Method for inverting ground visibility by using satellite data based on curvelet
By gridding and curvelet transforming satellite data, an estimation function of the mean of the directional coefficient matrix and visibility is established, which solves the problem of low accuracy of ground visibility inversion from satellite data and achieves higher-precision and more adaptable ground visibility monitoring.
Patent Information
- Application Number
- CN202511119078.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-11
- Publication Date
- 2025-10-17
AI Technical Summary
Existing methods for inverting ground visibility based on satellite data have problems such as low inversion accuracy, limited scope of application, and poor adaptability to complex atmospheric conditions.
An inversion method based on curvelet satellite data is adopted. By gridding the satellite band data, performing curvelet transformation, calculating the mean square error of the directional coefficient matrix, and performing regression analysis, an estimation function of the mean value of the directional coefficient matrix and visibility is established.
It improves the accuracy and effect of ground visibility inversion, adapts to complex atmospheric conditions, and meets the needs of real-time and refined monitoring.
Smart Images
Figure CN120804495A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to a meteorological monitoring method, in particular to a method for retrieving ground visibility based on satellite data using curvelet. BACKGROUND
[0002] Visibility is an important indicator of atmospheric transparency, and is also a routine atmospheric monitoring indicator. Visibility has a crucial impact on aviation, navigation, road transportation, military activities, energy development, etc. Low-visibility weather often leads to frequent traffic accidents, posing a serious threat to people's life and property safety. Visibility has clear atmospheric environmental indication significance and scientific characteristics, and has gradually received widespread attention in the field of environmental monitoring. Accurate acquisition of ground visibility index information is of great significance for ensuring traffic safety, reasonable arrangement of production and life, and meteorological research.
[0003] Traditional visibility monitoring mainly relies on ground meteorological observation stations, and visibility data is obtained through manual observation or instrument measurement. However, ground meteorological observation stations are sparsely distributed, especially in remote areas, ocean areas and complex terrain areas, and the coverage is insufficient, making it difficult to comprehensively and accurately reflect the visibility conditions of a large area. In addition, the spatio-temporal resolution of ground observation data is low, which cannot meet the demand for real-time and fine monitoring of visibility.
[0004] With the rapid development of satellite remote sensing technology, it is possible to obtain atmospheric information using satellite data. Satellites have the advantages of wide coverage, high observation frequency and fast data acquisition, and can provide rich data sources for visibility retrieval. The existing method for retrieving ground visibility based on satellite data establishes a functional relationship between the satellite data of a certain wave band and the ground visibility, but the satellite data of a certain wave band is not only affected by the ground visibility, but also affected by the air conditions between the ground and the satellite. And ground visibility mainly reflects the air conditions in the ground plane distance direction, which is affected by the air conditions in the ground plane distance direction, so relying only on the numerical value of the satellite data of a certain wave band (which mainly reflects the air conditions between the ground and the satellite) to retrieve the ground visibility will have the problems of low retrieval accuracy, limited application range and poor adaptability to complex atmospheric conditions. SUMMARY
[0005] The purpose of the present application is to provide a method for retrieving ground visibility based on curvelet satellite data, in order to improve the effect of retrieving ground visibility based on satellite data.
[0006] The purpose of the present application is achieved as follows: A method for retrieving ground visibility based on curvelet satellite data, comprising the following steps: a. performing grid processing on the satellite wave band data with the maximum geometric resolution; b. performing curvelet transform on the satellite band data after gridding processing; c. calculating the mean square deviation of the direction coefficient matrix of each direction of the next maximum scale; d. calculating the mean value of the mean square deviation of all direction coefficient matrices of the next maximum scale after curvelet transform; e. performing regression analysis on the mean value of the mean square deviation of all direction coefficient matrices and the measured value of the visibility; f. establishing an estimation function of the mean value of the mean square deviation of the direction coefficient matrix and the visibility according to the regression analysis result.
[0007] Further, step a is to select cloud-free satellite band data for gridding processing.
[0008] Further, the transform type of the curvelet transform in step b is real value transform, the transform coefficient type is curvelet, the transform scale is the maximum level, and the number of the second scale angle of the curvelet transform is 16.
[0009] Further, the next maximum scale in step c is the value of the maximum scale minus 1.
[0010] The present application uses the numerical difference of the satellite band data with the maximum geometric resolution at different geometric points to inverse the ground visibility, and uses the curvelet transform to transform the satellite data from the spatial domain to the frequency domain, and constructs the functional relationship with the ground measured visibility through the difference size of the frequency value, thereby improving the accuracy and effect of the satellite data based on the inverse ground visibility.
[0011] The present application is mainly applied to the fields of regional atmospheric environmental pollution monitoring, aviation, navigation, highway transportation, military activities, etc. BRIEF DESCRIPTION OF DRAWINGS
[0012] Figure 1 is the flow chart of the present application.
[0013] Figure 2 is the comparison chart of the curvelet inverse and the measured visibility in winter.
[0014] Figure 3 is the comparison chart of the curvelet inverse and the measured visibility in spring, summer and autumn.
[0015] Figure 4 is the regression curve chart in spring.
[0016] Figure 5 is the regression curve chart in autumn.
[0017] Figure 6 is the regression curve chart in winter. DETAILED DESCRIPTION
[0018] The present application will be further described in detail below in combination with the drawings and examples.
[0019] First, the satellite data is acquired within a set time and geographical range.
[0020] The satellite data is high-resolution visible light data of the third band of the Himawari-8 meteorological satellite (representing the numerical value of reflected light energy with a center wavelength of 0.64 um). The data information sent by the AHI sensor carried on the Himawari-8 meteorological satellite mainly includes data of sixteen bands, of which the data of the third band belongs to the visible light band (VIS) with a resolution of 0.5-1 km, mainly used for observing cloud distribution, ground features and weather changes during the day. The nadir geometric resolution of the third band satellite data is 0.5 km, the center wavelength is 0.64 um, and it belongs to red light in the visible light, which is the band with the highest geometric resolution among the sixteen bands of the Himawari-8 satellite data.
[0021] The geographical selection range is: 35° north latitude to 44° north latitude, 110.9° east longitude to 122.5° east longitude.
[0022] The selected time points are distributed throughout the four seasons of the year, and the selected time in spring, summer and autumn is: the integer point time from 11:00 to 13:00 on April 3, 2022, from 10:00 to 12:00 on June 1, 2022, from 10:00 to 13:00 on June 24, 2022, from 10:00 to 12:00 on October 22, 2022, from 10:00 to 12:00 on October 23, 2022, from 10:00 to 11:00 on June 15, 2023, and from 10:00 to 12:00 on June 21, 2023; the selected time in winter is: the integer point time from 10:00 to 12:00 on January 28, 2023, from 10:00 to 13:00 on January 29, 2023, from 10:00 to 12:00 on January 30, 2023, and from 10:00 to 11:00 on February 4, 2023. In this way, there are a total of 21 groups of satellite band data in spring, summer and autumn, and a total of 12 groups of satellite band data in winter.
[0023] The method for retrieving ground visibility based on the satellite data of the Qu-Band satellite according to the present application comprises the following steps: The first step is to grid the satellite band data with the maximum geometric resolution: extract the 33 groups of satellite band data selected above, and select the satellite band data without clouds. Since the satellite band data of the Himawari-8 satellite has the highest geometric resolution of the third band data, the third band data is selected for processing. The selected satellite band data is first grid processed according to the row and column 1160x901 to form grid data.
[0024] Step 2: Perform a curvelet transform on the gridded satellite band data: the transform type is real-valued transform, the transform coefficient type is curvelet, the number of second-scale angles is 16, and the transform scale is the maximum level 7. The minimum number of rows and columns in this embodiment is 901, so the maximum level of the curvelet transform is: .
[0025] Step 3: Count the mean square error of the directional coefficient matrix of each direction of the second largest scale; since the maximum scale is 7, the second largest scale is 6, and the number of angles in the second scale is 16, there are 32 directional coefficient matrices in the sixth scale, so it is necessary to count the mean square error of the coefficient matrices of the 32 directions of the sixth scale.
[0026] Step 4: Calculate the mean of the mean square error of the directional coefficient matrix of the total 32 directions of the sixth scale after the curvelet transform.
[0027] Step 5: Perform regression analysis on the mean of the mean square error of all directional coefficient matrices and the measured visibility value: Compare the mean of the curve direction coefficient of each period with the ground visibility, and obtain Figure 2 and Figure 3 The comparison results are shown, and regression analysis is performed based on the comparison results.
[0028] Step 6: Based on the regression analysis results, establish the estimation function of the mean square error of the directional coefficient matrix and visibility: from Figure 4 It can be seen that through the regression analysis of spring data, the estimation function established is: Moreover, the mean absolute error of the spring estimation function is 1.47 km, and the mean relative error is 5.73%.
[0029] from Figure 5 It can be seen that through the regression analysis of autumn data, the estimation function established is: Moreover, the mean absolute error of the estimation function in autumn is 0.93 km, and the mean relative error is 5.14%.
[0030] from Figure 6 It can be seen that through the regression analysis of winter data, the estimation function established is: Moreover, the average absolute error of the estimation function in winter is 1.71 km, and the average relative error is 9.14%.
[0031] from Figure 2The comparison chart of the winter curve wave inversion and the measured visibility can show that the trend of the satellite data inversion visibility is completely consistent with the trend of the ground measured visibility. Figure 3 The comparison chart of the spring, summer and autumn curve wave inversion and the measured visibility can show that the trend of the satellite data inversion visibility is almost consistent with the trend of the ground measured visibility. Figure 4-6 It can be seen that the relative error of the spring and autumn is about 5%, and the relative error of the winter is 9.14%, so the inversion effect is good, thereby proving the reliability of the inversion method.
Claims
1. A method for inverting ground visibility based on curvelet satellite data, characterized in that: The following steps are involved: a. Gridding the satellite band data with the maximum geometric resolution; b. Perform curvelet transform on the gridded satellite band data; c. Calculate the mean square error of the directional coefficient matrix in each direction of the second largest scale; d. Calculate the mean of the mean square error of all directional coefficient matrices at the next largest scale after the curvelet transform; e. Perform regression analysis on the mean of the mean square error of all directional coefficient matrices and the measured visibility value; f. Based on the regression analysis results, establish the estimation function of the mean square error of the directional coefficient matrix and visibility.
2. The method for inverting ground visibility using satellite data based on curvelet according to claim 1, characterized in that: Step a is to select cloud-free satellite band data for gridding processing.
3. The method for inverting ground visibility using satellite data based on curvelet according to claim 1, characterized in that: The transform type of the curvelet transform in step b is real-value transform, the transform coefficient type is curvelet, the transform scale is the maximum level, and the number of second-scale angles of the curvelet transform is 16.
4. The method for inverting ground visibility using satellite data based on curvelet according to claim 1, wherein: The second largest scale in step c is the value of the largest scale minus 1.