ITCZ ​​phenomenon exists, the centroid method is used to calculate the latitudinal weighted average of precipitation: Get the latitude range of the target area: la (sta) ~la (end) Longitude range: lo (sta) ~lo(end); In the same month, calculate latitude la. (sta) Longitude is lo (sta) ~lo (end) The sum of all precipitation in the corresponding region (pr) (sta) ; Similarly, calculate latitude la. (end) Longitude is lo (sta) ~lo (end) The sum of all precipitation in the corresponding region (pr) (end) ; Calculate the product of each latitude and the sum of all precipitation at the same latitude, then use pr (sta) To pr (end) The sum of the terms is used as the divisor to obtain the central latitude; Record the specific latitude and time of the monthly center latitude obtained using the centroid method; Extracting ITCZ ​​width: Find the two latitudinal positions on the monthly precipitation curve where the maximum precipitation in the ITCZ ​​reaches half of the maximum value in the north-south direction. The difference between the two latitudes is the width of the ITCZ. Record the ITCZ ​​width and time for each month. If the precipitation amounts in different regions within the target area are similar, the second moment of the precipitation distribution in the target area is calculated using the precipitation weighted variance method to define the width of the ITCZ.

4. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 3, characterized in that, The subsequent steps for feature extraction are as follows: Within the same month, obtain the latitudinal interval Δla and longitude interval Δlo for each grid space in the standardized data for that month; Get the center latitude lo of the month (con) Let pr(i,j) be the precipitation amount in the region with latitude i and longitude j. Calculate the average precipitation (pr) of all precipitation within the target area. (ave) : Calculate the second moment var: Calculate the width W of ITCZ (lo) : The maximum precipitation value within the corresponding latitude range of the target area is taken as the maximum precipitation in the ITCZ. We perform time-series analysis on standardized data corresponding to the central latitude region to predict the temporal pattern of precipitation in the central latitude region and precipitation in other regions.

5. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 4, characterized in that, The steps for predicting precipitation in a central latitude region are as follows: Extract the maximum precipitation data corresponding to the central latitude region from the standard data in chronological order, and use it as the target data; Using the autocorrelation function and the partial autocorrelation function, the difference order, autoregressive parameter and moving average parameter of the target data are calculated in sequence, as well as the seasonal difference order, seasonal autoregressive parameter, seasonal moving average parameter and seasonal cycle. Calculate the autocorrelation coefficient and moving average coefficient of the central latitude region based on the difference parameter, autoregression parameter, and moving average parameter. Calculate the seasonal autocorrelation coefficient and seasonal moving average coefficient of the central latitude region based on the seasonal difference parameter, seasonal autoregressive parameter, and seasonal moving average parameter. Calculate the average value aX of the target data, and subtract aX from each data point in the target data in turn to obtain the white noise for the first time series prediction. The lag factors and seasonal lag factors of the target data are calculated based on the difference order and seasonal difference order. The autocorrelation coefficient and moving average coefficient, along with the seasonal autocorrelation coefficient and seasonal moving average coefficient, are then substituted into the SARIMA model to calculate the precipitation (qpr) at the central latitude region after one seasonal cycle. (1) ; QPR (1) Incorporate it into the existing target data as the target data for the first time series prediction; repeat qpr (1) The calculation process predicts the precipitation (qpr) in the central latitude region after the second seasonal cycle. (2) ; Similarly, predict the subsequent precipitation in the central latitude region after other seasonal cycles.

6. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 4, characterized in that, The steps for predicting precipitation in other regions are as follows: Based on the specific latitude value of the center latitude and the specific time of each center latitude, construct a latitude function φc(t) that represents the change of the latitude value of the center latitude with time using a sine or cosine function; where t represents time. Calculate the standard deviation of all longitudes and latitudes in the target area to obtain σφ and σλ; Let the maximum rainfall at time t be Ra. (t) ; Let latitude be φ and longitude be λ; the range of values ​​for φ is: la (sta) ~la (end) The range of values ​​for λ is: lo (sta) ~lo (end) ; Let P be the rainfall in the region with latitude φ and longitude λ at time t. (φ，λ，t) Construct a distribution model: The predicted rainfall is calculated based on the distribution model at different times, at different center latitudes, and in other regions affected by the maximum rainfall.

7. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 1, characterized in that, The steps to determine the correspondence between precipitation and water vapor aggregation are as follows: Obtain specific humidity and wind field data for the target region each month; The water vapor flux in the isobaric layer is weighted and summed to obtain the water vapor flux Q in the meridional and zonal directions. Calculate horizontal divergence The difference between precipitation (P) and evaporation (E) is calculated using the water vapor budget equation. Comparison in terms of spatial distribution and temporal variation With (P-E); If the two are consistent, it indicates that precipitation is mainly driven by water vapor convergence, and precipitation migration is mainly controlled by changes in the low-level wind field and water vapor transport. If there is a deviation, it indicates that the sea surface temperature or thermal convection center in the target area is changing asynchronously. The contribution ratios of specific humidity and wind speed to flux can be decomposed to further analyze the causes of ITCZ.

8. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 7, characterized in that, contrast The steps for (P-E) are as follows: Comparison in terms of spatial distribution: At each time span, calculate (P-E) along the latitudinal direction. The distribution curves are used to find the latitude positions corresponding to the maximum values ​​in each curve, which are denoted as the latitude of the precipitation peak lap and the latitude of the water vapor convergence peak laq. Calculate the absolute value of the difference between lap and laq; If the absolute value is less than 1°, it indicates that precipitation in the target area is mainly formed by the transport of water vapor in the lower atmosphere. If the offset is in the medium range, it indicates that the target area is affected by unstable precipitation and thermal conditions. If the deviation exceeds 5° and remains stable over a long period, it indicates that thermal drive or sea surface temperature anomalies are the core factors driving precipitation in the target area. Comparison over time: Arrange the target region's monthly (P-E) data in chronological order. Organize into a time series; At the same latitude, the cross-correlation function corr is used to calculate (P-E) and The curve of the maximum correlation coefficient of the corresponding time series versus time lag τ; If the maximum correlation coefficient appears near τ = 0, it indicates that precipitation in the target area is formed by the transport of water vapor in the lower atmosphere. If the maximum correlation coefficient appears in the region where τ > 0, it indicates that the precipitation in the target area is still affected by the transport of water vapor in the lower layer, but the magnitude of the influence is weakened. If the maximum correlation coefficient appears in the region where τ < 0, it indicates that thermal driving or sea surface temperature anomalies are the core factors that dominate precipitation in the target area. judge Whether the comparison results with (P-E) in terms of spatial distribution and temporal variation are consistent, and whether the comparison results show that precipitation in the target area is mainly formed by the transport of low-level water vapor; If so, then do not process it; If not, then the contribution ratios of specific humidity and wind speed to flux are decomposed to further analyze the causes of ITCZ.

9. The method for analyzing the characteristics of the Intertropical Convergence Zone based on precipitation data according to claim 8, characterized in that, The steps to analyze the causes of ITCZ ​​are as follows: By comparing the specific humidity (q) and horizontal wind speed (v) over the time dimension, the ground state values ​​of specific humidity and horizontal wind speed are obtained. and Calculate the outliers q′ and v′; where, In the vertical direction, the integral flux Qs is calculated using discrete numerical integration of the isobaric layer: Where y represents the thickness of the cloud cover; Expand the integral flux Qs into three terms: This represents the change in water vapor flux caused by humidity anomalies under average wind conditions; This represents the change in water vapor flux caused by wind anomalies under average humidity. Qs (3) =∫(q / ×v / d(y) represents the change in water vapor flux caused by simultaneous anomalies in humidity and wind field; Calculate the divergence of the above three terms, and then perform area integration or latitude integration on the divergence of each term to obtain the time series corresponding to each term. Normalize the time series corresponding to each item and calculate the relative contribution percentage of each item; If Qs (1) If the contribution rate is the highest, it indicates that specific humidity is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area. If Qs (2) If the contribution is the highest, it indicates that wind speed is the main influencing factor of thermal drive or sea surface temperature anomaly in the target area. If Qs (3) If the contribution ratio is the highest, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area. If Qs (1) Qs (2) Qs (3) If at least two of the three factors have the highest contribution percentage, it indicates that specific humidity and wind speed are the main influencing factors of thermal drive or sea surface temperature anomalies in the target area.

Images