A method and system for virtual monitoring of regional precipitation isotopes based on atmospheric circulation model

Abstract

The application discloses a kind of regional precipitation isotope virtual monitoring method and system based on atmospheric circulation model, belong to hydrology meteorological monitoring and isotope hydrology technical field, method includes: based on engineering application demand obtains the reference point of target monitoring area, obtains the precipitation isotope simulation data at reference point and forms reference data sequence;Obtain the virtual precipitation isotope simulation data of multiple atmospheric circulation models, form candidate model set;From candidate model set, the virtual precipitation isotope simulation data of the position where reference point is located is extracted, and mode data sequence is obtained;The matching degree between mode data sequence and reference data sequence is calculated, and the optimal model and corresponding virtual precipitation isotope simulation data are selected from candidate model set;Based on the virtual precipitation isotope simulation data corresponding to optimal model, corresponding reference point in target monitoring area is filled with missing monitoring data or corrected original monitoring data, and the continuity of historical monitoring data of observation station is improved.

Description

Technical Field

[0001] This invention relates to the fields of hydrological and meteorological monitoring and isotope hydrology, and in particular to a method and system for virtual monitoring of regional precipitation isotopes based on atmospheric circulation models. Background Technology

[0002] Precipitation isotopic composition (e.g.) Precipitation isotopes (δD) can be used in research and applications such as water vapor transport, water vapor source analysis, water vapor recirculation, and groundwater recharge, which are of great significance for understanding regional hydrological processes. For example, by analyzing the distribution of precipitation isotopes, we can understand the contribution of different water vapor sources to precipitation, predict climate change, and reveal the use of water resources by vegetation, thereby helping to carry out ecological restoration and water resource management.

[0003] Current research on regional precipitation isotopes typically relies on field measurements, monitoring network data, or literature compilations. However, in arid and high-altitude regions, there are often issues such as sparse observation stations, discontinuous observation periods, and insufficient spatial coverage, making it difficult to obtain continuous precipitation isotope data for any location within the target monitoring area. This results in significant uncertainties regarding the resolution of regional water cycle processes, water vapor source tracing, climate prediction, and hydrological assessment results.

[0004] To compensate for insufficient observations, existing studies have introduced precipitation isotope simulation data from atmospheric circulation models for regional precipitation isotope analysis. However, the applicability of different atmospheric circulation models varies in different regions. Without the adaptation assessment and screening of reference point data, directly using the data output by unscreened atmospheric circulation models as the final monitoring results of regional precipitation isotopes can easily lead to significant deviations in the analysis results of the spatiotemporal distribution of precipitation isotopes. At the same time, current practices are mostly based on scientific research analysis and lack a unified process for engineering applications. This makes it difficult to objectively select the atmospheric circulation model that best fits the target region in practical applications, and also makes it difficult to effectively fill observation gaps and generate high-precision regional precipitation isotope continuous monitoring datasets. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for virtual monitoring of regional precipitation isotopes based on atmospheric circulation models. By calculating the matching degree, atmospheric circulation models that closely match actual reference points are selected. Virtual precipitation isotope simulation data obtained from the selected atmospheric circulation models are then used to supplement or repair the original precipitation isotope data at the reference points, solving the problem of data acquisition difficulties caused by sparse observation stations and insufficient spatial coverage. This invention is achieved through the following technical solutions.

[0006] In a first aspect, the present invention provides a method for virtual monitoring of regional precipitation isotopes based on atmospheric circulation models, comprising:

[0007] Within the target monitoring area, a reference point is determined, and the reference point data corresponding to the reference point is obtained;

[0008] Obtain virtual precipitation isotope simulation data of the reference point under various atmospheric circulation models;

[0009] Based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models, the matching degree evaluation results of various atmospheric circulation models are calculated.

[0010] Based on the matching degree evaluation results, the optimal atmospheric circulation model is selected from multiple atmospheric circulation models;

[0011] Based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model, a spatiotemporal prediction relationship is established for generating precipitation isotope results for the target monitoring area.

[0012] A virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. The original monitoring data is then repaired using the virtual monitoring dataset, thereby enabling virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

[0013] Optionally, the establishment of the spatiotemporal prediction relationship includes: constructing a triangular network based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model and the spatial node coordinates of the data within the target monitoring area;

[0014] The virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. This includes obtaining virtual precipitation isotope data at the reference point based on the spatial coordinates of the reference point in the triangular network. The virtual precipitation isotope data at the reference point based on the long-term series is the virtual monitoring dataset of precipitation isotopes for the target monitoring area.

[0015] Optionally, the reference point data originates from monitoring data, isotope monitoring network data, or literature compilation data. The reference point data includes historical precipitation isotope data and historical meteorological factor data. The historical precipitation isotope data includes δ¹⁸O⁻. 18 O and δD, the historical meteorological factor data including precipitation and temperature.

[0016] In this invention, meteorological factor data is used to assist in the study of precipitation isotope data.

[0017] Optionally, the reference point data can be preprocessed to obtain a reference data sequence, and the virtual precipitation isotope simulation data from various atmospheric circulation models can be preprocessed to obtain a model data sequence.

[0018] Optionally, the preprocessing includes outlier identification and removal, time scale unification, and data consistency processing. The time scale unification includes unifying the reference point data to one of the following: daily scale, monthly scale, seasonal scale, or annual scale.

[0019] Optionally, the calculation of the matching degree evaluation results of various atmospheric circulation models includes calculating the Nash efficiency coefficient, linear fit, and linear regression slope between the model data sequence and the reference data sequence; wherein, if there are multiple reference points, there are multiple reference data sequences, and the Nash efficiency coefficient, linear fit, and linear regression slope between the model data sequence and each reference data sequence are calculated respectively, resulting in multiple Nash efficiency coefficients, linear fits, and linear regression slopes. The optimal atmospheric circulation model is selected by comprehensively considering the case where the Nash efficiency coefficient, linear fit, and linear regression slope approach 1.

[0020] Optionally, the Nash efficiency coefficient is calculated using the following formula:

[0021] ,

[0022] In the formula, The Nash efficiency coefficient. This refers to the isotope sample number of precipitation, with values ​​ranging from 1 to... , This represents the total number of precipitation isotope samples. For the reference data sequence, the first The values ​​of precipitation isotopes, The average value of precipitation isotopes in the reference data series. The first candidate pattern in the sequence The values ​​of virtual precipitation isotopes.

[0023] Optionally, the linear fit is calculated using the following formula:

[0024] ,

[0025] In the formula, For linear fit, The mean value of precipitation isotopes in the candidate pattern sequence.

[0026] Optionally, the linear regression slope is calculated using the following formula:

[0027] ,

[0028] In the formula, The slope of the linear regression. The mean value of precipitation isotopes in the candidate pattern sequence.

[0029] Secondly, the present invention provides a regional precipitation isotope virtual monitoring system based on an atmospheric circulation model, comprising:

[0030] The region determination and reference data acquisition module is used to acquire the reference point data corresponding to the reference point;

[0031] The model data acquisition module is used to acquire virtual precipitation isotope simulation data of the reference point under various atmospheric circulation patterns;

[0032] The matching degree evaluation module is used to calculate the matching degree evaluation results of various atmospheric circulation models based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models.

[0033] The optimal mode selection module is used to select the optimal atmospheric circulation mode from multiple atmospheric circulation modes based on the matching degree evaluation results.

[0034] The spatiotemporal prediction relationship generation module is used to establish spatiotemporal prediction relationships for generating precipitation isotope results for the target monitoring area based on virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model.

[0035] The dataset generation module is used to generate a virtual monitoring dataset of precipitation isotopes in the target monitoring area based on the spatiotemporal prediction relationship. The virtual monitoring dataset of precipitation isotopes is then used to repair the original monitoring data, thereby realizing virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

[0036] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0037] The regional precipitation isotope virtual monitoring method based on atmospheric circulation models introduced in this invention can obtain an atmospheric circulation model that is well-suited to the reference points in the actual target monitoring area by calculating the matching degree between the model data sequence and the reference data sequence. The virtual precipitation isotope data obtained by using the atmospheric circulation model can then fill in the missing monitoring data or correct the original monitoring data for the corresponding reference points in the target monitoring area. This helps to transform incomplete original records into complete long-term continuous time series. It can generate continuous precipitation isotope sequences for locations without observation stations or with incomplete observations, effectively expanding the spatial coverage and temporal continuity of precipitation isotope data, and improving the completeness and usability of historical monitoring data from observation stations.

[0038] The regional precipitation isotope virtual monitoring method based on atmospheric circulation models introduced in this invention comprehensively considers the matching degree between the model data sequence and the reference data sequence by calculating the Nash efficiency coefficient (NSE), linear fit degree (R²), and linear regression slope index (K). This method can objectively select the optimal model that fits the target region from a variety of candidate atmospheric circulation models, avoiding the systematic bias that may be caused by directly using the output of a single unscreened atmospheric circulation model, and improving the accuracy of regional precipitation isotope virtual monitoring results. Attached Figure Description

[0039] Figure 1 The diagram shown is a flowchart of a regional precipitation isotope virtual monitoring method based on an atmospheric circulation model in one embodiment of the present invention.

[0040] Figure 2 The diagram shown is a schematic representation of the target monitoring area and station distribution in one embodiment of the present invention.

[0041] Figure 3 The diagram shown is a comparison of the Nash efficiency coefficient (NSE) in one embodiment of the present invention.

[0042] Figure 4 The figure shown is a schematic diagram comparing the linear fitting degree R² in one embodiment of the present invention;

[0043] Figure 5 The diagram shown is a comparison of the linear regression slope index K in one embodiment of the present invention.

[0044] Figure 6 The figure shown is a monthly precipitation at a reference point in one embodiment of the present invention. Schematic diagram of changes;

[0045] Figure 7 The figure shown is a time series diagram of virtual monitoring results in one embodiment of the present invention. Detailed Implementation

[0046] The following description, in conjunction with the accompanying drawings and specific embodiments, provides further details. In this description, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature.

[0047] Example 1

[0048] This embodiment introduces a method for virtual monitoring of regional precipitation isotopes based on atmospheric circulation models, such as... Figure 1 As shown, it includes the following:

[0049] Within the target monitoring area, a reference point is determined, and the reference point data corresponding to the reference point is obtained;

[0050] Obtain virtual precipitation isotope simulation data of the reference point under various atmospheric circulation models;

[0051] Based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models, the matching degree evaluation results of various atmospheric circulation models are calculated.

[0052] Based on the matching degree evaluation results, the optimal atmospheric circulation model is selected from multiple atmospheric circulation models;

[0053] Based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model, a spatiotemporal prediction relationship is established for generating precipitation isotope results for the target monitoring area.

[0054] A virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. The original monitoring data is then repaired using the virtual monitoring dataset, thereby enabling virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

[0055] Example 2

[0056] Based on Example 1, this example introduces the specific implementation process of a regional precipitation isotope virtual monitoring method based on atmospheric circulation models, which includes the following:

[0057] I. Determine the target monitoring area

[0058] In one specific embodiment of the present invention, a target monitoring area is first determined based on engineering application requirements to constrain the spatial range and output boundary of the virtual monitoring results. Corresponding to the target monitoring area, a target time range and output time scale are determined and used as unified constraints for subsequent pruning, time alignment, statistical summarization, and result output of reference data sequences, candidate pattern sets, and pattern data sequences.

[0059] In this embodiment, the selected target monitoring area is the Qaidam Basin, such as Figure 2 As shown, the Qaidam Basin includes nine stations: Mangya, Lenghu, Xiaozhaohuo, Dachaidan, Golmud, Nuomuhong, Delingha, Wulan, and Dulan (these stations serve as reference points). The target time range is set from 1979 to 2007, and the reference point precipitation isotopes... The monthly-scale series covers the period from 1992 to 2006 (because data exists at the reference point during this time period), and the isotopic data used to calculate the matching degree are the isotopic data from the period covered by both data. The precipitation isotopes of the monthly-scale series refer to the precipitation isotope values ​​of the same month over many years.

[0060] II. Reference data sequence, candidate pattern set, and pattern data sequence

[0061] 2.1 Reference Data Sequence

[0062] In one specific embodiment of the present invention, one or more reference points are determined within a selected target monitoring area, reference point data is acquired based on these reference points, and preprocessed to obtain a reference data sequence.

[0063] In one specific embodiment of the present invention, the reference point data originates from monitoring data, isotope monitoring network data, or literature compilation data. The reference point data includes historical precipitation isotope data and historical meteorological factor data. The historical precipitation isotope data includes... Historical meteorological factor data, including precipitation and temperature, is used to support the study of precipitation isotope data, along with δD.

[0064] In one specific embodiment of this invention, based on a set target time range, historical isotope values ​​of corresponding meteorological stations are collected according to the time field, and then preprocessed to form a reference data sequence of time fields and corresponding numerical fields. In practical applications, only the Delingha station among the nine stations was able to collect data. Other meteorological stations did not collect data due to high altitude, cold climate, and drought.

[0065] The preprocessing of the reference data sequence described above includes standardization and quality control. Standardization includes unifying the representation of time fields, units, and variable naming conventions, and marking data records that cannot be parsed or have missing key fields for removal or completion. Unifying the representation of time fields involves standardizing the reference point data to a daily, monthly, seasonal, or annual scale. Performing standardization on the original reference point data table ensures that records from different sources and batches have a unified data structure and parsing standards. This allows the reference data sequence to form a one-to-one comparable sequence with the candidate pattern set at the same time scale. Quality control includes outlier identification, duplicate record handling, missing record marking, and timestamp integrity checks. The results of handling outliers and missing records are marked with traceable tags to support subsequent evaluation of validity and interpretation.

[0066] 2.2 Candidate Pattern Set

[0067] In one specific embodiment of the present invention, virtual precipitation isotope simulation data obtained from 11 existing atmospheric circulation models are selected to form a candidate model set. These 11 atmospheric circulation models are CAM2, GISS-E(free), GISS-E(nudged), GISS-E(MERRA), GISS-E(NCEP), HadAM3, LMDZ(free), LMDZ(nudged), MIROC32, isoGSMnew(free), and isoGSMnew(nudged), and are defined as candidate models.

[0068] The virtual precipitation isotope simulation data acquired from each atmospheric circulation model undergoes uniform data organization and processing to meet the input format requirements for batch processing and automated evaluation. This uniform data organization and processing includes: cropping candidate model data to the target monitoring area; cropping time series to the target time range; and renaming or mapping variable fields to variable definitions consistent with the reference data series, thereby reducing structural differences in multi-source heterogeneous data during subsequent alignment stages.

[0069] 2.3 Pattern Data Sequence

[0070] In one specific embodiment of the present invention, based on the location of the reference point, i.e. the actual location of the Delingha station, the preprocessed virtual precipitation isotope simulation data output from each atmospheric circulation model is converted into a model data sequence at the reference point location.

[0071] Next, the model data sequence and the reference data sequence are aligned with timestamps and consistent with time scales, ensuring that they correspond item by item under the same timestamp set and time scale, forming a comparative sequence pair for which evaluation indicators can be directly calculated. When the timestamps of the comparative sequence pair are not completely consistent, the common coverage timestamp is determined as the evaluation window, and the uncovered timestamps are marked as missing data for completion or correction during the output stage.

[0072] III. Calculating the matching degree

[0073] In one specific embodiment of the present invention, the calculation of the matching degree evaluation results of various atmospheric circulation models includes calculating the Nash efficiency coefficient, linear fit degree, and linear regression slope between the model data sequence and the reference data sequence; wherein, if there are multiple reference points, there are multiple reference data sequences, and the Nash efficiency coefficient, linear fit degree, and linear regression slope between the model data sequence and each reference data sequence are calculated respectively, resulting in multiple Nash efficiency coefficients, linear fit degrees, and linear regression slopes, and the optimal atmospheric circulation model is selected by comprehensively considering the case where the Nash efficiency coefficient, linear fit degree, and linear regression slope approach 1.

[0074] After calculating multiple Nash efficiency coefficients, linear fit, and linear regression slopes between the reference point sequence and the corresponding pattern data sequence, the optimal pattern is selected by manual screening, taking into account that all three indicators are close to 1.

[0075] The Nash efficiency coefficient is calculated using the following formula:

[0076] ,

[0077] In the formula, The Nash efficiency coefficient. This refers to the isotope sample number of precipitation, with values ​​ranging from 1 to... , This represents the total number of precipitation isotope samples. For the reference data sequence, the first The values ​​of precipitation isotopes, The average value of precipitation isotopes in the reference data series. The first candidate pattern in the sequence Numerical values ​​of virtual precipitation isotopes. Nash efficiency coefficient. Used to measure the overall goodness of fit of the model data sequence to the reference data sequence, in order to evaluate the overall accuracy of the model output.

[0078] The linear fit is calculated using the following formula:

[0079] ,

[0080] In the formula, For linear fit, This represents the mean of precipitation isotopes in the candidate pattern sequences. Linearity fit. Used to measure the degree of linear correlation between the reference data sequence and the pattern data sequence within the same time window.

[0081] The slope of the linear regression is calculated using the following formula:

[0082] ,

[0083] In the formula, This represents the slope of the linear regression. Used to measure the magnitude consistency of the pattern data sequence with respect to the reference data sequence.

[0084] In one specific embodiment of the present invention, to conduct a dual-scale adaptability assessment, long-sequence data and monthly-scale data are constructed at a reference point. Virtual precipitation isotopes obtained from the candidate model at the Delingha meteorological station at the reference point are used. A complete time series within the effective time window is defined as a long-sequence pattern data series. Data from different years within the same month of a long-sequence pattern data series are treated as a group of samples. Multi-year monthly average historical precipitation isotopes and virtual precipitation isotopes are calculated to form monthly-scale candidate sequences and monthly-scale reference sequences, including multi-year weighted monthly average precipitation from January to December. The multi-year monthly average precipitation isotopes were calculated using the following formula:

[0085] ,

[0086] In the formula, δ represents the weighted average monthly precipitation over many years for month m. 18 O, where the overline indicates that the average value is taken from a multi-year sample; 𝑚 is the month number, ranging from 1 to 12; y is the year number; This represents the precipitation in year y within month m. This represents the δ of year y in month m. 18 O; Let m be the year number of the month m.

[0087] In the above, "multi-year monthly average" means the average precipitation isotope values ​​for the same month over many years.

[0088] Figures 3-5 Nash efficiency coefficients for matching candidate pattern data sequences with reference point data sequences, respectively. Comparison diagram, linear fit Comparison diagram and linear regression slope Comparison diagram. From Figures 3-5 It can be seen that the Nash efficiency coefficients of LMDZ(nudged), LMDZ(free), and isoGSMnew(nudged) are... The linear fit of LMDZ (nudged) is close to 1. The linear regression slope index is 0.73. The value is 0.72. The closer the three matching indicators mentioned above are to 1, the better the matching effect.

[0089] Figure 6 Monthly precipitation isotopes obtained at the reference point for candidate models The monthly precipitation isotopes actually obtained at the Delingha station Comparison diagram. Figure 6 In the middle, the monthly multi-year average value at the reference point Delingha station is used. The change sequence is used as the reference sequence, and the precipitation isotope change sequence of each candidate model at the same time scale at the reference point Delingha station is used as the candidate sequence.

[0090] Depend on Figure 6 It can be seen that the results obtained by the candidate pattern LMDZ (nudged) are closer to the numerical values ​​of the reference sequence. Combined with... Figures 3-5 Based on the combined data from the three indicators, LMDZ(nudged) was selected as the optimal model for virtual monitoring of precipitation isotopes in the Qaidam Basin. The virtual precipitation isotope results obtained from LMDZ(nudged) were used as the optimal model data for virtual monitoring of precipitation isotopes in the Qaidam Basin. That is, in this embodiment, the optimal atmospheric circulation model selected is LMDZ(nudged).

[0091] IV. Generating a Virtual Monitoring Dataset for Precipitation Isotopes

[0092] In one specific embodiment of the present invention, the establishment of the spatiotemporal prediction relationship includes: constructing a triangular network based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model and the spatial node coordinates of the data within the target monitoring area;

[0093] The virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. This includes obtaining virtual precipitation isotope data at the reference point based on the spatial coordinates of the reference point in the triangular network. The virtual precipitation isotope data at the reference point based on the long-term series is the virtual monitoring dataset of precipitation isotopes for the target monitoring area.

[0094] In practical applications, if precipitation isotope data is missing at a certain station within the target monitoring area, it can be filled in using virtual precipitation isotope data from the corresponding station within the precipitation isotope virtual monitoring dataset; conversely, if no data is missing, it can be corrected using virtual precipitation isotope data from the corresponding station within the precipitation isotope virtual monitoring dataset. For example... Figure 7 The figure shows a time series diagram of virtual monitoring results, realizing virtual monitoring of precipitation isotopes at reference points in the target monitoring area. This figure displays continuous precipitation isotope generation data for each station within the target monitoring area from 1979 to 2007. For stations with completely missing measured data (such as Mangya and Lenghu), this complete series is directly used as filler data; for stations with only partial measured data (such as Delingha), this series is used to fill in the gaps outside the period covered by their measured data (1992-2006), thus providing complete long-term continuous series support for the entire region. It can be seen that using the regional precipitation isotope virtual monitoring method based on atmospheric circulation models introduced in this invention can improve the completeness and usability of historical monitoring data from observation stations.

[0095] Example 3

[0096] This embodiment introduces a regional precipitation isotope virtual monitoring system based on an atmospheric circulation model, including:

[0097] The region determination and reference data acquisition module is used to acquire the reference point data corresponding to the reference point;

[0098] The model data acquisition module is used to acquire virtual precipitation isotope simulation data of the reference point under various atmospheric circulation patterns;

[0099] The matching degree evaluation module is used to calculate the matching degree evaluation results of various atmospheric circulation models based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models.

[0100] The optimal mode selection module is used to select the optimal atmospheric circulation mode from multiple atmospheric circulation modes based on the matching degree evaluation results.

[0101] The spatiotemporal prediction relationship generation module is used to establish spatiotemporal prediction relationships for generating precipitation isotope results for the target monitoring area based on virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model.

[0102] The dataset generation module is used to generate a virtual monitoring dataset of precipitation isotopes in the target monitoring area based on the spatiotemporal prediction relationship. The virtual monitoring dataset of precipitation isotopes is then used to repair the original monitoring data, thereby realizing virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

[0103] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for virtual monitoring of regional precipitation isotopes based on atmospheric circulation models, characterized in that, include: Within the target monitoring area, a reference point is determined, and the reference point data corresponding to the reference point is obtained; Obtain virtual precipitation isotope simulation data of the reference point under various atmospheric circulation models; Based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models, the matching degree evaluation results of various atmospheric circulation models are calculated. Based on the matching degree evaluation results, the optimal atmospheric circulation model is selected from multiple atmospheric circulation models; Based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model, a spatiotemporal prediction relationship is established for generating precipitation isotope results for the target monitoring area. A virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. The original monitoring data is then repaired using the virtual monitoring dataset, thereby enabling virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

2. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 1, characterized in that, The establishment of the spatiotemporal prediction relationship includes: constructing a triangular network based on the virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model and the spatial node coordinates of the data within the target monitoring area; The virtual monitoring dataset of precipitation isotopes for the target monitoring area is generated based on the spatiotemporal prediction relationship. This includes obtaining virtual precipitation isotope data at the reference point based on the spatial coordinates of the reference point in the triangular network. The virtual precipitation isotope data at the reference point based on the long-term series is the virtual monitoring dataset of precipitation isotopes for the target monitoring area.

3. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 1, characterized in that, The reference point data is derived from monitoring data, isotope monitoring network data, or literature compilation data. The reference point data includes historical precipitation isotope data and historical meteorological factor data. The historical precipitation isotope data includes δ¹⁸O⁻. 18 O and δD, the historical meteorological factor data including precipitation and temperature.

4. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 1, characterized in that, Reference data sequences are obtained by preprocessing reference point data, and model data sequences are obtained by preprocessing virtual precipitation isotope simulation data from various atmospheric circulation models.

5. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 4, characterized in that, The preprocessing includes outlier identification and removal, time scale unification, and data consistency processing. The time scale unification includes unifying the reference point data to one of the following: daily scale, monthly scale, seasonal scale, or annual scale.

6. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 4, characterized in that, The evaluation results of the matching degree of various atmospheric circulation models include calculating the Nash efficiency coefficient, linear fit, and linear regression slope between the model data sequence and the reference data sequence. If there are multiple reference points, there are multiple reference data sequences. The Nash efficiency coefficient, linear fit, and linear regression slope between the model data sequence and each reference data sequence are calculated separately to obtain multiple Nash efficiency coefficients, linear fits, and linear regression slopes. The optimal atmospheric circulation model is selected by comprehensively considering the case where the Nash efficiency coefficient, linear fit, and linear regression slope approach 1.

7. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 6, characterized in that, The Nash efficiency coefficient is calculated using the following formula: ， In the formula, The Nash efficiency coefficient. This refers to the isotope sample number of precipitation, with values ​​ranging from 1 to... , This represents the total number of precipitation isotope samples. For the reference data sequence, the first The values ​​of precipitation isotopes, The average value of precipitation isotopes in the reference data series. The first candidate pattern in the sequence The values ​​of virtual precipitation isotopes.

8. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 7, characterized in that, The linear fit degree is calculated using the following formula: ， In the formula, For linear fit, The mean value of precipitation isotopes in the candidate pattern sequence.

9. The regional precipitation isotope virtual monitoring method based on atmospheric circulation models according to claim 7, characterized in that, The linear regression slope is calculated using the following formula: ， In the formula, The slope of the linear regression. The mean value of precipitation isotopes in the candidate pattern sequence.

10. A regional precipitation isotope virtual monitoring system based on an atmospheric circulation model, characterized in that, include: The region determination and reference data acquisition module is used to acquire the reference point data corresponding to the reference point; The model data acquisition module is used to acquire virtual precipitation isotope simulation data of the reference point under various atmospheric circulation patterns; The matching degree evaluation module is used to calculate the matching degree evaluation results of various atmospheric circulation models based on the reference point data of the reference point and the virtual precipitation isotope simulation data of various atmospheric circulation models. The optimal mode selection module is used to select the optimal atmospheric circulation mode from multiple atmospheric circulation modes based on the matching degree evaluation results. The spatiotemporal prediction relationship generation module is used to establish spatiotemporal prediction relationships for generating precipitation isotope results for the target monitoring area based on virtual precipitation isotope simulation data corresponding to the optimal atmospheric circulation model. The dataset generation module is used to generate a virtual monitoring dataset of precipitation isotopes in the target monitoring area based on the spatiotemporal prediction relationship. The virtual monitoring dataset of precipitation isotopes is then used to repair the original monitoring data, thereby realizing virtual monitoring of precipitation isotopes at reference points in the target monitoring area.

Images