A meteorological data extension method considering mean-variance consistency

Abstract

The invention discloses a meteorological data extension method considering mean variance consistency, S1, grouping stations, and obtaining the initial data sequence of the reference station and the interpolation station; S2, obtaining the simple interpolation sequence of the interpolation station; S3, obtaining the correlation calculation sequence of the interpolation station and the reference station; S4, calculating the correlation coefficient between stations, and determining the interpolation reference station list of the interpolation station; S5, interpolating the interpolation station, obtaining the preliminary interpolation sequence of the interpolation station; S6, calculating the interpolation correction parameter based on the simple interpolation sequence and the preliminary interpolation sequence of the interpolation station; S7, correcting the preliminary interpolation sequence of the interpolation station, obtaining the corrected interpolation sequence of the interpolation station; S8, merging the corrected interpolation sequence and the simple interpolation sequence of the interpolation station, and obtaining the final interpolation station completed data sequence. The advantages are: making the mean and variance of the interpolation extension data series and the original data series at the same level or having a more reasonable change trend, avoiding obvious mutation or transition phenomenon, and improving the accuracy of interpolation data.

Description

Technical Field

[0001] The invention relates to the technical field of meteorological data processing, and in particular to a meteorological data extension method considering mean variance consistency. Background Art

[0002] Affected by many factors such as inconsistent meteorological station construction years or difficulties in collecting data, the actual meteorological data series used are of different lengths. In meteorological or hydrological analysis and research, data series of the same length are generally required, and stations with long data series (reference stations) are usually used to interpolate data from stations with short data series (interpolation stations). However, since the change patterns and trends of the reference station and the interpolation station are not completely consistent, there is a statistical inconsistency between the values ​​of the interpolation sequence obtained based on the reference station and the original sequence of the interpolation station, which introduces additional errors in the interpolation process. In special cases, there is even an obvious mean transition phenomenon, which affects the accuracy of the analysis and research conclusions. Summary of the invention

[0003] The object of the present invention is to provide a meteorological data extension method taking into account mean variance consistency, thereby solving the above-mentioned problems existing in the prior art.

[0004] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0005] A meteorological data extension method considering mean-variance consistency includes the following steps:

[0006] S1. According to the completeness of the data sequence, the collected meteorological station data are divided into the initial data sequence of the reference station with complete sequence and the initial data sequence of the interpolation station with missing data;

[0007] S2, performing simple interpolation according to the data missing in the initial data sequence of the interpolation station to obtain the simple interpolation sequence y of the interpolation station;

[0008] S3, using the monthly scale data sequence generated by the daily data of the interpolation station and the reference station, or directly using the corresponding annual or monthly scale sequence as the interpolation station correlation calculation sequence and the reference station correlation calculation sequence respectively;

[0009] S4, calculating the correlation coefficient between the interpolation station correlation calculation sequence and the reference station correlation calculation sequence, and determining the interpolation reference station list used by each interpolation station according to the magnitude relationship between the correlation coefficient and a preset threshold;

[0010] S5, performing interpolation calculation on the interpolation station based on the initial data sequence of the reference station in the interpolation reference station list of the interpolation station to obtain a preliminary interpolation sequence x of the interpolation station;

[0011] S6, calculating corresponding interpolation correction parameters according to the variance and mean of the sequence having data of the simple interpolation sequence y of the interpolation station and the preliminary interpolation sequence x of the interpolation station;

[0012] S7, correcting the preliminary interpolation sequence x of the interpolation station according to the interpolation correction parameter to obtain the corrected interpolation sequence y' of the interpolation station;

[0013] S8. Replace the missing data in the simple interpolation sequence y of the interpolation station according to the corrected interpolation sequence y' of the interpolation station to obtain the final interpolation station completed data sequence.

[0014] Preferably, the interpolation station initial data sequence includes a data part subsequence and a missing data part subsequence; the missing data part needs to be completed using the above steps in the end;

[0015] The meteorological data included in the initial data sequence are precipitation, temperature, sunshine, wind speed, and relative humidity; the time scale of the meteorological data is year, month, and day.

[0016] Preferably, step S2 specifically includes, when there are 1 to 2 missing data in the initial data sequence of the interpolation station, interpolating the precipitation data with 0 and other meteorological data with linear interpolation to obtain a simple interpolation sequence y of the interpolation station;

[0017] The interpolation station simple interpolation sequence y includes an interpolation station subsequence y0 with data and an interpolation station subsequence y1 without data.

[0018] Preferably, step S3 specifically includes, when interpolating daily data, firstly performing statistics on the initial data sequence on a monthly basis to obtain monthly scale data, and setting no monthly data for months with missing data exceeding 7 days, and using the monthly scale data sequences generated by the daily data of the interpolation station and the reference station as the correlation calculation sequence of the interpolation station and the correlation calculation sequence of the reference station respectively;

[0019] When interpolating at the annual and monthly scales, the corresponding annual or monthly scale sequences of the interpolation station and the reference station are directly used as the interpolation station correlation calculation sequence and the reference station correlation calculation sequence, respectively.

[0020] Preferably, step S4 specifically comprises calculating the correlation coefficient between the interpolation station and the reference station data sequence, and selecting the corresponding reference station as the reference station used for interpolation according to the correlation coefficient, so as to determine the interpolation reference station list used;

[0021] When calculating the correlation between the interpolation station and the reference station, the subsequence with data in the correlation calculation sequence of the two stations is used, and for a certain interpolation station, it is necessary to establish a relationship with all the reference stations one by one, and all available reference stations are used as reference stations for interpolation and added to the interpolation reference station list;

[0022] The selection criteria of the available reference stations are: select the reference stations with correlation coefficients greater than a preset threshold as the reference stations used for interpolation; if the correlation coefficients of all reference stations are less than the preset threshold, select the closest reference station as the reference station used for interpolation;

[0023] The calculation formula of the correlation coefficient is:

[0024]

[0025] Among them, r is the correlation coefficient; cov() is the covariance of the corresponding sequence; var() is the variance of the corresponding sequence; A is the reference station correlation calculation sequence; B is the interpolation station correlation calculation sequence.

[0026] Preferably, step S5 specifically comprises: using an interpolation method to interpolate the full series of data of each interpolation station based on the initial data sequence of the reference station in the interpolation reference station list, to obtain a preliminary interpolation sequence x of the interpolation station, wherein the data in the series are all calculated from the reference station data;

[0027] The interpolation methods used are Thiessen polygon method, inverse distance square method, Kriging method, and spline method; according to the requirements of the relevant algorithms, some or all of the reference stations in the interpolation reference station list are used for interpolation.

[0028] Preferably, step S6 specifically includes selecting sequences having data of both the simple interpolation sequence y of the interpolation station and the preliminary interpolation sequence x of the interpolation station, grouping them by month, and calculating the interpolation correction parameters respectively. The calculation formula is:

[0029]

[0030] b i =avg(y 0,i )-a i *avg(x 0,i )

[0031] Among them, y 0,i is the sequence consisting of the i-th month of the data subsequence y0 at the interpolation station; 0,i is the sequence consisting of the i-th month in the preliminary interpolation sequence x of the interpolation station and the data subsequence y0 of the interpolation station; a i , b i They are the monthly interpolation correction parameters respectively; var() is the variance of the corresponding sequence; avg() is the mean of the corresponding sequence; if the time scale of meteorological data is month and day, it is calculated by month, i=1~12; if the time scale is year, it is calculated by year and no longer subdivided into months, i=1.

[0032] Preferably, step S7 specifically comprises using the interpolation correction parameter to correct the values ​​in the preliminary interpolation sequence x of the interpolation station that are consistent with the time series of the subsequence y1 of the interpolation station without data, and obtaining the interpolation station corrected interpolation sequence y', and the calculation formula is:

[0033] y′=a i x1+b i

[0034] Among them, y′ is the corrected interpolation sequence of the interpolation station, the time series corresponds to the subsequence y1 of the interpolation station without data, and the data scale is year, month or day; x1 is the part of the preliminary interpolation sequence x of the interpolation station whose time series is consistent with the subsequence y1 of the interpolation station without data, and the data scale is the same as y ′ Consistent.

[0035] Preferably, step S8 specifically includes merging the interpolation station corrected interpolation sequence y' and the interpolation station existing data subsequence y0 to obtain a final interpolation station completed data sequence.

[0036] The beneficial effects of the present invention are as follows: the method of the present invention solves the problem of statistical inconsistency between the interpolated extended data series and the original data series of the short series meteorological stations by correcting the data after traditional interpolation, so that the mean and variance of the interpolated extended data series are at the same level as the mean and variance of the original data series or have a more reasonable change trend, avoiding obvious mutation or transition phenomena, and improving the accuracy of the interpolated data. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of a data extension method in an embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of example data of interpolation and extension in an embodiment of the present invention;

[0039] Figure 3 is a schematic diagram of the correlation between reference sites and interpolation sites in an embodiment of the present invention;

[0040] Figure 4 2 is a comparison diagram of interpolation data in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation methods described herein are only used to explain the present invention and are not used to limit the present invention.

[0042] like Figure 1As shown, in this embodiment, a method for extending meteorological data considering mean-variance consistency is provided. This method can make full use of the information of existing data and correct the data series after traditional interpolation, so that the mean and variance of the interpolated and extended data series are at the same level as the mean and variance of the original data series or have a more reasonable change trend, avoiding obvious mutations or transitions, and improving the accuracy of the interpolated data. The method includes the following parts:

[0043] 1. Divide meteorological stations and obtain the initial data sequence of reference stations and interpolation stations

[0044] According to the integrity of the data sequence, the collected meteorological station data are divided into two groups, one group is a complete sequence, called the reference station initial data sequence, and the other group is a missing data, called the interpolation station initial data sequence. The interpolation station initial data sequence includes a subsequence with data and a subsequence with missing data, wherein the missing data part eventually needs to be completed using the method of the present invention. The meteorological data contained in the initial data sequence are precipitation, temperature, sunshine, wind speed, relative humidity and other data, wherein -999 is used as a no data identifier. The time scale of meteorological data can be year, month, or day.

[0045] 2. Simple interpolation of interpolation stations, obtaining simple interpolation sequence of interpolation stations

[0046] According to the data missing in the initial data sequence of the interpolation station, simple interpolation is performed to obtain the simple interpolation sequence y of the interpolation station.

[0047] Specifically, for the case where 1 to 2 data are missing in the initial data sequence of the interpolation station (e.g., there are values ​​at positions 1 and 4, but values ​​are missing at positions 2 and 3; or there are values ​​at positions 1 and 3, but values ​​are missing at position 2), simple interpolation is performed, where precipitation data is 0, and other meteorological data are interpolated using linear interpolation to obtain the simple interpolation sequence y of the interpolation station. The simple interpolation sequence y of the interpolation station includes the subsequence y0 of the interpolation station with data and the subsequence y1 of the interpolation station without data.

[0048] 3. Obtaining the correlation calculation sequence of the interpolation station and the correlation calculation sequence of the reference station

[0049] The monthly scale data series generated by the daily data of the interpolation station and the reference station, or the corresponding annual or monthly scale series are directly used as the correlation calculation series of the interpolation station and the correlation calculation series of the reference station, respectively.

[0050] Specifically: when interpolating daily data, it is necessary to first count the initial data sequence by month to obtain monthly scale data, and assume that there is no monthly data for months with more than 7 days of missing data (using -999 instead), and use the monthly scale data sequence generated by the daily data of the interpolation station and the reference station as the correlation calculation sequence of the interpolation station and the correlation calculation sequence of the reference station respectively;

[0051] When interpolating at the annual and monthly scales, no statistics are performed, and the corresponding annual or monthly scale sequences of the interpolation station and the reference station are directly used as the interpolation station correlation calculation sequence and the reference station correlation calculation sequence, respectively.

[0052] 4. Calculate the correlation coefficient between the interpolation station and the reference station, and select the interpolation station interpolation reference station list

[0053] The correlation coefficient between the interpolation station correlation calculation sequence and the reference station correlation calculation sequence is calculated, and the interpolation reference station list used by each interpolation station is determined according to the size relationship between the correlation coefficient and a preset threshold.

[0054] Specifically: calculate the correlation coefficient between the interpolation station and the reference station data series, select the corresponding reference station as the reference station used for interpolation according to the correlation coefficient, so as to determine the interpolation reference station list used;

[0055] When calculating the correlation between the interpolation station and the reference station, the subsequence with data in the correlation calculation sequence of the two stations is used, and for a certain interpolation station, it is necessary to establish a relationship with all the reference stations one by one, and all available reference stations are used as reference stations for interpolation and added to the interpolation reference station list;

[0056] The selection criteria for the available reference stations are: select the reference stations with correlation coefficients greater than the preset threshold as the reference stations used for interpolation; if the correlation coefficients of all reference stations are less than the preset threshold (taken as 0.6), select the closest reference station as the reference station used for interpolation;

[0057] The calculation formula of the correlation coefficient is:

[0058]

[0059] Among them, r is the correlation coefficient; cov() is the covariance of the corresponding sequence; var() is the variance of the corresponding sequence; A is the reference station correlation calculation sequence; B is the interpolation station correlation calculation sequence.

[0060] 5. Use the interpolation method to interpolate the interpolation station length series to obtain the preliminary interpolation sequence of the interpolation station

[0061] Based on the initial data sequence of the reference station in the interpolation reference station list of the interpolation station, interpolation calculation is performed on the interpolation station to obtain a preliminary interpolation sequence x of the interpolation station.

[0062] Specifically: an interpolation method is used to interpolate the full series of data of each interpolation station based on the initial data sequence of the reference station in the interpolation reference station list to obtain a preliminary interpolation sequence x of the interpolation station. The data in this series are all calculated from the reference station data.

[0063] The interpolation methods used are Thiessen polygon method, inverse distance square method, Kriging method, spline method and other methods; according to the requirements of relevant algorithms, some or all reference stations in the interpolation reference station list are used for interpolation.

[0064] 6. Obtain interpolation correction parameters based on the simple interpolation sequence of the interpolation station and the preliminary interpolation sequence of the interpolation station

[0065] According to the variance and mean of the sequence for which both the simple interpolation sequence y at the interpolation station and the preliminary interpolation sequence x at the interpolation station have data, the corresponding interpolation correction parameters are calculated.

[0066] Specifically: select the sequences with data in both the simple interpolation sequence y of the interpolation station and the preliminary interpolation sequence x of the interpolation station (i.e., the interpolation station has a data subsequence y0 and a preliminary interpolation sequence x0 of the interpolation station extracted with the same series length), and group them by month, and calculate the interpolation correction parameters respectively. The calculation formula is:

[0067]

[0068] b i =avg(y 0,i )-a i *avg(x 0,i )

[0069] Among them, y 0,i is the sequence consisting of the i-th month of the data subsequence y0 at the interpolation station; 0,i is the sequence consisting of the i-th month in the preliminary interpolation sequence x of the interpolation station and the data subsequence y0 of the interpolation station; a i , b i They are the monthly interpolation correction parameters respectively; var() is the variance of the corresponding sequence; avg() is the mean of the corresponding sequence; if the time scale of meteorological data is month and day, it is calculated by month, i=1~12; if the time scale is year, it is calculated by year and no longer subdivided into months, i=1.

[0070] 7. Correct the preliminary interpolation sequence of the interpolation station according to the interpolation correction parameters to obtain the corrected interpolation sequence of the interpolation station

[0071] The preliminary interpolation sequence x of the interpolation station is corrected according to the interpolation correction parameter to obtain the corrected interpolation sequence y' of the interpolation station.

[0072] Specifically: using the interpolation correction parameters, the values ​​in the preliminary interpolation sequence x of the interpolation station that are consistent with the time series of the subsequence y1 of the interpolation station without data are corrected to obtain the interpolation station corrected interpolation sequence y'. The calculation formula is:

[0073] y′=a i x1+b i

[0074] Among them, y′ is the corrected interpolation sequence of the interpolation station, the time series corresponds to the subsequence y1 of the interpolation station with no data, and the data scale is year, month or day; x1 is the part of the preliminary interpolation sequence x of the interpolation station whose time series is consistent with the subsequence y1 of the interpolation station with no data, and the data scale is consistent with y′.

[0075] 8. Interpolation station simple interpolation sequence completion

[0076] The missing data in the simple interpolation sequence y of the interpolation station is replaced according to the corrected interpolation sequence y' of the interpolation station to obtain the final interpolation station completed data sequence.

[0077] Specifically: the interpolation station corrected interpolation sequence y' and the interpolation station existing data subsequence y0 are merged to obtain the final interpolation station completed data sequence.

[0078] In this embodiment, the Thiessen polygon method is used to interpolate the site data. The interpolation results are shown in Figures 2 to 4 . Figure 2 The relevant data with interpolation extension are shown in (assuming annual data). Figure 2 The relevant sequences in are specifically: Full series of reference stations: reference stations obtained by screening according to correlation > 0.6 (the Thiessen polygon method takes the nearest station as the reference station, so there is only one reference station). Simple interpolation sequence y of interpolation stations: a sequence after simple interpolation of stations with missing data, and the data at position -999 needs to be completed. Preliminary interpolation sequence x of interpolation stations: an interpolation sequence calculated in step 5. Since the Thiessen polygon method is used for interpolation in this embodiment, the series value is equal to the full series of reference station data. Corrected interpolation sequence y' of interpolation stations: a corrected interpolation sequence calculated in step 7. Corrected and completed series of interpolation stations: a sequence obtained by splicing the corrected interpolation sequence y' of interpolation stations and the simple interpolation sequence y of interpolation stations. Uncorrected and completed series of interpolation stations: a sequence obtained by splicing the preliminary interpolation sequence x of interpolation stations and the simple interpolation sequence y of interpolation stations, used for comparison effects. Figure 3 The correlation between the reference and interpolated stations is shown in Figure 2. Assuming the correlation is greater than 0.6, the Figure 2 The reference sites in serve as interpolation benchmarks. Figure 4The middle grey dotted line is the uncorrected data series supplemented by the Thiessen polygon method, and it can be seen that there is an obvious data mutation. The Thiessen polygon method uses the data of the nearest station as the data of the interpolated station, that is, the interpolated data is the original station data. The corrected result (black solid line) of the present invention is statistically relatively consistent with the original data (triangle) without mutation.

[0079] In this embodiment, it is assumed that the change trends of the mean and variance of the preliminary interpolation sequence x of the interpolation station in the fifth part are consistent. Due to the inconsistency of the data change rules of the reference station and the interpolation station, the interpolation station data subsequence y0 and the corresponding time series x0 in the preliminary interpolation sequence x of the interpolation station with the same series length have different means and variances, which is also the reason for the sudden change or transition in meteorological data statistics. In the present invention, according to the interpolation station data subsequence y0, the mean and variance of the corresponding time series x0 in the preliminary interpolation sequence x of the interpolation station are corrected to the y0 level, and the interpolation correction coefficient of the corresponding month / year is obtained, and the remaining sequence x1 in the preliminary interpolation sequence of the interpolation station is corrected to obtain the corrected interpolation sequence y' used at last. This method can reduce the interpolation error to a certain extent and avoid the sudden change of the meteorological data sequence after interpolation and the initial sequence of the interpolation station.

[0080] By adopting the above technical solution disclosed in the present invention, the following beneficial effects are obtained:

[0081] The present invention provides a meteorological data extension method taking into account mean-variance consistency, and solves the problem of statistical inconsistency between the interpolated extended data series and the original data series of short-series meteorological stations by correcting the data after traditional interpolation, so that the mean and variance of the interpolated extended data series are at the same level as the mean and variance of the original data series or have a more reasonable change trend, avoiding obvious mutation or transition phenomena, and improving the accuracy of the interpolated data.

[0082] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be considered as the scope of protection of the present invention.

Claims

1. A meteorological data extension method considering mean-variance consistency, characterized by: The following steps are included: S1. According to the completeness of the data sequence, the collected meteorological station data are divided into the initial data sequence of the reference station with complete sequence and the initial data sequence of the interpolation station with missing data; S2, performing simple interpolation according to the data missing in the initial data sequence of the interpolation station to obtain the simple interpolation sequence y of the interpolation station; S3, using the monthly scale data sequence generated by the daily data of the interpolation station and the reference station, or directly using the corresponding annual or monthly scale sequence as the interpolation station correlation calculation sequence and the reference station correlation calculation sequence respectively; S4, calculating the correlation coefficient between the interpolation station correlation calculation sequence and the reference station correlation calculation sequence, and determining the interpolation reference station list used by each interpolation station according to the magnitude relationship between the correlation coefficient and a preset threshold; S5, performing interpolation calculation on the interpolation station based on the initial data sequence of the reference station in the interpolation reference station list of the interpolation station to obtain a preliminary interpolation sequence x of the interpolation station; S6, calculating corresponding interpolation correction parameters according to the variance and mean of the sequence having data of the simple interpolation sequence y of the interpolation station and the preliminary interpolation sequence x of the interpolation station; Step S6 specifically includes selecting sequences that have data for both the simple interpolation sequence y of the interpolation station and the preliminary interpolation sequence x of the interpolation station, grouping them by month, and calculating interpolation correction parameters respectively. The calculation formula is: b i =avg(y 0,i )-a i *avg(x 0,i ) Among them, y 0,i is the sequence consisting of the i-th month of the data subsequence y0 at the interpolation station; 0,i is the sequence consisting of the i-th month in the preliminary interpolation sequence x of the interpolation station and the data subsequence y0 of the interpolation station; a i , b i are the monthly interpolation correction parameters respectively; var() is the variance of the corresponding sequence; avg() is the mean of the corresponding sequence; if the time scale of meteorological data is month and day, it is calculated by month, i=1~12; if the time scale is year, it is calculated by year and no longer subdivided into months, i=1; S7, correcting the preliminary interpolation sequence x of the interpolation station according to the interpolation correction parameter to obtain the corrected interpolation sequence y' of the interpolation station; Step S7 specifically comprises using the interpolation correction parameter to correct the values ​​in the preliminary interpolation sequence x of the interpolation station that are consistent with the time series of the subsequence y1 of the interpolation station without data, and obtaining the interpolation station corrected interpolation sequence y'. The calculation formula is: and ′ =a i x1+b i Among them, y ′ is the corrected interpolation sequence of the interpolation station, the time series corresponds to the subsequence y1 of the interpolation station without data, and the data scale is year, month or day; x1 is the part of the preliminary interpolation sequence x of the interpolation station whose time series is consistent with the subsequence y1 of the interpolation station without data, and the data scale is the same as y ′ Consistency; S8. Replace the missing data in the simple interpolation sequence y of the interpolation station according to the corrected interpolation sequence y' of the interpolation station to obtain the final interpolation station completed data sequence.

2. The meteorological data extension method considering mean-variance consistency according to claim 1, characterized in that: The interpolation station initial data sequence includes a data part subsequence and a missing data part subsequence; the missing data part needs to be completed using the above steps in the end; The meteorological data included in the initial data sequence are precipitation, temperature, sunshine, wind speed, and relative humidity; the time scale of the meteorological data is year, month, and day.

3. The meteorological data extension method considering mean-variance consistency according to claim 1, characterized in that: Step S2 specifically includes, when there are 1 to 2 missing data in the initial data sequence of the interpolation station, interpolation is performed by using 0 for precipitation data and linear interpolation for other meteorological data to obtain a simple interpolation sequence y of the interpolation station; The interpolation station simple interpolation sequence y includes an interpolation station subsequence y0 with data and an interpolation station subsequence y1 without data.

4. The meteorological data extension method considering mean-variance consistency according to claim 1 is characterized in that: Step S3 specifically includes: when interpolating daily data, firstly, statistically analyzing the initial data sequence by month to obtain monthly scale data, and setting no monthly data for months with missing data exceeding 7 days, and using the monthly scale data sequence generated by the daily data of the interpolation station and the reference station as the correlation calculation sequence of the interpolation station and the correlation calculation sequence of the reference station respectively; When interpolating at the annual and monthly scales, the corresponding annual or monthly scale sequences of the interpolation station and the reference station are directly used as the interpolation station correlation calculation sequence and the reference station correlation calculation sequence, respectively.

5. The meteorological data extension method considering mean-variance consistency according to claim 1, characterized in that: Step S4 specifically includes calculating the correlation coefficient between the interpolation station and the reference station data sequence, and selecting the corresponding reference station as the reference station used for interpolation according to the correlation coefficient, so as to determine the interpolation reference station list used; When calculating the correlation between the interpolation station and the reference station, the subsequence with data in the correlation calculation sequence of the two stations is used, and for a certain interpolation station, it is necessary to establish a relationship with all the reference stations one by one, and all available reference stations are used as reference stations for interpolation and added to the interpolation reference station list; The selection criteria of the available reference stations are: select the reference stations with correlation coefficients greater than a preset threshold as the reference stations used for interpolation; if the correlation coefficients of all reference stations are less than the preset threshold, select the closest reference station as the reference station used for interpolation; The calculation formula of the correlation coefficient is: Among them, r is the correlation coefficient; cov() is the covariance of the corresponding sequence; var() is the variance of the corresponding sequence; A is the reference station correlation calculation sequence; B is the interpolation station correlation calculation sequence.

6. The meteorological data extension method considering mean-variance consistency according to claim 1, characterized in that: Step S5 specifically includes: using an interpolation method to interpolate the full series of data of each interpolation station based on the initial data sequence of the reference station in the interpolation reference station list, and obtaining a preliminary interpolation sequence x of the interpolation station, wherein the data in the series are all calculated from the reference station data; The interpolation methods used were Thiessen polygon method, inverse distance square method, Kriging method, and spline method; Use some or all of the reference stations in the interpolation reference station list for interpolation according to the requirements of the relevant algorithm.

7. The meteorological data extension method considering mean-variance consistency according to claim 1, characterized in that: Step S8 specifically includes merging the interpolation station corrected interpolation sequence y' and the interpolation station existing data subsequence y0 to obtain a final interpolation station completed data sequence.

Images