A method for generating meteorological raster data based on satellite data

Abstract

The present invention belongs to the field of meteorological management technology, and specifically discloses a method for generating meteorological raster data based on satellite data, which performs time scale fluctuation analysis and spatial difference analysis by retrieving historical meteorological data of various regions within the coverage of meteorological observations, and screens out temporally heterogeneous regions and spatially heterogeneous regions to achieve targeted and localized adjustment of temporal and spatial resolution. This method distinguishes the needs of different regions, reasonably allocates computing resources, avoids wasting resources in regions that do not require high-resolution data, and after screening, clarifies the temporal and spatial resolutions required by different regions by combining the temporal and spatial resolutions of historical meteorological observations and the fluctuation and difference characteristics of temporally heterogeneous regions and spatially heterogeneous regions. This method establishes a unified standard for temporal and spatial resolution adjustment, and greatly improves the operability and scientificity of temporal and spatial resolution adjustment.

Description

Technical Field

[0001] The invention belongs to the technical field of meteorological management, and specifically discloses a method for generating meteorological raster data based on satellite data. Background Art

[0002] Meteorological observations have become an indispensable information resource for various regions due to their direct impact on the operational efficiency of many industries, especially the importance of agriculture, forestry and transportation. However, meteorological phenomena usually do not exist in isolation, but show a certain spatial correlation. By converting the scattered observation site data into a gridded (rasterized) form, the distribution characteristics of meteorological variables (such as temperature, humidity, wind speed, etc.) in geographic space can be effectively described, thereby enhancing the spatial continuity and visualization of the data. In addition, a unified raster data format can also help promote data sharing among different institutions and improve data accessibility and interoperability.

[0003] Resolution must be considered when designing and generating meteorological grid data, because resolution directly affects the accuracy of meteorology. Resolution covers temporal resolution and spatial resolution. In view of the fact that the terrain differences in different regions lead to different characteristics of meteorological changes in time and space dimensions, the grasp of meteorological observation resolution now shows obvious regional differences in different regions. For example, the elastic meteorological grid design method disclosed in the Chinese invention patent with the prior art publication number CN112149294B sets different precisions for meteorological data at different grid points in the meteorological grid, and the precision includes time attributes and space attributes, thereby generating meteorological grid data in a coarse and fine combination manner to improve the utility ratio of grid data. This design method does not give a clear setting method when setting the precision of different grid points corresponding to the region, resulting in poor implementation. In addition, it is ignored that not all regions need to adjust the temporal and spatial resolutions. If the resolution of all regions is improved, the computational burden of data processing will increase, especially in the case of limited computing resources, which may lead to unnecessary waste of resources. At the same time, high-resolution data usually takes up more storage space. If the resolution of all regions is improved, the demand for data storage will be greatly increased, resulting in an increase in storage costs. Summary of the invention

[0004] To this end, one purpose of an embodiment of the present application is to provide a method for generating meteorological raster data based on satellite data, by screening out temporally heterogeneous areas and spatially heterogeneous areas from the areas within the coverage of meteorological observations with the help of historical meteorological data, while maintaining the historical observation resolution in other areas, thereby effectively solving the problems mentioned in the background technology.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A method for generating meteorological raster data based on satellite data, comprising the following steps: S1, retrieving historical meteorological data of various regions within the coverage of meteorological observation, and constructing their respective historical data time series sets according to different meteorological parameters.

[0006] S2. Perform time scale fluctuation analysis on the time series data of each meteorological parameter in each region to select the regions that need to improve the time resolution as time heterogeneous regions.

[0007] S3. Locate the geographical location of each region, thereby screening out the neighboring regions corresponding to each region, and then conduct spatial difference analysis on the historical data time series sets of each meteorological parameter corresponding to each region and the historical data time series sets of each meteorological parameter corresponding to its neighboring regions, so as to screen out the regions that need to improve the spatial resolution and record them as spatially heterogeneous regions.

[0008] S4. Set the required temporal resolution and spatial resolution for temporally heterogeneous areas and spatially heterogeneous areas based on the temporal resolution and spatial resolution of historical meteorological observations.

[0009] S5. Arrange multiple meteorological observation points in spatially heterogeneous areas based on the required spatial resolution, and then use meteorological observation terminals to observe meteorological data at meteorological observation points in spatially heterogeneous areas to judge in real time whether the observation times of each meteorological observation point are synchronized.

[0010] S6. When the observation times of meteorological observation points in a spatially heterogeneous area are not synchronized, an adaptive interpolation method is selected for interpolation correction.

[0011] S7. Meteorological grid data are generated in a meteorological grid by inverting meteorological data observed in a temporally heterogeneous region according to a required temporal resolution and meteorological data observed in a spatially heterogeneous region according to a required spatial resolution or interpolation correction data.

[0012] In combination with all the above technical solutions, the positive effects of the present invention are as follows: 1. The present invention retrieves historical meteorological data from various regions within the coverage of meteorological observations to perform time scale fluctuation analysis and spatial difference analysis, and screens out temporally heterogeneous regions and spatially heterogeneous regions to achieve targeted and localized adjustment of spatiotemporal resolution. This method distinguishes the needs of different regions and reasonably allocates computing resources to avoid wasting resources in areas where high-resolution data is not needed. At the same time, improving the resolution only in areas where high-resolution data is needed can effectively reduce the total amount of data, save storage space, and thus improve the efficiency of meteorological raster data generation.

[0013] 2. The present invention clarifies the time resolution and spatial resolution required for different regions by combining the time and space resolution of historical meteorological observations and the fluctuation and difference characteristics of temporally heterogeneous regions and spatially heterogeneous regions. This method establishes a unified standard for adjusting the time and space resolution, greatly improving the operability and scientificity of the time and space resolution adjustment. In addition, the adjustment is based on historical observation data to ensure the reliability and accuracy of the adjustment operation.

[0014] 3. When the present invention conducts meteorological observations in spatially heterogeneous areas according to the required spatial resolution, it judges the synchronization of the observation times of different meteorological observation points at the same period. When it is judged to be out of sync, the interpolation method is selected according to the distribution characteristics of the observation data of different meteorological observation points at the same period, thereby realizing the adaptive interpolation of meteorological data in an asynchronous state, improving the reliability of the interpolated data, ensuring the authenticity and representativeness of the data, and the adaptive interpolation method can reduce unnecessary data processing time and improve the efficiency of data processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0016] Figure 1 The present invention is a flowchart of the steps for implementing the method.

[0017] Figure 2 It is a schematic diagram of the screening of neighboring areas in the present invention.

[0018] Figure 3 This is a flow chart for implementing the dynamic adjustment of time resolution in the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] See also Figure 1 As shown, the present invention proposes a method for generating meteorological raster data based on satellite data, comprising the following steps: S1, retrieving historical meteorological data of various regions within the coverage of meteorological observation, and constructing respective historical data time series sets according to different meteorological parameters.

[0021] The above-mentioned meteorological data include but are not limited to temperature, humidity, wind speed, visibility, air pressure, etc.

[0022] In the feasible implementation of the above scheme, historical meteorological data can be retrieved periodically in stages, and the retrieval must be limited to the current time. Assuming that a retrieval cycle is set to 3 months, the historical meteorological data within 3 months from the current time can be retrieved for the first retrieval to screen the temporally heterogeneous areas and spatially heterogeneous areas, and the historical meteorological data within 3 months from the current time can be retrieved again for the next retrieval. This can achieve dynamic screening of temporally heterogeneous areas and spatially heterogeneous areas. In this way, the system can adjust the temporal and spatial resolution according to the latest data to better adapt to changes in actual conditions, and has a stronger adjustment timeliness. At the same time, compared with obtaining many years of data at one time, staged periodic retrieval can reduce the amount of data that needs to be processed and stored at a single time, thereby reducing storage and processing costs.

[0023] The construction of respective historical data time series sets mentioned above can facilitate the subsequent separate time series analysis of each meteorological parameter.

[0024] The above construction of a historical meteorological data time series usually involves extracting meteorological data by calendar date within the retrieval period. Considering that meteorological parameters will change dynamically within a day, it is recommended to collect data at a specific time of each date (such as 12 noon every day). Such data can be obtained from the meteorological service center of the relevant region or the official meteorological information source.

[0025] S2. Perform time scale fluctuation analysis on the time series data of each meteorological parameter in each region to select the regions that need to improve the time resolution as time heterogeneous regions.

[0026] In the optimized implementation of the above scheme, the time scale fluctuation analysis of the historical data time series set corresponding to each meteorological parameter in each region is implemented as follows: the maximum data is extracted from the historical data time series set corresponding to each meteorological parameter in each region for difference, wherein the maximum data extraction can be achieved by traversing the data of each date, the maximum data includes the maximum data and the minimum data, and the difference result is divided by the time length of the historical data time series set to obtain the volatility of each meteorological parameter.

[0027] Compare the volatility of each meteorological parameter in each region with the set volatility threshold, and use the expression Calculate the weather fluctuation index for each region. Indicates The volatility of meteorological parameters, The number representing the meteorological parameter. , Indicates the number of meteorological parameters, Indicates the volatility threshold set. Indicates Influencing factors of meteorological parameters. The influencing factors are the weights of meteorological parameters. The influencing factors of each meteorological parameter can be quantified by the historical importance of each meteorological parameter. Specifically, it can be evaluated by assessing the degree of influence of different meteorological parameters on a specific target (such as crop yield). In detailed operations, it is necessary to first collect crop yield data and meteorological data for each region in the same period, and then calculate the correlation coefficient through correlation analysis such as Pearson correlation coefficient. Since the value of the correlation coefficient is between -1 and 1, it is necessary to convert the correlation coefficient to between 0 and 1 to obtain the influencing factors of each meteorological parameter.

[0028] It should be noted that the influencing factors of meteorological parameters are not fixed because they may vary in each region.

[0029] The present invention calculates a meteorological fluctuation index by combining the fluctuation rate of each meteorological parameter in each region with a fluctuation rate threshold, thereby reflecting the degree of meteorological fluctuation in each region.

[0030] In the further optimization and implementation of the above scheme, the median meteorological volatility index is extracted from the meteorological volatility index of all regions, and then the meteorological volatility index of each region is compared with the median meteorological volatility index, and the regions with higher meteorological volatility index than the median meteorological volatility index are screened out as regions where the temporal resolution needs to be improved.

[0031] It should be understood that the above scheme helps to distinguish those areas with more severe fluctuations by selecting the median meteorological fluctuation index as a benchmark, which can avoid the influence of extreme values ​​and make the distinction results more objective and robust.

[0032] S3. Locate the geographical location of each region, thereby screening out the neighboring regions corresponding to each region, and then conduct spatial difference analysis on the historical data time series sets of each meteorological parameter corresponding to each region and the historical data time series sets of each meteorological parameter corresponding to its neighboring regions, so as to screen out the regions that need to improve the spatial resolution and record them as spatially heterogeneous regions.

[0033] In the above scheme, the neighboring areas corresponding to each region are screened out and implemented as follows: each region is marked on the map according to its geographical location, where the geographical location can be longitude and latitude coordinates, and a circle is drawn with the geographical location of each region as the center and a limited distance as the radius. For example, the limited distance is 50km, and the area within the circle is the neighboring range of each region.

[0034] The altitude of each region can be obtained based on its geographical location. Specifically, the altitude of each region can be obtained using terrain data or geographic information systems, and a contour line can be drawn based on the altitude of these regions within the vicinity of each region. The contour line drawing function in the geographic information system tool can be used to draw the contour line. Finally, the regions falling on the contour line are selected as the neighboring regions of the region. Figure 2 shown.

[0035] It should be noted that after determining the proximity of each region, the altitude is taken into consideration and contour lines are drawn, and then the regions falling on the contour lines within the proximity are selected as neighboring regions, so that the selected neighboring regions are not only geographically adjacent but also have similar terrain characteristics, which can make the definition of neighboring regions more scientific and reasonable, and avoid the errors caused by selecting neighboring regions based solely on distance; in addition, spatial heterogeneity refers to the characteristics of different regions with significant spatial differences. By considering the altitude and drawing contour lines to select neighboring regions, spatially heterogeneous regions can be screened in regions that are geographically adjacent and have similar terrain characteristics, which greatly narrows the screening range and is conducive to improving the efficiency and accuracy of screening spatially heterogeneous regions.

[0036] As a feasible approach to the above scheme, the spatial difference analysis of the historical data time series sets corresponding to each meteorological parameter in each region and the historical data time series sets corresponding to each meteorological parameter in its neighboring regions is carried out as follows: the historical data time series sets corresponding to each meteorological parameter in each region and its neighboring regions are decomposed into a control group of historical data of each meteorological parameter in the region and its neighboring regions on each date.

[0037] The premise for the above decomposition into control groups is to align the historical meteorological data of each region and its neighboring regions according to the same date, and to construct a control group for each date.

[0038] In the above example, suppose that a region A has a neighboring region B, and the historical data time series set of each meteorological parameter of region A is , the historical data time series set of each meteorological parameter corresponding to the neighboring area B is , in the collection Indicates the temperature, humidity, and wind speed in meteorological parameters. Indicates the date number, , then the above three sets are decomposed into the historical data control groups of temperature, humidity and wind speed in the local area and the neighboring areas on each date, among which the historical data control group of temperature in the local area and the neighboring areas on each date is , … The historical data comparison group of humidity in this area and neighboring areas on each date is , … .

[0039] It should be pointed out that the historical data control group is composed of the local region and a single neighboring region. When there are multiple neighboring regions, the local region and each neighboring region constitute the historical data control group respectively.

[0040] The meteorological difference between this region and neighboring regions is obtained by taking the absolute value of the difference between the meteorological data of this region and neighboring regions in the historical data control group and dividing it by the meteorological data of this region.

[0041] The number of neighboring areas screened out by each area is counted. If a certain area has more than one neighboring area screened out, the distance between the area and each neighboring area is obtained from the map, and the reciprocal of the distance is used as the weight value of each neighboring area. The meteorological difference between the area and each neighboring area is combined with the weight value for weighted average calculation to obtain the meteorological difference between the area and the neighboring areas.

[0042] It is important to understand that using the inverse of the interval distance as the weight value can quantify the proximity relationship. The closer the distance is, the higher the weight value is, reflecting that the neighboring region has a greater impact on the local region. This method can more accurately reflect the proximity relationship between regions and their interactions.

[0043] It should also be explained that the meteorological difference calculated by weighted average can reasonably quantify the meteorological differences between this region and its neighboring regions, and can improve the accuracy and explanatory power of the calculation.

[0044] The average meteorological difference between the region and the neighboring regions on each date is calculated by averaging the meteorological parameters corresponding to the region.

[0045] As a further feasible way of the above scheme, the areas that need to improve the spatial resolution are screened out as follows: the average meteorological difference between each region and the neighboring regions on each date is compared with the set difference threshold, wherein the purpose of setting the difference threshold is to assist in the division of dates. When the average meteorological difference on a certain date reaches the difference threshold, it means that the meteorological difference between this region and the neighboring regions on that date is large, and then the proportion of dates that reach the difference threshold in each region is counted.

[0046] The median proportion is extracted from the proportion of the number of dates that reach the difference threshold in each region, and then the proportion of the number of dates that reach the difference threshold in each region is compared with the median proportion, and the regions with higher proportions than the median are selected as the regions where the spatial resolution needs to be improved.

[0047] It is important to note that the reason for using the median share as a benchmark is the same as the reason for choosing the median meteorological volatility index as a benchmark.

[0048] S4. Set the required temporal resolution and spatial resolution of temporally heterogeneous areas and spatially heterogeneous areas based on the temporal resolution and spatial resolution of historical meteorological observations. The specific settings are as follows: Substitute the temporal resolution of historical meteorological observations and the meteorological fluctuation index of temporally heterogeneous areas into the calculation formula Get the required temporal resolution for temporally heterogeneous regions , where represents the meteorological volatility index for temporally heterogeneous regions, represents the median weather volatility index, It represents the temporal resolution of historical meteorological observations. From this formula, we can see that the higher the meteorological fluctuation index in a temporally heterogeneous region is than the median meteorological fluctuation index, the greater the temporal resolution required in the temporally heterogeneous region.

[0049] Substitute the spatial resolution of historical meteorological observations and the proportion of the number of dates of the difference threshold in spatially heterogeneous regions into the calculation formula Get the required spatial resolution for spatially heterogeneous areas , where Indicates the proportion of dates that reach the difference threshold in spatially heterogeneous areas, represents the median proportion, It indicates the spatial resolution of historical meteorological observations. From this formula, we can see that the higher the proportion of dates reaching the difference threshold in spatially heterogeneous regions is than the median proportion, the greater the spatial resolution required in the spatially heterogeneous regions.

[0050] It should be explained that temporal resolution refers to the time interval between meteorological data observations. High temporal resolution means more frequent data updates, which can better track rapid changes occurring in a short period of time.

[0051] Spatial resolution refers to the actual geographic area represented by each grid cell. High spatial resolution means more detailed geographic coverage and can capture smaller-scale meteorological changes. Generally speaking, in areas with complex and changeable terrain, higher spatial resolution is required to capture local climate changes, while in areas with relatively flat terrain, lower spatial resolution may be sufficient.

[0052] S5. Deploy multiple meteorological observation points in spatially heterogeneous areas based on the required spatial resolution, and then use meteorological observation terminals to observe meteorological data at meteorological observation points in spatially heterogeneous areas. In real time, judge whether the time of observation at each meteorological observation point during the same period is synchronized. The specific judgment is as follows: By comparing the timestamps of observations at each meteorological observation point during the same period, extract the earliest time and the latest time, and calculate the time difference between the two as the deviation duration.

[0053] In the implementation of the above scheme, multiple meteorological observation points are deployed in spatially heterogeneous areas based on the required spatial resolution. In the specific operation, the number of meteorological observation points currently deployed in the spatially heterogeneous area can be obtained first, and then the required spatial resolution is compared with the spatial resolution of historical meteorological observations, and then the number of meteorological observation points that need to be added is obtained according to the difference in spatial resolution, and the meteorological observation points are deployed accordingly. By deploying observation points in this way, it can be ensured that the number of newly added observation points is just right, and neither too many observation points will lead to waste of resources, nor will insufficient observation points affect the resolution and quality of the data.

[0054] In the above implementation example, suppose that there are currently 10 meteorological observation points in a spatially heterogeneous area, distributed in an area of ​​100 square kilometers, that is, there is an observation point for every 10 square kilometers on average. Assuming that the required spatial resolution needs to reach 1 observation point for every 5 square kilometers, the total number of observation points required is: 100 square kilometers × 0.2 / square kilometer = 20. Then the number of newly added observation points is 20-10 = 10.

[0055] The deviation duration is compared with the preset allowable deviation duration. For example, the allowable deviation duration is 0.1s. If the deviation duration does not exceed the allowable deviation duration, the time of the observations at each meteorological observation point during the same period is judged to be synchronized. Otherwise, the time of the observations at each meteorological observation point during the same period is judged to be asynchronous.

[0056] The meteorological observation terminal mentioned above may be a satellite.

[0057] S6. When the observation times of meteorological observation points in the spatially heterogeneous region are not synchronized, an adaptive interpolation method is selected for interpolation correction. The adaptive interpolation method is selected as follows:

[0058] The meteorological parameter data observed at each meteorological observation point during the same period are analyzed for changes to obtain the degree of change of each meteorological parameter, wherein the analysis of the degree of change of the meteorological parameter can be calculated using the variance of the meteorological parameter data observed at each meteorological observation point during the same period.

[0059] Based on the degree of change of each meteorological parameter, the adaptive interpolation method corresponding to each meteorological parameter is extracted from the interpolation selection correspondence relationship.

[0060] In the application of the above scheme, the corresponding relationship of interpolation selection is: linear interpolation is suitable for scenarios where data changes relatively slowly, and polynomial interpolation is suitable for scenarios where data fluctuates to a certain extent. Specifically, a critical value of the degree of change can be set to distinguish the degree of smoothness of data changes, and the degree of change of each meteorological parameter can be compared with the critical value of the degree of change. If the degree of change of a meteorological parameter is less than the critical value of the degree of change, the adaptive interpolation method of the meteorological parameter is linear interpolation, otherwise the adaptive interpolation method of the meteorological parameter is polynomial interpolation.

[0061] The present invention is an intuitive and effective method to distinguish the degree of change of data by variance. A small variance indicates that the data changes slowly and is suitable for linear interpolation. A large variance indicates that the data has a certain fluctuation, and polynomial interpolation may provide a better fitting effect.

[0062] S7. Meteorological grid data are generated in a meteorological grid by inverting meteorological data observed in a temporally heterogeneous region according to a required temporal resolution and meteorological data or interpolation correction data observed in a spatially heterogeneous region according to a required spatial resolution.

[0063] It should be explained that the main purpose of inverting the observation data after meteorological observation is to extract more accurate and meaningful information from the original observation data, so that the meteorological grid data generated after inversion can better serve various meteorological research and practical application needs.

[0064] In the innovative implementation of the above scheme, observing meteorological data in time-heterogeneous areas according to the required time resolution also includes real-time monitoring of load status during the observation process. For details, see the following process: when using a meteorological observation terminal to observe meteorological data in time-heterogeneous areas according to the required time resolution, the CPU usage and memory usage are monitored in real time according to the monitoring period to obtain load status indicators at each monitoring time.

[0065] In the above, CPU usage indicates the percentage of active time of the CPU per unit time. Memory usage indicates the percentage of used memory to total memory.

[0066] The monitoring period is set in the above for dynamic load monitoring, which is convenient for dynamic adjustment of subsequent time resolution. For example, the monitoring period can be set to 2 hours.

[0067] The load index at each monitoring time is calculated by multiplying the load state indicators corresponding to each monitoring time within the monitoring cycle.

[0068] The reason why the load index is calculated by multiplication is that it is very sensitive to significant changes in any one indicator. If any indicator is close to its upper limit, the result of the multiplication will be large even if the other indicator is at a lower level. This allows the load index to quickly reflect when some indicators are close to their limits.

[0069] By comparing the load index at each monitoring time with the load index threshold, it is possible to quickly identify when the system begins to enter an overload state. For example, the load index threshold is 0.7, from which the monitoring time above the load index threshold is identified as the overload monitoring time.

[0070] When the overload monitoring time is identified, the overload duration is obtained. The calculation of the overload duration can quantify the length of time the system is in an overload state, and then the overload duration and the average load index during this period are weighted averaged to obtain the load overload performance value of the monitoring period.

[0071] The above-mentioned load overload performance value provides a comprehensive evaluation indicator that can more intuitively understand the overall load situation of the system during the monitoring period.

[0072] In further innovative implementation, see Figure 3 As shown, real-time monitoring of load status also includes dynamic adjustment of time resolution based on load status monitoring results and meteorological data observation results, as follows: compare the load overload performance value of the monitoring period with the allowable load overload performance value. If the load overload performance value of a certain monitoring period is higher than the allowable load overload performance value, it means that the load status at this time is difficult to maintain the current time resolution, then it is identified that the time resolution needs to be adjusted, otherwise the time resolution does not need to be adjusted.

[0073] It should be added that the setting of the above-mentioned load index threshold and the allowable load overload performance value can be obtained from the load usage instructions according to the load properties and load specifications of the meteorological observation terminal, so as to ensure that the set thresholds and performance values ​​are based on the actual capabilities and design specifications of the equipment.

[0074] When it is identified that the time resolution needs to be adjusted, the observed meteorological parameters are analyzed for changes with the meteorological parameters of several previous observation times to obtain the change difference of each meteorological parameter, wherein the change difference can be obtained by extracting the maximum observation data and the minimum observation data from the meteorological parameters of the current observation and the meteorological parameters of the previous observation time, and then arranging them in order from large to small in terms of change difference, thereby reducing the time resolution of the meteorological parameter with the last change difference, and then continuing to monitor the load status after the reduction to obtain the load index, and comparing it with the load index threshold. If it is still higher than the load index, the meteorological parameters ranked last two from the change difference arrangement result are extracted to reduce the time resolution until the load index obtained by monitoring is no higher than the load index threshold, and the time resolution adjustment is stopped.

[0075] When the present invention conducts meteorological observation in a time-heterogeneous area according to the required time resolution, it can dynamically adjust the required time resolution according to the load status. When the equipment load is high, it can appropriately reduce the time resolution to reduce the equipment burden, prevent overload, and extend the service life of the equipment.

[0076] Furthermore, by giving priority to retaining the time resolution of meteorological parameters with larger variability and reducing the time resolution of meteorological parameters with smaller variability when dynamically adjusting the required time resolution, it is possible to capture important changes in key meteorological parameters while releasing system resources, which has great practical value in actual applications.

[0077] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A method for generating meteorological raster data based on satellite data, characterized in that: The following steps are involved: (1) Retrieve historical meteorological data from various regions within the coverage area of ​​meteorological observation and construct their own time series of historical data based on different meteorological parameters; (2) Performing time scale fluctuation analysis on the historical data time series of each meteorological parameter in each region to screen out the regions that need to improve the temporal resolution as temporally heterogeneous regions; the temporal resolution refers to the time interval between meteorological data observations; (3) Locate the geographical location of each region, thereby screening out the neighboring regions corresponding to each region, and then conduct spatial difference analysis on the historical data time series sets of each meteorological parameter corresponding to each region and the historical data time series sets of each meteorological parameter corresponding to its neighboring regions, so as to screen out the regions that need to improve spatial resolution and record them as spatially heterogeneous regions; (4) Determine the required temporal resolution and spatial resolution for temporally heterogeneous areas and spatially heterogeneous areas based on the temporal resolution and spatial resolution of historical meteorological observations; (5) Deploy multiple meteorological observation points in spatially heterogeneous areas based on the required spatial resolution, and then use meteorological observation terminals to observe meteorological data at meteorological observation points in spatially heterogeneous areas to judge in real time whether the observation times of each meteorological observation point are synchronized; (6) When the observation times of meteorological observation points in a spatially heterogeneous area are not synchronized, an adaptive interpolation method is selected for interpolation correction; (7) Generate meteorological grid data in the meteorological grid by inverting the meteorological data observed in the temporally heterogeneous area at the required temporal resolution and the meteorological data observed in the spatially heterogeneous area at the required spatial resolution or the interpolated corrected data; The spatial difference analysis is as follows: Decompose the historical data time series set of each meteorological parameter of each region and its neighboring regions into a control group of historical data of each meteorological parameter of the region and its neighboring regions on each date; The difference between the meteorological data of the local area and the neighboring areas in the historical data control group is taken and the absolute value is divided by the meteorological data of the local area to obtain the meteorological difference between the local area and the neighboring areas; The number of neighboring regions screened out by each region is counted. If a region has more than one neighboring region screened out, the distance between the region and each neighboring region is obtained from the map, and the reciprocal of the distance is used as the weight value of each neighboring region. The meteorological difference between the region and each neighboring region is combined with the weight value for weighted average calculation to obtain the meteorological difference between the region and the neighboring regions; The average difference between the meteorological parameters of the region and the neighboring regions on the same date is calculated by averaging the meteorological differences between the region and the neighboring regions to obtain the average meteorological difference between each region and the neighboring regions on each date. The screening of the areas where the spatial resolution needs to be improved is achieved as follows: Compare the average meteorological difference between each region and its neighboring regions on each date with the set difference threshold, and then count the percentage of dates that reach the difference threshold in each region; The median proportion is extracted from the proportion of the number of dates that reach the difference threshold in each region, and then the proportion of the number of dates that reach the difference threshold in each region is compared with the median proportion, and the regions with a higher proportion than the median proportion are selected as the regions that need to improve the spatial resolution; The process of setting the required temporal resolution and required spatial resolution of the temporally heterogeneous region and the spatially heterogeneous region is as follows: Substitute the temporal resolution of historical meteorological observations and the meteorological fluctuation index of temporally heterogeneous regions into the calculation formula Get the required temporal resolution for temporally heterogeneous regions , where represents the meteorological volatility index for temporally heterogeneous regions, represents the median weather volatility index, Indicates the temporal resolution of historical meteorological observations; Substitute the spatial resolution of historical meteorological observations and the proportion of the number of dates of the difference threshold in spatially heterogeneous regions into the calculation formula Get the required spatial resolution for spatially heterogeneous areas , where Indicates the proportion of dates that reach the difference threshold in spatially heterogeneous areas, represents the median proportion, Represents the spatial resolution of historical meteorological observations.

2. The method for generating meteorological raster data based on satellite data according to claim 1, characterized in that: The time scale fluctuation analysis of the historical data time series set corresponding to each meteorological parameter in each region is performed to screen out the regions where the time resolution needs to be improved as follows: Extract the maximum value data from the historical data time series set corresponding to each meteorological parameter in each region, and divide the difference result by the time length of the historical data time series set to obtain the volatility of each meteorological parameter; Compare the volatility of each meteorological parameter in each region with the set volatility threshold, and use the expression Calculate the weather fluctuation index for each region. Indicates The volatility of meteorological parameters, The number representing the meteorological parameter. , Indicates the number of meteorological parameters, Indicates the volatility threshold set. Indicates Influencing factors of meteorological parameters; The median meteorological volatility index is extracted from the meteorological volatility index of all regions, and then the meteorological volatility index of each region is compared with the median meteorological volatility index, from which regions with higher meteorological volatility index than the median are selected as regions where the temporal resolution needs to be improved.

3. The method for generating meteorological raster data based on satellite data according to claim 1, characterized in that: The method of screening out the neighboring regions corresponding to each region is implemented as follows: Mark each region on the map according to its geographical location, and draw a circle with the geographical location of each region as the center and the specified distance as the radius. The area within the circle is the neighboring area of ​​each region. The altitude of each region is obtained according to its geographical location, and a contour line is drawn according to the altitude of these regions within the vicinity of each region, and then the regions falling on the contour line are screened out as the neighboring regions of the region.

4. The method for generating meteorological raster data based on satellite data according to claim 1, characterized in that: The process of judging whether the time of observation at each meteorological observation point is synchronized is as follows: By comparing the timestamps of each meteorological observation point observed during the same period, the earliest time and the latest time are extracted, and the time difference between the two is calculated as the deviation duration; The deviation duration is compared with the preset allowable deviation duration. If the deviation duration does not exceed the allowable deviation duration, the time observed at each meteorological observation point during the same period is judged to be synchronized. Otherwise, the time observed at each meteorological observation point during the same period is judged to be asynchronous.

5. The method for generating meteorological raster data based on satellite data according to claim 1, characterized in that: The process of selecting the adaptive interpolation method is as follows: The meteorological parameter data observed at each meteorological observation point during the same period are analyzed for changes to obtain the degree of change of each meteorological parameter; Based on the degree of change of each meteorological parameter, the adaptive interpolation method corresponding to each meteorological parameter is extracted from the interpolation selection correspondence relationship.

6. The method for generating meteorological raster data based on satellite data according to claim 1, characterized in that: The observation of meteorological data in a time-heterogeneous region according to the required time resolution also includes real-time monitoring of the load status during the observation process, as shown in the following process: When using the meteorological observation terminal to observe meteorological data in time-heterogeneous areas according to the required time resolution, the CPU usage and memory usage are monitored in real time according to the monitoring period to obtain the load status indication at each monitoring time; The load index of each monitoring time is calculated by multiplying the load state indicators corresponding to each monitoring time within the monitoring cycle; Compare the load index of each monitoring time with the load index threshold, and identify the monitoring time that is higher than the load index threshold as the overload monitoring time; When the overload monitoring time is identified, the overload duration is obtained, and then the overload duration and the average load index within the duration are weighted averaged to obtain the load overload performance value of the monitoring period.

7. A method for generating meteorological raster data based on satellite data as claimed in claim 6, characterized in that: The real-time monitoring of the load state also includes dynamically adjusting the time resolution based on the load state monitoring results and the meteorological data observation results, see the following process: Compare the load overload performance value of the monitoring period with the allowable load overload performance value. If the load overload performance value of a monitoring period is higher than the allowable load overload performance value, it is identified that the time resolution needs to be adjusted; When it is identified that the time resolution needs to be adjusted, the observed meteorological parameters are analyzed for changes with the meteorological parameters of several previous observation times to obtain the change differences of each meteorological parameter, and the change differences are arranged in order from large to small, thereby reducing the time resolution of the meteorological parameter with the last change difference. After the reduction, the load status monitoring is continued to obtain the load index, and compared with the load index threshold. If it is still higher than the load index, the last two meteorological parameters are extracted from the change difference arrangement results to reduce the time resolution until the monitored load index is no higher than the load index threshold, and the time resolution adjustment is stopped.

Images