Multi-source collaborative rain gauge precipitation data quality control method, system, device and medium

Abstract

The application discloses a multi-source cooperative rain gauge station precipitation data quality control method, system, equipment and medium, the method constructs a plurality of local area small groups; the correct data in the initial quality control observation precipitation data corresponding to each local area small group is detected by the adjacent station, and the adjacent station detection result is obtained; the main station with suspicious data in the adjacent station detection result is taken as a to-be-detected station, and the main station without an adjacent station or with a number of adjacent stations less than a preset number threshold in a preset distance is taken as a to-be-detected station; the observation precipitation data corresponding to the to-be-detected station is detected by radar, and the to-be-detected station corresponding to the radar detection judgment as suspicious data or error data is listed in a blacklist; the data in the blacklist, the error data or the suspicious data and the missing data in the initial quality control observation precipitation data are subjected to interpolation processing, and target rain gauge station precipitation real product is obtained. The application can perform rain gauge station precipitation data quality control in real time, and improve the accuracy of rain gauge station precipitation data quality control.

Description

Technical Field

[0001] This application relates to the field of meteorological data quality control technology, and in particular to a method, system, equipment and medium for quality control of precipitation data from multi-source collaborative rain gauge stations. Background Technology

[0002] Rain gauge precipitation data is a core data source for weather and meteorological monitoring, climate analysis, and watershed hydrological simulation, with extremely wide applications. Rain gauges operate based on the core principles of precipitation collection and measurement sensing. They collect natural precipitation through a collection device and utilize metering methods such as tipping bucket induction and weighing sensors to convert the amount of liquid precipitation into quantifiable and recordable electrical signals and numerical data, enabling quantitative monitoring of precipitation over time periods and long time series. However, during long-term field observations, rain gauge data is susceptible to interference from multiple factors, including the natural environment, instrumentation, transmission systems, and human operation and maintenance, resulting in random errors and systematic biases that severely affect the accuracy and usability of precipitation data. Therefore, conducting systematic and refined data quality control is a necessary prerequisite for ensuring the scientific application of precipitation data.

[0003] Although existing research has gradually introduced technologies such as collaborative quality control and fusion correction of multi-source data from surrounding stations and radar remote sensing, there are still significant shortcomings in current multi-source quality control technologies. They generally adopt a static fusion verification mode with fixed weights, which does not take into account the differences in accuracy of multi-source data, spatiotemporal adaptability biases, and data uncertainties in complex terrain. They cannot dynamically adjust verification rules according to precipitation intensity and weather scenarios, resulting in poor quality control refinement and scale adaptability. Furthermore, they are all offline post-correction modes, lacking real-time online verification capabilities for data streams, making it difficult to support high-timeliness hydrological business scenarios such as flash flood warnings and real-time watershed monitoring.

[0004] Therefore, existing related technologies still have problems such as uncontrolled precipitation data from rain gauge stations, low availability, difficulty in real-time quality control of precipitation data from rain gauge stations, and inaccurate quality control of precipitation data from rain gauge stations. Summary of the Invention

[0005] This application aims to propose a multi-source collaborative method, system, equipment, and medium for quality control of precipitation data from rain gauge stations. It solves the problems of uncontrolled precipitation data and low availability of precipitation data from rain gauge stations, and enables real-time quality control of precipitation data from rain gauge stations, improving the accuracy of quality control and enhancing the overall availability of precipitation data from rain gauge stations.

[0006] In a first aspect, embodiments of this application provide a method for quality control of precipitation data from multi-source collaborative rain gauge stations, the method comprising: The acquired precipitation data from rain gauge stations are preprocessed to obtain a standardized dataset for quality control. Metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks are performed on the data in the standardized dataset for quality control to obtain initial quality control precipitation data. The initial quality control precipitation data includes erroneous or suspicious data as well as correct data. Arbitrarily select a rain gauge station as the master station, and filter out multiple neighboring stations within a preset radius centered on the master station; form a local spatial group with each master station and its corresponding multiple neighboring stations to construct multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the master station; For each of the local spatial groups, the correct data in the initial quality control observation precipitation data corresponding to the local spatial group are checked by neighboring stations to obtain neighboring station detection results containing correct data and suspicious data; The main station with suspicious data in the neighboring station detection results is designated as the station to be inspected, and the main station with no neighboring station within a preset distance or with a number of neighboring stations less than a preset threshold is designated as the station to be inspected. Radar detection is performed on the observed precipitation data corresponding to the stations to be inspected, and the stations to be inspected that are judged by radar detection as having suspicious or erroneous data are blacklisted. Interpolation processing is performed on the data of the stations to be inspected in the blacklist, as well as the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, to obtain the precipitation real-time product of the target rain gauge station.

[0007] In some implementations, the step of performing metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks on the data in the standardized dataset to be quality controlled, to obtain initial quality control observation precipitation data, includes: Metadata checks are performed on the data in the standardized dataset to be quality controlled to obtain metadata check results; the metadata check is to check the temporal continuity and integrity of the data in the standardized dataset to be quality controlled. The data in the metadata check results are subjected to eigenvalue checks to obtain a cleaned and valid observation dataset; the eigenvalue check is to remove invalid data samples, including those with missing data and no observed precipitation data. The data in the cleaned effective observation dataset are subjected to a climate limit value check to obtain data that passes the climate limit value check; the climate limit value check is to check whether the data in the cleaned effective observation dataset exceeds the set climate limit value check threshold. The data that passed the climate limit value check is then subjected to a station limit value check to obtain data that passed the station limit value check; the station limit value check is to check whether the data that passed the climate limit value check exceeds the set station limit value. The data that passed the station boundary value check were subjected to a dead value check to obtain the initial quality control observation precipitation data; the dead value check was to check whether the data that passed the station boundary value check exceeded the set dead value check threshold.

[0008] In some implementations, for each of the local spatial groups, the correct data in the initial quality control precipitation data corresponding to the local spatial group are subjected to neighbor station detection to obtain neighbor station detection results containing correct data and suspicious data, including: Based on the observed precipitation data corresponding to the main station in the current local spatial group, the precipitation amount of the main station in the current local spatial group is calculated; based on whether the precipitation amount of the main station in the current local spatial group is 0, it is determined whether the observed precipitation data corresponding to the main station is correct data or suspicious data. If the observed precipitation data corresponding to the main station is determined to be suspicious, a first-level comparison is performed on the precipitation of the main station in the current local spatial group to obtain the first-level comparison result; based on the first-level comparison result, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal; the first-level comparison is to compare the precipitation of the main station in the current local spatial group with multiple different thresholds; If the observed precipitation data corresponding to the main station is determined to be abnormal, the Grubbs method is used to determine the observed precipitation data corresponding to the main station based on the cumulative precipitation of all nearby stations within a preset radius within a preset time period, and the determination result is obtained. If the discrimination result indicates that the observed precipitation data corresponding to the main station is abnormal, then a second-level comparison is performed on the precipitation of the main station in the current local spatial group to obtain a second-level comparison result; based on the second-level comparison result, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal; the second-level comparison is to compare the precipitation of the main station in the current local spatial group with multiple different thresholds.

[0009] In some implementations, determining whether the observed precipitation data corresponding to the master station is correct or questionable based on whether the precipitation data of the master station in the current local spatial group is 0 includes: Determine whether the cumulative precipitation at the main station in the current local spatial group is 0 within a preset time. If the cumulative precipitation is 0, obtain the number of first neighboring stations with cumulative precipitation greater than 0 within the preset time and the number of second neighboring stations with precipitation greater than 0 per hour within the preset time. Construct a first discrimination condition based on the first and second neighboring station numbers. If the observed precipitation data corresponding to the neighboring stations meets the first discrimination condition, determine that the observed precipitation data corresponding to the main station is suspicious data. If the observed precipitation data corresponding to the neighboring stations does not meet the first discrimination condition, determine that the observed precipitation data corresponding to the main station is correct data. The system counts the number of times within a preset time period when the precipitation at the national station and the neighboring stations in the current local spatial group exceeds a first preset value, and constructs a second discrimination condition based on the number of such times. If the observed precipitation data corresponding to the national station or the neighboring station does not meet the second discrimination condition, the observed precipitation data corresponding to the main station is judged to be suspicious data. If the observed precipitation data corresponding to the national station or the neighboring station meets the second discrimination condition, the observed precipitation data corresponding to the main station is judged to be correct data. If the precipitation data of the master station in the current local spatial group is greater than 0, the Grubbs method is used to determine whether the observed precipitation data corresponding to the master station is correct or suspicious.

[0010] In some implementations, determining whether the observed precipitation data corresponding to the main station is correct or abnormal based on the first hierarchical comparison result includes: If the first classification comparison result indicates that the precipitation at the current main station is less than the first threshold, then it is determined whether there is precipitation at the nearest national station within the preset range in the current hour and the previous hour. If there is precipitation, then it is determined that the observed precipitation data corresponding to the current main station is correct. The first number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the first number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the first classification comparison result indicates that the current precipitation at the main station is less than the second threshold, then it is determined whether there is precipitation in the current hour and the previous hour at the nearest national station within the preset range. If there is precipitation and there is a time-based precipitation greater than or equal to the second preset value, then it is determined that the observed precipitation data corresponding to the current main station is correct. The second number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the second number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the first classification comparison result indicates that the current precipitation at the main station is less than the third threshold, then it is determined whether there is precipitation in the current hour and the previous hour at the nearest national station within the preset range. If there is precipitation and there is a time-based precipitation greater than or equal to the third preset value, then the observed precipitation data corresponding to the current main station is determined to be correct. The third number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the third number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. The first threshold, the second threshold, and the third threshold are preset values, and the first threshold, the second threshold, and the third threshold are not equal.

[0011] In some implementations, determining whether the observed precipitation data corresponding to the main station is correct or abnormal based on the second classification comparison result includes: If the second classification comparison result indicates that the current precipitation at the main station is less than the fourth threshold, then it is determined whether the cumulative precipitation at the nearest national station within the preset range for three hours is greater than 0. If the cumulative precipitation for three hours is greater than 0, then it is determined that the observed precipitation data corresponding to the current main station is correct. The fourth number of neighboring stations with cumulative precipitation for three hours greater than 0 is counted, and based on the fourth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the second classification comparison result indicates that the current precipitation at the main station is less than the fifth threshold, then the cumulative precipitation of the nearest national station within the preset range over three hours is counted. If the cumulative precipitation over three hours is greater than or equal to the first cumulative precipitation threshold, then the observed precipitation data corresponding to the current main station is determined to be correct. The fifth number of neighboring stations with cumulative precipitation over three hours greater than 0 is counted. Based on the fifth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the second classification comparison result indicates that the current precipitation at the main station is less than the sixth threshold, then the cumulative precipitation of the nearest national station within the preset range over three hours is counted. If the cumulative precipitation over three hours is greater than or equal to the second cumulative precipitation threshold, then the observed precipitation data corresponding to the current main station is determined to be correct. The sixth number of neighboring stations with cumulative precipitation over three hours greater than 0 is counted. Based on the sixth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the cumulative precipitation at the main station over three hours is greater than the seventh threshold, and the maximum cumulative precipitation at a nearby station over three hours is less than the third cumulative precipitation threshold, then the observed precipitation data corresponding to the main station is determined to be abnormal; the fourth threshold, the fifth threshold, the sixth threshold, and the seventh threshold are preset values, and the fourth threshold, the fifth threshold, the sixth threshold, and the seventh threshold are not equal.

[0012] In some implementations, the step of performing radar detection on the observed precipitation data corresponding to the stations to be inspected, and adding the stations corresponding to the radar detection as suspicious or erroneous data to the blacklist, includes: The first observed precipitation value of the station to be inspected for one hour is statistically analyzed, and the second observed precipitation value of the station to be inspected for three hours is statistically analyzed. Multiple grid data points are selected centered on the latitude and longitude of the station to be inspected; the target radar quantitatively estimated precipitation with the smallest difference from the observed precipitation data corresponding to the station to be inspected is selected from the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points; and the maximum radar quantitatively estimated precipitation is selected from the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points. Based on the first observed precipitation value, the area is divided into a first range, a second range, a third range, a fourth range, a fifth range, and a sixth range; If the first observed precipitation value meets the first range, then based on the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points and the target radar quantitatively estimated precipitation, the observed precipitation data corresponding to the station to be inspected is judged to be zero-value precipitation, and the station to be inspected corresponding to the judged suspicious data or erroneous data is blacklisted. If the first observed precipitation value meets any of the second, third, fourth, fifth, and sixth ranges, or if the second observed precipitation value meets the seventh range, then based on the first observed precipitation value, the second observed precipitation value, the quantitatively estimated precipitation by the target radar, and the maximum quantitatively estimated precipitation by the radar, the observed precipitation data corresponding to the station to be inspected is judged as false precipitation and outliers, and the stations to be inspected corresponding to the data judged as suspicious or erroneous are blacklisted.

[0013] Secondly, embodiments of this application also provide a multi-source collaborative rain gauge precipitation data quality control system, the system comprising: The data inspection module is used to preprocess the acquired precipitation data from rain gauge stations to obtain a standardized dataset for quality control; the data in the standardized dataset for quality control are then subjected to metadata inspection, feature value inspection, climate limit value inspection, station limit value inspection, and dead value inspection to obtain initial quality control precipitation data; the initial quality control precipitation data includes erroneous or suspicious data as well as correct data; The neighboring station filtering module is used to arbitrarily select a rain gauge station as the main station and filter out multiple neighboring stations within a preset radius centered on the main station; each main station and its corresponding multiple neighboring stations are combined into a local spatial group to construct multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the main station; The neighbor station detection module is used to perform neighbor station detection on the correct data in the initial quality control observation precipitation data corresponding to each local spatial group, and obtain the neighbor station detection result containing correct data and suspicious data. The station to be inspected module is used to designate the main station with suspicious data in the detection results of the nearby stations as the station to be inspected, and to designate the main station with no nearby stations within a preset distance or with the number of nearby stations less than a preset threshold as the station to be inspected. The radar detection module is used to perform radar detection on the observed precipitation data corresponding to the stations to be inspected, and to blacklist the stations to be inspected that are determined by radar detection to be suspicious or erroneous data. The interpolation processing module is used to interpolate the data of the stations to be inspected in the blacklist, as well as the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, to obtain the precipitation data of the target rain gauge station.

[0014] Thirdly, embodiments of this application also provide an electronic device, including at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which are executed by the at least one control processor to enable the at least one control processor to perform a multi-source collaborative rain gauge precipitation data quality control method as described above.

[0015] Fourthly, embodiments of this application also provide a computer-readable storage medium storing computer-executable instructions for causing a computer to execute a multi-source collaborative rain gauge precipitation data quality control method as described above.

[0016] Compared with the prior art, this application has the following beneficial effects: This application preprocesses the acquired precipitation data from rain gauge stations to obtain a standardized dataset for quality control. The standardized dataset undergoes metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks to obtain initial quality control precipitation data. This initial quality control precipitation data includes erroneous or suspicious data, missing data, and correct data. A rain gauge station is arbitrarily selected as the master station, and multiple neighboring stations within a preset radius centered on the master station are selected. Each master station and its corresponding multiple neighboring stations form a local spatial group, resulting in multiple local spatial groups. The multiple neighboring stations are rain gauge stations adjacent to the master station. Each local spatial group performs neighbor station detection on the correct data in the initial quality control precipitation observation data corresponding to the local spatial group, obtaining neighbor station detection results containing both correct and suspicious data; the main station with suspicious data in the neighbor station detection results is designated as the station to be inspected, as are the main stations with no neighboring stations within a preset distance or with fewer than a preset threshold number of neighboring stations; the precipitation observation data corresponding to the stations to be inspected is subjected to radar detection, and the stations corresponding to the radar detection that are identified as having suspicious or erroneous data are added to a blacklist; the data in the blacklist, the erroneous or suspicious data in the initial quality control precipitation observation data, and the missing data are interpolated to obtain the actual precipitation product of the target rain gauge station.

[0017] Thus, by performing metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks on the data in the standardized dataset to be quality controlled, relatively high-quality initial quality control observation precipitation data can be obtained. Furthermore, by constructing a local spatial group, the quality of the observation precipitation data of the main station in the local spatial group is quickly checked using neighboring stations. Suspicious data that fails the neighboring station check and observation precipitation data corresponding to the main station with no neighboring stations or fewer than a preset threshold are subjected to radar detection. This solves the problems of uncontrolled precipitation data and low availability at rain gauge stations, enabling real-time quality control of rain gauge station precipitation data and further improving the accuracy of quality control. Finally, interpolation processing is performed on data in the blacklist, erroneous or suspicious data in the initial quality control observation precipitation data, and missing data, which improves the overall availability of rain gauge station data. Attached Figure Description

[0018] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart illustrating an embodiment of the multi-source collaborative rain gauge precipitation data quality control method provided in this application; Figure 2 This is a schematic diagram of an embodiment of the multi-source collaborative rain gauge precipitation data quality control system provided in this application; Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device provided in this application. Detailed Implementation

[0019] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0020] In the description of this application, the use of terms such as "first," "second," etc., is for the purpose of distinguishing technical features only and should not be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated or the order of the technical features indicated.

[0021] In the description of this application, it should be understood that the orientation descriptions, such as up, down, etc., are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0022] In the description of this application, it should be noted that, unless otherwise explicitly defined, terms such as "setup," "installation," and "connection" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this application in conjunction with the specific content of the technical solution.

[0023] To address the problems in related technologies, such as uncontrolled precipitation data from rain gauge stations, low availability, difficulty in real-time quality control of precipitation data from rain gauge stations, and inaccurate quality control of precipitation data from rain gauge stations, this application proposes a multi-source collaborative method, system, equipment, and medium for quality control of precipitation data from rain gauge stations.

[0024] Reference Figure 1 This application provides a flowchart illustrating a multi-source collaborative rain gauge precipitation data quality control method. This method is applied to electronic devices, such as servers or mobile terminals. Figure 1 As shown, the multi-source collaborative rain gauge precipitation data quality control method may include the following steps S101 to S106.

[0025] Step S101: Preprocess the acquired precipitation data from rain gauge stations to obtain a standardized dataset for quality control; perform metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks on the data in the standardized dataset for quality control to obtain initial quality control precipitation data; the initial quality control precipitation data includes erroneous or suspicious data as well as correct data.

[0026] In this step, the provincial-level Tianqing interface acquires real-time precipitation data from provincial-level natural resources rain gauge stations and provincial-level transportation rain gauge stations. The raw data containing precipitation data from these stations is then uniformly converted into a standard data format, including field naming, data type, and timestamp format, forming a standardized dataset for quality control.

[0027] Metadata checks are performed on the data in the standardized dataset to be quality controlled, yielding metadata check results. Metadata checks verify the temporal continuity and completeness of the data in the standardized dataset. Feature value checks are then performed on the data in the metadata check results to obtain a cleaned, valid observation dataset. Feature value checks remove invalid data samples, including those with missing data and no observed precipitation data. Climate boundary value checks are then performed on the cleaned, valid observation dataset to obtain data that passes the climate boundary value check. Climate boundary value checks verify whether the data in the cleaned, valid observation dataset exceeds the set climate boundary value check threshold. Station boundary value checks are then performed on the data that passes the climate boundary value check to obtain data that passes the station boundary value check. Station boundary value checks verify whether the data that passes the climate boundary value check exceeds the set station boundary value. Finally, a dead value check is performed on the data that passes the station boundary value check to obtain initial quality control observation precipitation data. Dead value checks verify whether the data that passes the station boundary value check exceeds the set dead value check threshold.

[0028] Step S102: Randomly select a rain gauge station as the main station, and filter out multiple neighboring stations within a preset radius centered on the main station; form a local spatial group with each main station and its corresponding multiple neighboring stations to construct multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the main station.

[0029] In this step, based on the acquired latitude and longitude information of each automatic station, one automatic station is first arbitrarily selected as the master station. Then, using the master station as the center and a preset radius (e.g., 15km) as the range, other automatic stations within this preset radius are selected as neighboring stations. The master station and its corresponding neighboring stations together form a local spatial group, which contains n automatic stations. For each of the N automatic stations across the province, the above steps are performed sequentially, thereby constructing a local spatial group corresponding to each of the N automatic stations as the master station.

[0030] Step S103: For each local spatial group, the correct data in the initial quality control observation precipitation data corresponding to the local spatial group is checked by neighboring stations to obtain the neighboring station detection results containing correct data and suspicious data.

[0031] In this step, based on the observed precipitation data corresponding to the master station in the current local spatial group, the precipitation amount of the master station in the current local spatial group is calculated; based on whether the precipitation amount of the master station in the current local spatial group is 0, it is determined whether the observed precipitation data corresponding to the master station is correct or suspicious; if the observed precipitation data corresponding to the master station is determined to be suspicious, a first-level comparison is performed on the precipitation amount of the master station in the current local spatial group to obtain the first-level comparison result; based on the first-level comparison result, it is determined whether the observed precipitation data corresponding to the master station is correct or abnormal; the first-level comparison involves comparing the precipitation amount of the master station in the current local spatial group with multiple different thresholds. The process involves several steps: First, a comparison is performed. If the observed precipitation data corresponding to the main station is determined to be anomalous, the Grubbs method is used to evaluate the observed precipitation data corresponding to the main station based on the cumulative precipitation of all neighboring stations within a preset radius over a preset time period, resulting in a evaluation result. If the evaluation result indicates that the observed precipitation data corresponding to the main station is anomalous, a second-level comparison is performed on the precipitation data of the main station in the current local spatial group, resulting in a second-level comparison result. Based on the second-level comparison result, it is determined whether the observed precipitation data corresponding to the main station is correct or anomalous. The second-level comparison involves comparing the precipitation data of the main station in the current local spatial group with multiple different thresholds.

[0032] The above method, based on whether the precipitation data of the master station in the current local spatial group is zero, determines whether the observed precipitation data of the master station is correct or questionable. This includes: Determine if the cumulative precipitation at the main station in the current local spatial group is 0 within a preset time. If the cumulative precipitation is 0, obtain the number of first neighboring stations with cumulative precipitation greater than 0 within the preset time and the number of second neighboring stations with hourly precipitation greater than 0 within the preset time. Construct a first discrimination condition based on the number of first and second neighboring stations. If the observed precipitation data corresponding to a neighboring station meets the first discrimination condition, determine that the observed precipitation data corresponding to the main station is suspicious data. If the observed precipitation data corresponding to a neighboring station does not meet the first discrimination condition, determine that the observed precipitation data corresponding to the main station is positive. The system first obtains accurate data; it then counts the number of times within a preset time period that the precipitation at the national station and neighboring stations in the current local spatial group exceeds a first preset value, and constructs a second discrimination criterion based on these times; if the observed precipitation data corresponding to the national station or neighboring stations does not meet the second discrimination criterion, the observed precipitation data corresponding to the main station is judged as suspicious data; if the observed precipitation data corresponding to neighboring stations meets the second discrimination criterion, the observed precipitation data corresponding to the main station is judged as correct data; if the precipitation at the main station in the current local spatial group is greater than 0, the Grubbs method is used to determine whether the observed precipitation data corresponding to the main station is correct data or suspicious data.

[0033] Based on the first-level comparison results, the above determination of whether the observed precipitation data corresponding to the main station is correct or abnormal includes: If the first-level comparison result indicates that the current precipitation at the main station is less than the first threshold, then it is determined whether there was precipitation at the nearest national station within the preset range in the current hour and the previous hour. If there was precipitation, the observed precipitation data corresponding to the current main station is determined to be correct. The first number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the first number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the first-level comparison result indicates that the current precipitation at the main station is less than the second threshold, then it is determined whether there was precipitation at the nearest national station within the preset range in the current hour and the previous hour. If there was precipitation and at least one hourly precipitation is greater than or equal to the second preset value, then the current main station is determined to be correct. The observed precipitation data corresponding to the station is correct; count the second number of nearby stations with precipitation in the current hour or the previous hour, and determine whether the observed precipitation data corresponding to the main station is correct or abnormal based on the second number; if the first-level comparison result indicates that the current precipitation of the main station is less than the third threshold, then determine whether there is precipitation in the current hour and the previous hour of the nearest national station within the preset range. If there is precipitation and there is a time-based precipitation greater than or equal to the third preset value, then the observed precipitation data corresponding to the current main station is determined to be correct; count the third number of nearby stations with precipitation in the current hour or the previous hour, and determine whether the observed precipitation data corresponding to the main station is correct or abnormal based on the third number.

[0034] The above-mentioned determination of whether the observed precipitation data corresponding to the main station is correct or abnormal based on the first quantity includes: Count the first number of precipitation events recorded at nearby stations in the current hour or the previous hour. When the number of adjacent stations is 1, 2, or 3, that is, if If the observed precipitation data corresponding to the current main station is correct, then the data is considered correct; if the number of neighboring stations is greater than 3, then... If so, then the observed precipitation data corresponding to the current main station is considered correct. This indicates the total number of rain gauges in the current local spatial group.

[0035] The above-mentioned determination of whether the observed precipitation data corresponding to the main station is correct or abnormal based on the second quantity includes: Count the second number of stations with precipitation in the current hour or the previous hour. When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the nearby stations is greater than or equal to 1.0 mm, the observed precipitation data for the current main station is considered correct; if the number of nearby stations is greater than 3, then... If the hourly precipitation at one of the stations is greater than or equal to 1.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0036] The above-mentioned third quantity is used to determine whether the observed precipitation data corresponding to the main station is correct or abnormal, including: Count the number of stations with precipitation in the current hour or the previous hour. When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 5.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If the hourly precipitation at one of the stations is greater than or equal to 5.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0037] Based on the second-level comparison results, the above determination of whether the observed precipitation data corresponding to the main station is correct or abnormal includes: If the second-level comparison result indicates that the current main station's precipitation is less than the fourth threshold, then it is determined whether the three-hour cumulative precipitation of the nearest national station within the preset range is greater than 0. If the three-hour cumulative precipitation is greater than 0, then the observed precipitation data corresponding to the current main station is determined to be correct. The fourth threshold is then counted for neighboring stations with three-hour cumulative precipitation greater than 0. Based on this fourth threshold, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the second-level comparison result indicates that the current main station's precipitation is less than the fifth threshold, then the three-hour cumulative precipitation of the nearest national station within the preset range is counted. If the three-hour cumulative precipitation is greater than or equal to the first cumulative precipitation threshold, then the observed precipitation data corresponding to the current main station is determined to be correct. The third threshold is then counted for neighboring stations with three-hour cumulative precipitation greater than or equal to the fifth threshold. The fifth threshold is set at 0. Based on this fifth threshold, the observed precipitation data corresponding to the main station is determined to be correct or abnormal. If the second-level comparison result indicates that the current precipitation at the main station is less than the sixth threshold, the cumulative precipitation over three hours at the nearest national station within the preset range is calculated. If the cumulative precipitation over three hours is greater than or equal to the second cumulative precipitation threshold, the observed precipitation data corresponding to the current main station is determined to be correct. The sixth threshold is set at the number of neighboring stations whose cumulative precipitation over three hours is greater than 0. Based on this sixth threshold, the observed precipitation data corresponding to the main station is determined to be correct or abnormal. If the cumulative precipitation over three hours at the main station is greater than the seventh threshold, and the maximum cumulative precipitation over three hours at neighboring stations is less than the third cumulative precipitation threshold, the observed precipitation data corresponding to the main station is determined to be abnormal.

[0038] The above-mentioned fourth quantity determines whether the observed precipitation data corresponding to the main station is correct or abnormal, including: Count the number of nearby stations with cumulative precipitation greater than 0 over 3 hours. (i.e., the fourth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the observed precipitation data corresponding to the current main station is correct, then the data is considered correct; if the number of neighboring stations is greater than 3, then... If the observed precipitation data corresponding to the current main station is correct, then it is considered that the data is correct.

[0039] Based on the fifth quantity, the above determines whether the observed precipitation data corresponding to the main station is correct or abnormal, including: Count the number of nearby stations with cumulative precipitation greater than 0 over 3 hours. (i.e., the fifth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 2.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If the hourly precipitation at one of the stations is greater than or equal to 2.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0040] Based on the sixth quantity, the above determines whether the observed precipitation data corresponding to the main station is correct or abnormal, including: Count the number of nearby stations with cumulative precipitation greater than 0 over 3 hours. (i.e., the sixth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 9.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If the hourly precipitation at one of the stations is greater than or equal to 9.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0041] Step S104: Select the main station with suspicious data in the neighboring station detection results as the station to be inspected, and select the main station with no neighboring stations within a preset distance or with a number of neighboring stations less than a preset threshold as the station to be inspected.

[0042] In this step, for suspicious data that did not pass the neighbor station detection in step S103, the main station corresponding to the suspicious data is designated as a station to be inspected, and main stations with no neighbor stations within a preset distance or with fewer than a preset threshold number of neighbor stations are also designated as stations to be inspected. For example, if there are no neighbor stations within 10km (i.e., within the preset distance) or the number of neighbor stations is insufficient (less than the preset threshold number), then the process of determining false precipitation and anomalies in radar detection is initiated.

[0043] Step S105: Perform radar detection on the observed precipitation data corresponding to the stations to be inspected, and add the stations to be inspected that are judged as having suspicious or erroneous data by radar detection to the blacklist.

[0044] In this step, the first observed precipitation value of the station to be inspected for one hour and the second observed precipitation value of the station to be inspected for three hours are statistically analyzed. Multiple grid data points are selected centered on the latitude and longitude of the station to be inspected. From the radar quantitatively estimated precipitation values ​​corresponding to each of the multiple grid data points, the target radar quantitatively estimated precipitation with the smallest difference from the observed precipitation data corresponding to the station to be inspected is selected. Also, from the radar quantitatively estimated precipitation values ​​corresponding to each of the multiple grid data points, the maximum radar quantitatively estimated precipitation is selected. The area is divided into a first range, a second range, a third range, a fourth range, a fifth range, and a sixth range based on the first observed precipitation value. If the first observed precipitation value meets the requirements of the first range, then based on the multiple grid data... The radar quantitatively estimated precipitation for each point and the target radar quantitatively estimated precipitation are used to determine zero-value precipitation in the observed precipitation data for the stations to be inspected. Stations with suspected or erroneous data are blacklisted. If the first observed precipitation value meets any of the second, third, fourth, fifth, and sixth ranges, or the second observed precipitation value meets the seventh range, then based on the first observed precipitation value, the second observed precipitation value, the target radar quantitatively estimated precipitation, and the maximum radar quantitatively estimated precipitation, the observed precipitation data for the stations to be inspected are judged as false precipitation and outliers. Stations with suspected or erroneous data are blacklisted.

[0045] The above-mentioned zero precipitation determination can be made by using the quantitative estimation of precipitation by radar corresponding to multiple grid data points and the quantitative estimation of precipitation by the target radar when the first observed precipitation value of the station under inspection is 0 or trace precipitation in one hour. This can be used to determine whether the observed precipitation data corresponding to the station under inspection is really 0 or trace precipitation.

[0046] The above-mentioned determination of false precipitation and outliers can be made when the first observed precipitation value of the station under inspection is greater than 0 in one hour. Based on the first observed precipitation value, the second observed precipitation value, the quantitatively estimated precipitation by the target radar, and the maximum quantitatively estimated precipitation by the radar, it can be determined whether the observed precipitation data corresponding to the station under inspection is false precipitation or an outlier.

[0047] Step S106: Interpolate the data of the stations to be inspected in the blacklist, the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, and obtain the precipitation real-time product of the target rain gauge station.

[0048] In this step, a fixed-point nearest-neighbor grid method is used to interpolate erroneous, suspicious, and missing data in the blacklisted data, initial quality control precipitation observation data, and other abnormal precipitation data to obtain grid interpolated values. These grid interpolated values ​​are then used to replace the erroneous, suspicious, and missing data in the blacklisted data, initial quality control precipitation observation data, and other abnormal precipitation observation data to obtain the actual precipitation product from the target rain gauge station.

[0049] To facilitate understanding by those skilled in the art, a set of preferred embodiments is provided below: Compared to conventional meteorological observation stations, rain gauges, specifically designed for the natural resources and hydrology industries, are often deployed in remote mountainous areas and complex river basins. These stations (rain gauges) experience low maintenance frequency and are subject to numerous field interference factors. In addition to general observation errors, they commonly suffer from specific errors such as missing data, observation sequence drift, blockage of tipping bucket rain gauges, and mismatched observation periods. This combination of errors results in insufficient data integrity, poor temporal continuity, significant observation system bias, low spatiotemporal matching, and inconsistent data in the observed precipitation data. This significantly reduces the accuracy and reliability of long-sequence precipitation data, directly impacting the accuracy and scientific rigor of research and operational work related to watershed hydrological simulation, water resource assessment, and rainstorm and flood early warning.

[0050] Existing precipitation quality control technologies are ill-suited to the observation characteristics and operational scenarios of rain gauges in the hydrology and natural resources sectors, exhibiting significant technical shortcomings and failing to meet the quality control requirements for high-precision precipitation data. Specifically, the following core technical issues exist: 1. Existing quality control technologies have extremely poor industry adaptability. Current three-level quality control algorithms, multi-source quality control algorithms, and verification rules are all based on the observation specifications, equipment characteristics, and application scenarios of conventional meteorological rain gauge stations. They do not address issues such as the mixed standards for rain gauge station construction, inconsistent instrument models, numerous long-sequence observation breakpoints, and frequent occurrences of specific hidden errors. Therefore, their adaptability and relevance to the quality control of precipitation data observed by rain gauge stations in the industry are seriously insufficient.

[0051] 2. Existing multi-source quality control technologies suffer from dual defects in terms of application scenarios and timeliness. Existing multi-source data fusion quality control solutions are mostly applied to the compilation of historical data from meteorological stations, and have not been optimized for watershed monitoring scenarios at industry rain gauge stations. Furthermore, they cannot adapt to the quality control needs of complex terrain and differentiated precipitation levels. In addition, the overall approach is an offline, non-real-time correction mode, and lacks the ability to verify observation data streams online in real time.

[0052] 3. The existing quality control system lacks a quality control mechanism for industry rain gauge stations, making it impossible to efficiently complete the quality control of industry rainfall observation precipitation data. This results in a low availability rate of current industry rainfall observation precipitation data, which directly affects the accuracy and reliability of core hydrological operations such as watershed hydrological simulation, water resources assessment, and rainstorm and flood early warning.

[0053] The purpose of this embodiment is to address the problem of low availability of uncontrolled precipitation data from industry rain gauge stations, providing a complete quality control system for these stations. This system includes a self-detection module and a multi-source data collaborative detection module, which can effectively resolve issues such as false jumps and abnormally low values ​​in industry rain gauge data, accurately identify abnormal data, and create a dynamic blacklist. This blacklist can dynamically, reversibly, and accurately manage abnormal stations based on historical quality information and real-time multi-source quality control results. This embodiment incorporates observed precipitation data from provincial natural resource rain gauge stations and provincial transportation rain gauge stations into the Tianheng real-time quality control scope. After quality control, provincial real-time gridded data interpolation technology is used to interpolate abnormal and erroneous observation data from industry rain gauge stations, effectively improving the overall availability of observed precipitation data and generating standardized secondary-processed meteorological data products, ultimately forming provincial industry service rain gauge station precipitation products (i.e., target rain gauge station precipitation products). The specific contents of this embodiment include the following: I. The self-checking module in this embodiment is used to verify the rationality of the observed precipitation data from provincial natural resource rain gauge stations and provincial transportation rain gauge stations (hereinafter referred to as industry rain gauge stations). The self-checking module includes self-checking functions such as metadata checking, feature value checking, climate limit value checking, station limit value checking, and dead value checking, and each check has a result. Data with a quality control code of 7 needs to be removed when acquiring data. According to relevant industry standards and climate limit values, combined with local station historical data statistics, the observed precipitation data of each rain gauge station is self-checked, and the dead value check threshold is determined for each station according to precipitation intensity. The station thresholds are then gridded to form a high spatial resolution threshold parameter file. Specifically: (1) Data acquisition and preprocessing: The observed precipitation data of the provincial natural resources rain gauge station and the provincial traffic rain gauge station are acquired in real time through the provincial Tianqing interface. The raw data containing precipitation data of the provincial natural resources rain gauge station and the provincial traffic rain gauge station are uniformly converted into a standard data format, including field naming, data type and timestamp format, etc., to form a standardized dataset to be quality controlled (i.e. the raw data after preprocessing).

[0054] (2) Metadata Check: Used to control the timing of the algorithm and handle late-arriving precipitation data. In this embodiment, the metadata check first determines the execution frequency of the control algorithm (1 time / hour). The control algorithm starts executing half an hour after the hour. For late-arriving and missing data, it prompts the stations with missing data to supplement the data and repair the instruments, and no further quality control is performed for that time period. Specifically, the input of this step is the preprocessed raw data. It checks the temporal continuity and completeness of the preprocessed raw data, and the output is to mark the missing and late-arriving times and generate a list of valid time data (i.e., the metadata check result). This list serves as the input for the next step to ensure that only valid time data is checked subsequently.

[0055] (3) Feature Value Check: This step identifies invalid feature values ​​in the observed precipitation data and removes invalid data samples. It uses meteorological element feature values ​​according to the domestic meteorological industry standard QX / T515-2019. Invalid data includes 999999 (missing measurement value) and 999998 (no data or no observation). These feature values ​​should be directly excluded from the list of valid time-time data and not participate in subsequent quality control steps. Specifically, the input for this step is the data in the list of valid time-time data. Invalid data samples with feature values ​​of 999999 (missing measurement data) and 999998 (no observed precipitation data) are removed, and the cleaned valid observation dataset is output as the input for the next step.

[0056] (4) Climate Limit Value Check: This check verifies the reasonableness of precipitation data by verifying whether it exceeds the meteorological extreme value range. The climate limit value check threshold uses the latest 1-hour precipitation extreme value published by the World Meteorological Organization (WMO), which is 305 mm. Observed precipitation values ​​exceeding the climate limit value are assigned a quality control code of 2 to indicate data errors. Specifically, the input for this step is the cleaned and valid observation dataset. Hourly checks are performed to determine whether the precipitation exceeds the WMO-specified 1-hour precipitation climate extreme value (305 mm). If it does, the quality control code for the observed precipitation data is marked as 2 (indicating erroneous data). Data that passes the climate limit value check is used as the input for the next step.

[0057] (5) Station Boundary Value Check: Used to identify anomalies in single-station precipitation data based on historical extreme values. The boundary value check threshold is calculated station-by-station and month-by-month based on historical 1-hour precipitation data from national stations. 98 extreme values ​​(obtained from the information center interface) are obtained based on historical 1-hour precipitation data from 98 national stations. The station boundary value is calculated as 150% of the extreme value. Regional stations are divided according to administrative regions. Specifically, the input for this step is the data that passes the climate boundary value check. The station boundary value is determined by 150% of the historical 1-hour precipitation extreme value for each national station. The check is performed station-by-station and month-by-month, and data exceeding the station boundary value is marked as suspicious data. The data that passes the station boundary value check is used as the input for the next step.

[0058] The aforementioned national stations can be national-level meteorological observation stations (i.e., national basic meteorological stations), which are high-precision meteorological observation stations certified by national standards, with reliable data quality, and serve as benchmark reference stations for quality control verification.

[0059] (6) Stiffness Check: Used to construct a refined grid precipitation stiffness verification threshold system. The stiffness check threshold is determined station by station according to precipitation intensity, and the station threshold is gridded to a 1°×1° grid to form a high spatial resolution threshold parameter file. The quality-controlled data forms the initial provincial industry service rain gauge precipitation data. Specifically, the input for this step is the data that has passed the station boundary value check. By checking whether the precipitation has remained unchanged for a long time (stiffness), the data that exceeds the stiffness check threshold is marked as abnormal data. After all the data has undergone stiffness check, the initial quality control observation precipitation data is finally obtained.

[0060] The aforementioned dead value check threshold can be dynamically determined based on the actual precipitation conditions at the stations (i.e., rain gauges). The maximum duration of consecutive identical precipitation values ​​in historical data is statistically analyzed for each station according to precipitation intensity level (such as 0 to 1 mm / h, 1 to 7 mm / h, and 7 to 15 mm / h, etc.), and this duration is used as the dead value check threshold for that station at that precipitation intensity level.

[0061] The above-mentioned station thresholds can be the dead value check thresholds corresponding to each station. Each rain gauge station calculates its corresponding dead value check threshold (i.e., the maximum duration of continuous identical values ​​allowed for that station under that precipitation intensity) based on its historical observed precipitation data.

[0062] Each of the above five inspection steps generates a quality control mark (correct or incorrect), and the most stringent principle is used for final integration: if any step (2) to step (6) determines that the observed precipitation data is incorrect, the final quality control result is incorrect; if all steps pass, the observed precipitation data is marked as correct. The integrated quality control result serves as the final output of the self-testing module and also provides preliminary screening results for the multi-source data collaborative testing module.

[0063] II. The multi-source data collaborative detection module in this embodiment combines simultaneous radar quantitative precipitation estimation and observed precipitation data from nearby stations to further evaluate the results of the self-detection module (i.e., the initial quality control observed precipitation data). This module includes two methods: nearby station detection and radar detection. The final quality-controlled data forms a provincial-level industry service rain gauge precipitation data product. Specifically: 1. Nearby Station Detection: Considering the horizontal range of small-scale structures in surface precipitation and the distribution density of automatic weather stations, near-station detection is performed using data from nearby stations within a certain radius. Local spatial analysis and the Grubbs method are used to quickly filter hourly and cumulative precipitation (determined through the precipitation process), enabling long-term dynamic quality control. Nearby station detection includes the following specific steps: (1) Based on the latitude and longitude information of each automatic station, firstly, arbitrarily select one automatic station as the master station; then, with the master station as the center and a preset radius (e.g., 15km) as the range, filter out other automatic stations within the preset radius as neighboring stations. The master station and its corresponding neighboring stations together form a local spatial group, which contains n automatic stations. For the N automatic stations in the province, traverse each automatic station in turn and perform the above steps, thereby constructing the local spatial group corresponding to each of the N automatic stations as the master station.

[0064] The aforementioned automatic station can refer to any automatic station selected as the center for quality control judgment in the current local space group. In this embodiment, it can be a rain gauge station.

[0065] (2) For each local spatial group, the observed precipitation data of the main station are classified into two categories based on whether the precipitation value of the main station is 0 mm. The specific discrimination rules and judgment process are as follows: When the precipitation at the main station is 0 mm, the following judgment logic is executed with a 10 km range from the target location as a constraint: Determine the cumulative precipitation at the main station over the past 12 hours. If the cumulative precipitation over 12 hours is 0, and the precipitation over the next three hours after 12 hours is also 0, proceed to the next step.

[0066] Calculate the number of nearby stations with a 6-hour cumulative precipitation greater than 0 mm (m, i.e., the number of the first nearby stations) and the number of nearby stations with an hourly precipitation greater than 0 mm in the first 6 hours (m1, m2, ..., m6, i.e., the number of the second nearby stations). For the first criterion: if m and m1 to m6 are all equal to... (Or when m is greater than 4, m and m1 to m6 are all greater than or equal to 4.) This is because when the number of neighboring stations is small, m and m1 to m6 are all greater than or equal to 1. (In the case of) the recent 6-hour precipitation at the national station is greater than or equal to 13 mm, (When m is greater than 4) If, within a 6-hour period, all neighboring stations have at least two instances of precipitation exceeding 7 mm and at least four instances of precipitation exceeding 0 mm per hour within the same 6-hour period, the observed precipitation data corresponding to the current main station is deemed suspicious and proceeds to radar detection and judgment; otherwise, it proceeds to the next step of judgment. The aforementioned instances can be hourly observation instances, i.e., precipitation data points observed once per hour.

[0067] Calculate the number of times within 6 hours that the hourly precipitation at the national station and neighboring stations exceeds 15 mm (i.e., the first preset value). For the second criterion: if the number of times the precipitation at the national station or two neighboring stations exceeds 15 mm is greater than or equal to 2 hours, and greater than or equal to... If all neighboring stations have more than 3 instances of hourly precipitation greater than 7 mm and more than 3 instances of hourly precipitation greater than 0 within 6 hours, then the observed precipitation data corresponding to the current main station is considered suspicious and enters the radar detection and judgment stage.

[0068] When the precipitation at the main station is greater than 0 mm, the Grubbs algorithm is used to detect outliers at neighboring stations. The formula for the Grubbs algorithm is: ; in, Sort the sample values ​​(which can be observed precipitation data) from smallest to largest. It is the first Each sample value It is the sample mean. It is the sample standard deviation, calculated from the main station. At a certain significance level If Greater than the critical value If the sample value is abnormal, then in this embodiment, the value is taken as follows: The purpose of this embodiment is to select stations with abnormal precipitation. The station with the highest or lowest precipitation in the local spatial area (i.e., the rain gauge station) meets the abnormality requirements. The local spatial method and Grubbs method are used to quickly screen hourly precipitation and cumulative precipitation (determined by the precipitation process) to achieve long-term dynamic quality control.

[0069] The threshold is adjusted so that the reliability is relatively low when the number of neighboring stations is small. hour, ,when hour, ,when hour, ,when hour, .

[0070] (3) For the above-mentioned suspicious data, the following judgment steps shall be performed in sequence according to the current precipitation level of the main station (i.e., the first level): If the current precipitation at the main station is less than 1.6 mm (i.e., the first threshold), the following judgment is made: Determine whether there was precipitation in the current hour and the previous hour at the nearest national station within 50km (i.e., within the preset range). If there was precipitation, the observed precipitation data corresponding to the current main station is considered correct. Count the number of nearby stations that had precipitation in the current hour or the previous hour. (i.e., the first quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the observed precipitation data corresponding to the current main station is correct, then the data is considered correct; if the number of neighboring stations is greater than 3, then... If the observed precipitation data at the current main station is correct, then the data is considered correct. Otherwise, the observed precipitation data at the current main station is considered suspicious.

[0071] If the current precipitation at the main station is less than 7mm (i.e., the second threshold), the following judgment is made: Determine if there was precipitation at the nearest national station within 50km in the current hour and the previous hour. If there was precipitation, and at least one hourly precipitation level is greater than or equal to 1.0mm (i.e., the second preset value), the observed precipitation data corresponding to the current main station is considered correct. Count the number of nearby stations that had precipitation in the current hour or the previous hour. (i.e., the second quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 1.0, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If at least one station has an hourly precipitation of 1.0 mm or more, the observed precipitation data for the current main station is considered correct. Otherwise, the observed precipitation data for the current main station is considered suspicious.

[0072] If the current precipitation at the main station is less than 15mm (i.e., the third threshold), the following judgment is made: Determine if there was precipitation at the nearest national station within 50km in the current hour and the previous hour. If there was precipitation, and at least one hourly precipitation level is greater than or equal to 5.0mm (i.e., the third preset value), the observed precipitation data corresponding to the current main station is considered correct. Count the number of nearby stations that had precipitation in the current hour or the previous hour. (i.e., the third quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 5.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If at least one station has an hourly precipitation of 5.0 mm or more, the observed precipitation data for the current main station is considered correct. Otherwise, the observed precipitation data for the current main station is considered suspicious.

[0073] (4) For suspicious data that fails the test in step (3) above, nearby stations are used to detect the precipitation process. The 3-hour cumulative precipitation of the current main station and nearby stations within 15km is calculated and tested using the Grubbs algorithm. The precipitation of the main station is calculated. ,when Greater than the critical value If so, the observed precipitation data corresponding to that main station is judged to be abnormal data.

[0074] (5) For abnormal data that fails the detection of the adjacent stations mentioned above, the following judgment steps shall be performed in sequence according to the current precipitation classification of the main station (i.e., the second classification comparison): If the current precipitation at the main station is less than 3.0 mm (fourth threshold), the following judgment is made: Determine if the 3-hour cumulative precipitation at the nearest national station within 50km is greater than 0. If it is, the observed precipitation data at the current main station is considered correct. Count the number of nearby stations with 3-hour cumulative precipitation greater than 0. (i.e., the fourth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the observed precipitation data corresponding to the current main station is correct, then the data is considered correct; if the number of neighboring stations is greater than 3, then... If the observed precipitation data corresponding to the current main station is correct, then it is considered that the data is correct.

[0075] If the current precipitation at the main station is less than 10mm (the fifth threshold), the following judgment is made: To determine the 3-hour cumulative precipitation at the nearest national station within 50km, if the 3-hour cumulative precipitation at the national station is greater than or equal to 2.0mm (the first cumulative precipitation threshold), the observed precipitation data for the current main station is considered correct. The number of neighboring stations with 3-hour cumulative precipitation greater than 0 is also counted. (i.e., the fifth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 2.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If the hourly precipitation at one of the stations is greater than or equal to 2.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0076] If the current precipitation at the main station is less than 20mm (sixth threshold), the following judgment is made: To determine the 3-hour cumulative precipitation at the nearest national station within 50km, if the 3-hour cumulative precipitation at the national station is greater than or equal to 9.0mm (the second cumulative precipitation threshold), the observed precipitation data corresponding to the current main station is considered correct. The number of neighboring stations with 3-hour cumulative precipitation greater than 0 is also counted. (i.e., the sixth quantity). When the number of adjacent stations is 1, 2, or 3, that is, if If the hourly precipitation at one of the neighboring stations is greater than or equal to 9.0 mm, then the observed precipitation data corresponding to the current main station is considered correct; when the number of neighboring stations is greater than 3, that is, if If the hourly precipitation at one of the stations is greater than or equal to 9.0 mm, then the observed precipitation data corresponding to the current main station is considered correct.

[0077] Secondly, if the current main station's 3-hour cumulative precipitation is greater than 60mm (seventh threshold), and the maximum 3-hour cumulative precipitation of nearby stations is less than 10mm (third cumulative precipitation threshold), then the observed value (i.e., observed precipitation data) corresponding to the current main station is determined to be suspicious (or abnormal), indicating that the precipitation at this station is too high; if there are no nearby stations within 10km (i.e., within the preset distance) or the number of nearby stations is insufficient (less than the preset number threshold), then the radar detection will proceed to the determination of false precipitation and abnormal values.

[0078] 2. Radar Detection: Radar detection utilizes quantitative radar precipitation estimation products to assess the quality of precipitation values. It divides the data into different intervals based on the hourly rainfall recorded by automatic weather stations, applying different comparison thresholds to each interval. For suspicious (or abnormal) precipitation data that fails the neighboring station detection, and for precipitation data from stations with no or insufficient neighboring stations within a 10km radius, it classifies them as zero-value precipitation, false precipitation, or outliers. The specific radar detection process is as follows: Using the latitude and longitude of the industry rain gauge station (i.e., the station currently under inspection) as the center, select eight directions (including east, west, south, north, northeast, northwest, southeast, and southwest). Within a 5km × 5km area, select eight grid data points whose radar quantitatively estimated precipitation is closest to the automatic weather station's precipitation data (i.e., the observed precipitation data of the station currently under inspection). The target radar quantitatively estimated precipitation is denoted as R. The maximum radar quantitatively estimated precipitation among the grid data points is denoted as... (i.e., maximum radar quantitative precipitation estimation). The observed precipitation value of the one-hour automatic weather station (i.e., the station currently under inspection) is recorded as G (i.e., the first observed precipitation value), and the observed precipitation value of the three-hour automatic weather station is recorded as G3 (i.e., the second observed precipitation value). Based on the statistical results, the automatic weather station precipitation G is divided into 6 ranges for quality control: G=0, 0<G<7, 7≤G<15, 15≤G<40, 40≤G<50, and G≥50, with units of mm / h. Based on the latitude and longitude of the rain gauge station, find the radar hourly average precipitation data within a corresponding 5km×5km range. Determine if there is a valid precipitation value within the 5km×5km range (determine if the precipitation value is not 0). If a valid precipitation value exists, proceed to the next step; otherwise, no processing is performed. Compare the valid precipitation value within the 5km×5km range with the hourly observed precipitation value of the automatic weather station to find the closest value R, and the maximum value among the valid precipitation values. The data is divided into different intervals based on the hourly rainfall recorded at industry rain gauge stations. Different comparison thresholds are used in different intervals to obtain the final quality control results. The specific process is as follows: (1) Determination of zero precipitation.

[0079] If the observed precipitation value G=0 or is trace precipitation (i.e., the first range) of the current station to be inspected is 0, then the number of grid points (i.e., grid points) with radar quantitatively estimated precipitation greater than 0.1 mm is determined. If the proportion of grid points with precipitation is less than 1 / 2 of the total, then the observed precipitation data corresponding to the current station to be inspected is considered correct. Otherwise, the value R closest to the radar quantitatively estimated precipitation is determined. If R≤5 mm, then the observed value (i.e., observed precipitation data) corresponding to the current station to be inspected is considered correct; otherwise, the observed value corresponding to the current station to be inspected is considered incorrect, and the current station to be inspected is added to the blacklist, indicating that the station's precipitation value is too low.

[0080] The observed precipitation value in this embodiment can be the observed precipitation data within a preset time period.

[0081] (2) False precipitation and outlier determination.

[0082] If the observed precipitation value G at the current monitoring station is ≥ 0.3 mm, the maximum precipitation value can be quantitatively estimated using radar. Make a judgment, if When G is less than 1.6 mm, the radar reflectivity factor for the current half hour is statistically analyzed within a 40 km × 40 km radius centered on the automatic rain gauge. The average value of the valid (non-negative) radar reflectivity data for this area within the previous and subsequent half hours is calculated. If the maximum average radar reflectivity factor is less than 10, the observation value corresponding to the current station under inspection is determined to be suspicious. When G is less than 7 mm, the average radar reflectivity factor for the current half hour is statistically analyzed within a 20 km × 20 km radius centered on the automatic rain gauge. If the maximum average radar reflectivity factor is less than 15, the observation value corresponding to the current station under inspection is determined to be suspicious. When G is greater than or equal to 7 mm, the observation value corresponding to the current station under inspection is determined to be suspicious. Stations with suspicious data are added to a blacklist.

[0083] like If the condition is met, then proceed directly to the following rule for judgment: 1) If the observed precipitation value at the current station under inspection is 0.3 < G ≤ 7 (second range), determine the difference between the radar quantitatively estimated precipitation value R closest to the station under inspection and the observed precipitation value G at the station under inspection. If the observed value is incorrect, the station will be added to the blacklist and a message will be displayed indicating that the precipitation value is too low; otherwise, the observed value will be correct.

[0084] 2) If the current observed precipitation value at the station under inspection is 7 < G ≤ 15 (third range), determine the difference between the radar quantitatively estimated precipitation value R closest to the station under inspection and the observed precipitation value G at the station under inspection. If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist, and a message will be displayed indicating that the precipitation value at the station is too low. If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist, and a message will be displayed indicating that the precipitation value at the station is too high. If the observed value of the current station to be inspected is deemed suspicious, the station with the suspicious data will be added to the blacklist; otherwise, the observed value of the current station to be inspected will be deemed correct.

[0085] 3) If the current observed precipitation value at the station under inspection is 15 < G ≤ 40 (fourth range), determine the difference between the radar quantitatively estimated precipitation value R closest to the station under inspection and the observed precipitation value G at the station under inspection. If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist, and a message will be displayed indicating that the precipitation value at the station is too low. If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist, and a message will be displayed indicating that the precipitation value at the station is too high. If the observed value of the current station to be inspected is deemed suspicious, the station with suspicious data will be added to the blacklist; otherwise, if the maximum radar reflectivity factor within a 10km radius is less than 25 in the preceding and following half-hours, the observed value of the current station to be inspected is deemed suspicious, the station with suspicious data will be added to the blacklist, and the station's precipitation value will be indicated as too high; otherwise, the observed value of the current station to be inspected will be deemed correct.

[0086] 4) If the current precipitation value observed at the station under inspection is 40 < G ≤ 50 (the fifth range), determine the value R that is closest to the radar quantitative estimate of precipitation at the current station under inspection. If false precipitation is judged as suspicious, the observed value corresponding to the current station to be inspected is determined to be incorrect; if false precipitation is judged as correct, the observed value corresponding to the current station to be inspected is determined to be suspicious, and the stations judged as having suspicious or incorrect data are added to the blacklist. If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist, and a message will be displayed indicating that the precipitation value at the station is too high. and If the observed value of the station to be inspected is incorrect, the station will be added to the blacklist and a message will be sent indicating that the precipitation value of the station is too high. Otherwise, if the maximum radar reflectivity factor within a 10km radius is less than 35 in the preceding and following half-hour intervals, the observed value of the station to be inspected will be considered suspicious, and the station with suspicious data will be added to the blacklist and a message will be sent indicating that the precipitation value of the station is too high. Otherwise, the observed value of the station to be inspected will be considered correct.

[0087] 5) If the observed precipitation value G at the current station under inspection is greater than 50 (sixth range), and the maximum radar reflectivity factor within a 10km radius is less than 40 in the preceding and following half-hour intervals, then the observed value at the current station under inspection is deemed suspicious, and the station with suspicious data is added to the blacklist; otherwise, the maximum precipitation value is quantitatively estimated by radar. Make a judgment, if and If so, the observation value corresponding to the current station to be inspected is determined to be correct; if and If so, the observation value corresponding to the current station to be inspected is determined to be correct; if and If the observation value is correct, the station to be inspected is determined to be correct; otherwise, the station to be inspected is determined to be incorrect, the station is added to the blacklist, and an abnormal precipitation value is indicated.

[0088] 6) If the current precipitation value observed at the station to be inspected (Seventh range) If the maximum radar reflectivity factor within a 10km range and within 3 hours is less than 35, the observation value corresponding to the current station to be inspected is judged to be suspicious. The stations to be inspected with suspicious data will be added to the blacklist and the station's precipitation value will be indicated as too high.

[0089] Third, the dynamic blacklist module in this embodiment is used to implement automatic alarm and information feedback functions. This embodiment implements automatic alarm and information feedback functions for important erroneous data of key elements. It promptly sends the identification results to data quality control duty posts at all levels through automatic telephone voice, SMS, and WeChat reminders, and automatically records quality control feedback information.

[0090] The dynamic blacklist module, based on prior data quality information, temporarily suspends observation stations with abnormally high error rates within a specific time period. This prior data quality information is established based on the quality control results of observed precipitation data from the rain gauge self-detection module and the multi-source data collaborative detection module. If a rain gauge has more than one error record in the past 24 hours, that station is blacklisted. Subsequent data from that station will still undergo self-detection and multi-source data collaborative detection. When the quality of the station's observed precipitation data improves and no longer meets the criteria for being blacklisted, it will be restored to normal use.

[0091] IV. Provincial-level real-time gridded data interpolation technology is employed to interpolate abnormal and erroneous observation data from industry rain gauge stations, effectively improving the overall availability of observation data and generating standardized secondary-processed meteorological data products (i.e., target rain gauge station precipitation data products). Based on the list of abnormal and erroneous stations output by the self-detection module, multi-source data collaborative detection module, and dynamic blacklist module, a real-time gridded data interpolation module is constructed. This module uses industry rain gauge stations identified by quality control as erroneous, suspected of being erroneous and requiring replacement, lacking data due to communication interruption, or temporarily disabled by the dynamic blacklist as interpolation targets. It extracts station number, longitude, latitude, province, and altitude to form a time-based interpolation station file and starts in three modes: scheduled task, data source triggered, and historical backtesting. In data source-triggered mode, the module can dynamically obtain the same list of sites at the same time from the monitoring interface for missing and abnormal sites. If the dynamic list is not available, it automatically reverts to a pre-configured list of sites (which can be a business fault tolerance mechanism; it is a pre-configured static list of sites used to ensure that the interpolation process can continue to run when the dynamic data source is unavailable, reflecting the design considerations for high availability and continuity in real-time business scenarios in this embodiment). In historical backtesting mode, the interpolation sequence is reconstructed hourly according to the start and end times, thus enabling this embodiment to simultaneously adapt to real-time quality control, disaster emergency response, and post-event evaluation and correction scenarios. Specifically, it includes: (1) Spatial legality and uniqueness checks are performed on the stations to be interpolated, and data with duplicate station numbers, missing latitude and longitude, longitudes outside the range of 70° to 140°, or latitudes outside the range of 0° to 60° are removed to prevent erroneous metadata from entering the grid point positioning process. For each time period to be interpolated, the real-time grid point data fixed-point interpolation module generates a calculation task according to UTC and converts it to the corresponding Beijing time to match the real-time grid point product files in the provincial Tianqing catalog. The grid point data used includes: The multi-source fusion precipitation data analysis product (CMPAS) provides hourly precipitation data by province, and the high-resolution land surface data assimilation system (HRCLDAS) provides hourly temperature, humidity, wind pressure, and maximum wind data by province. Precipitation is the core element for interpolation of industry rain gauges, while other meteorological elements are output synchronously as standardized products.

[0092] (2) Based on the latitude and longitude of the station to be interpolated, a coordinate index is constructed. The two-dimensional latitude and longitude axes corresponding to the station location are read in the grid field, and the grid-based estimated value of the station location (i.e., the grid interpolation value) is obtained by taking the value of the nearest neighbor grid point. This method is different from the interpolation method that relies solely on the simple averaging of surrounding stations or manual recording. It can maintain spatial integrity by using the continuously covered real-time grid field when observations around the station fail simultaneously, and avoids excessive smoothing that weakens the center of heavy precipitation. It is suitable for scenarios with abrupt changes in precipitation gradient, such as mountainous areas, watershed boundaries, and local strong convection. For elements such as temperature and air pressure, this module simultaneously completes the unit conversion from Kelvin to Celsius and from Pascal to Hectopascal. For wind field elements, this module reads the U and V wind components respectively, obtains the wind speed through vector synthesis, and converts the arctangent angle into a wind direction value in the range of 0° to 360°.

[0093] The aforementioned method of obtaining values ​​from the nearest neighbor grid point can be achieved by finding the nearest grid point in a regular grid based on the coordinates of the target point and directly reading the value from that grid point for interpolation. However, the method of obtaining values ​​from the nearest neighbor grid point takes values ​​from a high-quality grid field that has undergone complex multi-source fusion analysis. This grid field itself is the output of national-level operational products such as CMPAS (China Regional Multi-Source Fusion Precipitation Analysis Product) and HRCLDAS (High Resolution Land Surface Data Assimilation System). Its creation process integrates multiple data sources, including radar quantitative precipitation estimation, satellite remote sensing, ground automatic station observations, and numerical model background fields, and is not simply station data.

[0094] (3) After obtaining the grid interpolation value, this module introduces an observation-priority fusion mechanism constrained by quality codes. That is, for the same station, the same time, and the same element, if the Tianqing surface hourly observation data exists and the corresponding quality code meets the preset credibility conditions, the original observation value is retained; if the station is determined by the quality control process of this embodiment to be erroneous, missing, communication interrupted, or on a dynamic blacklist, the grid interpolation value is used as the replacement value, and it is distinguished from the original observation value in the product service by the interpolation data identifier or quality control code. Thus, this embodiment avoids the repeated replacement of normal precipitation observation data, and can form traceable quality control post-supplementary values ​​for industry rain gauges that have been identified as abnormal.

[0095] (4) This module first forms an intermediate file from the interpolation results of each element, and then merges multiple elements according to station number, longitude, latitude, province and altitude to generate a secondary processed meteorological data product containing precipitation, temperature, wind speed, wind direction, maximum wind speed, relative humidity and air pressure. The formal product adopts UTF-8 text encoding. The file name includes the province code, product generation time and data time. The first line of the file has a fixed header of "#station number, observation time, latitude, longitude, altitude, precipitation, temperature, wind speed, wind direction, maximum wind speed, relative humidity and air pressure", and outputs with precision of 0.1 mm precipitation, 0.0001° longitude and latitude, 0.1℃ temperature, 0.1 m / s wind speed, 1° wind direction, 1% relative humidity and 0.1 hPa air pressure.

[0096] (5) After the secondary processed meteorological data products are generated, this module encapsulates them by province according to the provincial code, and completes the collection, distribution, decoding and storage of files through the provincial CTS, provincial DPC and national CTS links; at the same time, it writes the file index table and sends the Tianjing DI monitoring information, recording the data code, product time, algorithm scheduling time, processing start and end time, number of interpolation stations, station list name, file name and latitude and longitude legality check results. During the file transmission process, the method of writing temporary files and then atomically renaming them can be used to reduce the risk of half files being collected, so as to realize that the entire process of interpolation products from generation, transmission, storage to service release can be monitored and traced.

[0097] Through the aforementioned real-time grid data interpolation module, this embodiment connects "anomaly identification, dynamic deactivation, real-time interpolation, standard products, and monitoring feedback" into a closed-loop technical solution. On the one hand, abnormal and erroneous observations from industry rain gauge stations are no longer simply processed by rejection, but are provided with physically consistent replacement values ​​by multi-source fusion of real-time grid fields from the same time and spatial location, significantly reducing data gaps during rainstorms and floods. On the other hand, the interpolated values ​​enter the Tianqing service view as independently processed products with identifiable quality identifiers, preserving the boundaries between original observations, quality control identification, and interpolation replacement, facilitating subsequent evaluation of indicators such as deviation and root mean square error using recovered observation data. This embodiment improves the continuity, integrity, and operational availability of precipitation data from industry rain gauge stations under extreme weather, communication interruption, and instrument malfunction scenarios, enhancing the real-time support capabilities for flash flood warnings, watershed flood control scheduling, and emergency decision-making services.

[0098] Reference Figure 2 This application also provides a multi-source collaborative rain gauge precipitation data quality control system, which includes: The data inspection module 100 is used to preprocess the acquired rain gauge observation precipitation data to obtain a standardized dataset for quality control; the data in the standardized dataset for quality control are inspected for metadata, feature values, climate limits, station limits, and dead values ​​to obtain initial quality control observation precipitation data; the initial quality control observation precipitation data includes erroneous or suspicious data as well as correct data; The neighboring station filtering module 200 is used to arbitrarily select a rain gauge station as the main station and filter out multiple neighboring stations within a preset radius centered on the main station; each main station and its corresponding multiple neighboring stations are combined into a local spatial group to construct multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the main station; The nearest station detection module 300 is used to perform nearest station detection on the correct data in the initial quality control observation precipitation data corresponding to each local spatial group, and obtain the nearest station detection results containing correct data and suspicious data. The inspection station construction module 400 is used to select the main station with suspicious data in the detection results of the neighboring stations as the inspection station, and to select the main station with no neighboring stations within a preset distance or with the number of neighboring stations less than a preset number threshold as the inspection station. The radar detection module 500 is used to perform radar detection on the observed precipitation data corresponding to the stations to be inspected, and to blacklist the stations to be inspected that are judged to have suspicious or erroneous data by radar detection. The interpolation processing module 600 is used to interpolate the data of the stations to be inspected in the blacklist, as well as the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, to obtain the precipitation real-time product of the target rain gauge station.

[0099] It should be noted that since the multi-source collaborative rain gauge precipitation data quality control system in this embodiment is based on the same inventive concept as the multi-source collaborative rain gauge precipitation data quality control method described above, the corresponding content in the method embodiment is also applicable to this system embodiment, and will not be described in detail here.

[0100] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with relevant regulations. The acquisition, storage, use and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations.

[0101] Reference Figure 3 This application also provides an electronic device, which includes: At least one memory; At least one processor; At least one program; The program is stored in memory, and the processor executes at least one program to implement the multi-source collaborative rain gauge precipitation data quality control method described in the embodiments of this application.

[0102] This electronic device can be any smart terminal, including mobile phones, tablets, personal digital assistants (PDAs), and in-vehicle computers.

[0103] The electronic devices according to embodiments of this application will now be described in detail.

[0104] The processor 1600 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 1700 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 1700 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1700 and is called and executed by the processor 1600 to execute the multi-source collaborative rain gauge precipitation data quality control method of the embodiments of this application.

[0105] The input / output interface 1800 is used to implement information input and output. The communication interface 1900 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 2000 transmits information between various components of the device (e.g., processor 1600, memory 1700, input / output interface 1800, and communication interface 1900); The processor 1600, memory 1700, input / output interface 1800 and communication interface 1900 are connected to each other within the device via bus 2000.

[0106] This application embodiment also provides a storage medium, which is a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the above-described multi-source collaborative rain gauge precipitation data quality control method.

[0107] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0108] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0109] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0110] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0111] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0112] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0113] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0114] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0115] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0116] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0117] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks. The embodiments of this application have been described in detail above with reference to the accompanying drawings, but this application is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of this application.

[0118] The embodiments of this application have been described in detail above with reference to the accompanying drawings. However, this application is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of this application.

Claims

1. A method for quality control of precipitation data from multi-source collaborative rain gauge stations, characterized in that, The method includes: The acquired precipitation data from rain gauge stations are preprocessed to obtain a standardized dataset for quality control. Metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks are performed on the data in the standardized dataset for quality control to obtain initial quality control precipitation data. The initial quality control precipitation data includes erroneous or suspicious data as well as correct data. Arbitrarily select a rain gauge station as the master station, and filter out multiple neighboring stations within a preset radius centered on the master station; form a local spatial group with each master station and its corresponding multiple neighboring stations to construct multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the master station; the neighboring stations are other rain gauge stations within a preset radius centered on the master station, excluding the master station. For each of the local spatial groups, the correct data in the initial quality control observation precipitation data corresponding to the local spatial group are checked by neighboring stations to obtain neighboring station detection results containing correct data and suspicious data; The main station with suspicious data in the neighboring station detection results is designated as the station to be inspected, and the main station with no neighboring station within a preset distance or with a number of neighboring stations less than a preset threshold is designated as the station to be inspected. Radar detection is performed on the observed precipitation data corresponding to the stations to be inspected, and the stations to be inspected that are judged by radar detection as having suspicious or erroneous data are blacklisted. Interpolation processing is performed on the data of the stations to be inspected in the blacklist, as well as the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, to obtain the precipitation data of the target rain gauge station.

2. The method for quality control of multi-source collaborative rain gauge precipitation data according to claim 1, characterized in that, The process of performing metadata checks, eigenvalue checks, climate limit value checks, station limit value checks, and dead value checks on the data in the standardized dataset to be quality controlled yields initial quality control observation precipitation data, including: Metadata checks are performed on the data in the standardized dataset to be quality controlled to obtain metadata check results; the metadata check is to check the temporal continuity and integrity of the data in the standardized dataset to be quality controlled. The data in the metadata check results are subjected to eigenvalue checks to obtain a cleaned and valid observation dataset; the eigenvalue check is to remove invalid data samples, including those with missing data and no observed precipitation data. The data in the cleaned effective observation dataset are subjected to a climate limit value check to obtain data that passes the climate limit value check; the climate limit value check is to check whether the data in the cleaned effective observation dataset exceeds the set climate limit value check threshold. The data that passed the climate limit value check is then subjected to a station limit value check to obtain data that passed the station limit value check; the station limit value check is to check whether the data that passed the climate limit value check exceeds the set station limit value. The data that passed the station boundary value check were subjected to a dead value check to obtain the initial quality control observation precipitation data; the dead value check was to check whether the data that passed the station boundary value check exceeded the set dead value check threshold.

3. The multi-source collaborative rain gauge precipitation data quality control method according to claim 1, characterized in that, For each of the local spatial groups, the correct data in the initial quality control precipitation data corresponding to the local spatial group are checked by neighboring stations to obtain neighboring station detection results containing both correct and suspicious data, including: Based on the observed precipitation data corresponding to the main station in the current local spatial group, the precipitation amount of the main station in the current local spatial group is calculated; based on whether the precipitation amount of the main station in the current local spatial group is 0, it is determined whether the observed precipitation data corresponding to the main station is correct data or suspicious data. If the observed precipitation data corresponding to the main station is determined to be suspicious, a first-level comparison is performed on the precipitation of the main station in the current local spatial group to obtain the first-level comparison result; based on the first-level comparison result, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal; the first-level comparison is to compare the precipitation of the main station in the current local spatial group with multiple different thresholds; If the observed precipitation data corresponding to the main station is determined to be abnormal, the Grubbs method is used to determine the observed precipitation data corresponding to the main station based on the cumulative precipitation of all nearby stations within a preset radius within a preset time period, and the determination result is obtained. If the discrimination result indicates that the observed precipitation data corresponding to the main station is abnormal, then a second-level comparison is performed on the precipitation of the main station in the current local spatial group to obtain a second-level comparison result; based on the second-level comparison result, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal; the second-level comparison is to compare the precipitation of the main station in the current local spatial group with multiple different thresholds.

4. The multi-source collaborative rain gauge precipitation data quality control method according to claim 3, characterized in that, The step of determining whether the observed precipitation data corresponding to the main station in the current local spatial group is correct or questionable based on whether the precipitation is 0 includes: Determine whether the cumulative precipitation at the main station in the current local spatial group is 0 within a preset time. If the cumulative precipitation is 0, obtain the number of first neighboring stations with cumulative precipitation greater than 0 within the preset time and the number of second neighboring stations with precipitation greater than 0 per hour within the preset time. Construct a first discrimination condition based on the first and second neighboring station numbers. If the observed precipitation data corresponding to the neighboring stations meets the first discrimination condition, determine that the observed precipitation data corresponding to the main station is suspicious data. If the observed precipitation data corresponding to the neighboring stations does not meet the first discrimination condition, determine that the observed precipitation data corresponding to the main station is correct data. The number of times within a preset time period that the precipitation at the national station and the neighboring stations in the current local spatial group exceeds a first preset value is counted, and a second discrimination condition is constructed based on the number of times. If the observed precipitation data corresponding to the national station or the neighboring station does not meet the second discrimination condition, the observed precipitation data corresponding to the main station is judged as suspicious data; if the observed precipitation data corresponding to the national station or the neighboring station meets the second discrimination condition, the observed precipitation data corresponding to the main station is judged as correct data; the national station is a national-level meteorological observation station. If the precipitation data of the master station in the current local spatial group is greater than 0, the Grubbs method is used to determine whether the observed precipitation data corresponding to the master station is correct or suspicious.

5. The multi-source collaborative rain gauge precipitation data quality control method according to claim 3, characterized in that, The step of determining whether the observed precipitation data corresponding to the main station is correct or abnormal based on the first hierarchical comparison result includes: If the first classification comparison result indicates that the precipitation at the current main station is less than the first threshold, then it is determined whether there is precipitation at the nearest national station within the preset range in the current hour and the previous hour. If there is precipitation, then it is determined that the observed precipitation data corresponding to the current main station is correct. The first number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the first number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the first classification comparison result indicates that the current precipitation at the main station is less than the second threshold, then it is determined whether there is precipitation in the current hour and the previous hour at the nearest national station within the preset range. If there is precipitation and there is a time-based precipitation greater than or equal to the second preset value, then it is determined that the observed precipitation data corresponding to the current main station is correct. The second number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the second number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the first classification comparison result indicates that the current precipitation at the main station is less than the third threshold, then it is determined whether there is precipitation in the current hour and the previous hour at the nearest national station within the preset range. If there is precipitation and there is a time-based precipitation greater than or equal to the third preset value, then the observed precipitation data corresponding to the current main station is determined to be correct. The third number of nearby stations with precipitation in the current hour or the previous hour is counted, and based on the third number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. The first threshold, the second threshold, and the third threshold are preset values, and the first threshold, the second threshold, and the third threshold are not equal.

6. The multi-source collaborative rain gauge precipitation data quality control method according to claim 3, characterized in that, The step of determining whether the observed precipitation data corresponding to the main station is correct or abnormal based on the second-level comparison result includes: If the second classification comparison result indicates that the current precipitation at the main station is less than the fourth threshold, then it is determined whether the cumulative precipitation at the nearest national station within the preset range for three hours is greater than 0. If the cumulative precipitation for three hours is greater than 0, then it is determined that the observed precipitation data corresponding to the current main station is correct. The fourth number of neighboring stations with cumulative precipitation for three hours greater than 0 is counted, and based on the fourth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the second classification comparison result indicates that the current precipitation at the main station is less than the fifth threshold, then the cumulative precipitation of the nearest national station within the preset range over three hours is counted. If the cumulative precipitation over three hours is greater than or equal to the first cumulative precipitation threshold, then the observed precipitation data corresponding to the current main station is determined to be correct. The fifth number of neighboring stations with cumulative precipitation over three hours greater than 0 is counted. Based on the fifth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the second classification comparison result indicates that the current precipitation at the main station is less than the sixth threshold, then the cumulative precipitation of the nearest national station within the preset range over three hours is counted. If the cumulative precipitation over three hours is greater than or equal to the second cumulative precipitation threshold, then the observed precipitation data corresponding to the current main station is determined to be correct. The sixth number of neighboring stations with cumulative precipitation over three hours greater than 0 is counted. Based on the sixth number, it is determined whether the observed precipitation data corresponding to the main station is correct or abnormal. If the cumulative precipitation at the main station over three hours is greater than the seventh threshold, and the maximum cumulative precipitation at a nearby station over three hours is less than the third cumulative precipitation threshold, then the observed precipitation data corresponding to the main station is determined to be abnormal; the fourth threshold, the fifth threshold, the sixth threshold, and the seventh threshold are preset values, and the fourth threshold, the fifth threshold, the sixth threshold, and the seventh threshold are not equal.

7. The multi-source collaborative rain gauge precipitation data quality control method according to claim 1, characterized in that, The process of performing radar detection on the observed precipitation data corresponding to the stations to be inspected, and adding the stations corresponding to the radar detection that are determined to have suspicious or erroneous data to the blacklist, includes: The first observed precipitation value of the station to be inspected for one hour is statistically analyzed, and the second observed precipitation value of the station to be inspected for three hours is statistically analyzed. Multiple grid data points are selected centered on the latitude and longitude of the station to be inspected; the target radar quantitatively estimated precipitation with the smallest difference from the observed precipitation data corresponding to the station to be inspected is selected from the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points; and the maximum radar quantitatively estimated precipitation is selected from the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points. Based on the first observed precipitation value, the area is divided into a first range, a second range, a third range, a fourth range, a fifth range, and a sixth range; If the first observed precipitation value meets the first range, then based on the radar quantitatively estimated precipitation corresponding to each of the multiple grid data points and the target radar quantitatively estimated precipitation, the observed precipitation data corresponding to the station to be inspected is judged to be zero-value precipitation, and the station to be inspected corresponding to the judged suspicious data or erroneous data is blacklisted. If the first observed precipitation value meets any one of the second, third, fourth, fifth, and sixth ranges, or if the second observed precipitation value meets the seventh range, then based on the first observed precipitation value, the second observed precipitation value, the target radar quantitatively estimated precipitation, and the maximum radar quantitatively estimated precipitation, the observed precipitation data corresponding to the station to be inspected is judged as false precipitation and outliers, and the stations to be inspected corresponding to the data judged as suspicious or erroneous are blacklisted; the false precipitation refers to the absence of actual precipitation.

8. A multi-source collaborative rain gauge precipitation data quality control system, characterized in that, The system includes: The data inspection module is used to preprocess the acquired precipitation data from rain gauge stations to obtain a standardized dataset for quality control; the data in the standardized dataset for quality control are then subjected to metadata inspection, feature value inspection, climate limit value inspection, station limit value inspection, and dead value inspection to obtain initial quality control precipitation data; the initial quality control precipitation data includes erroneous or suspicious data as well as correct data; The neighboring station filtering module is used to arbitrarily select a rain gauge station as the main station and filter out multiple neighboring stations within a preset radius centered on the main station; each main station and its corresponding multiple neighboring stations are grouped into a local spatial group, thus constructing multiple local spatial groups; the multiple neighboring stations are multiple rain gauge stations adjacent to the main station; the neighboring stations are other rain gauge stations within a preset radius centered on the main station, excluding the main station. The neighbor station detection module is used to perform neighbor station detection on the correct data in the initial quality control observation precipitation data corresponding to each local spatial group, and obtain the neighbor station detection result containing correct data and suspicious data. The station to be inspected module is used to designate the main station with suspicious data in the detection results of the nearby stations as the station to be inspected, and to designate the main station with no nearby stations within a preset distance or with the number of nearby stations less than a preset threshold as the station to be inspected. The radar detection module is used to perform radar detection on the observed precipitation data corresponding to the stations to be inspected, and to blacklist the stations to be inspected that are determined by radar detection to be suspicious or erroneous data. The interpolation processing module is used to interpolate the data of the stations to be inspected in the blacklist, as well as the erroneous or suspicious data and missing data in the initial quality control observation precipitation data, to obtain the precipitation data of the target rain gauge station.

9. An electronic device, characterized in that, It includes at least one control processor and a memory for communicatively connecting to the at least one control processor; the memory stores instructions executable by the at least one control processor, which, when executed by the at least one control processor, enable the at least one control processor to perform the multi-source collaborative rain gauge precipitation data quality control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the multi-source collaborative rain gauge precipitation data quality control method as described in any one of claims 1 to 7.

Images