CN103984748B - A kind of solar wind data partition, coding and the access method of large scale solar-terrestrial physics - Google Patents

Abstract

The present invention provides a solar wind data subdivision, coding and access method in large-scale solar-terrestrial space. The method includes: receiving the original solar wind data output by a three-dimensional solar wind model; The PDQG-R grid of the corresponding resolution level is automatically constructed, wherein the PDQG-R grid is: when the radial subdivision level is greater than the meridional subdivision level, continue on the basis of the PDQG grid Carry out independent subdivision in the radial direction; code the PDQG-R grid, that is, give each divided sub-grid a unique identifier, so that the coordinates and codes of the grid form a one-to-one correspondence; calculate each Coordinates of the center point of the PDQG‑R grid; according to the calculated coordinates of the center point of the PDQG‑R grid, put the original solar wind data into the PDQG‑R grid, when the original data matches a PDQG‑R grid After that, the attribute value of the original solar wind data is assigned to the grid center point.

Description

A Method for Segmentation, Coding and Access of Solar Wind Data in Large-Scale Solar-terrestrial Space

technical field

The invention relates to the field of rapid organization of mass data of solar-terrestrial space physical fields, in particular to a method for segmenting, encoding and accessing solar wind data in large-scale solar-terrestrial space.

Background technique

The development of geographic information system, especially the development of digital earth technology, has enabled the research of earth science to have a powerful means of data organization and management. In the face of the rapid development of space science, with the massive growth of space detection data, how to organize and manage the detection data in the solar system, how to improve the efficiency of data access and data visualization, has become an urgent need for space scientists. The problems solved have important theoretical research significance and broad application prospects.

Since there is no actual solar wind data of actual detection at present, what the present invention adopts is the data output by the solar interplanetary conservation element solution element (SIP-CESE) three-dimensional solar wind model developed by the SIGMA research group of the State Key Laboratory of Space Weather in my country. Provide technical support for the organization, management, and analysis of the actual detection data of the solar wind.

The space of the entire solar system is huge, and the amount of solar wind data is astonishingly large. In order to improve the retrieval and access speed of massive data, it is necessary to build a two-dimensional subdivision model and generate data of different resolution levels, thereby improving the data organization of the spherical space of the entire solar system Efficiency, so as to achieve the purpose of fast and even real-time rendering of massive data.

At present, there are three typical two-dimensional space subdivision models, namely latitude and longitude subdivision model, regular polyhedron subdivision model and Voronoi spherical adaptive grid model.

To be applied to the subdivision of the solar wind, the polyhedral subdivision model and the adaptive grid model can be excluded. The coordinate calculation of the polyhedral model is relatively complicated, and the organization form is different from the existing surveying and mapping and remote sensing data, so it is difficult to integrate. The sub-model is a subdivision of a single scale, which is difficult to combine with multi-resolution.

Contents of the invention

The object of the present invention is to provide a method for segmenting, encoding and accessing solar wind data in large-scale solar-terrestrial space in order to overcome the above-mentioned problems.

In order to achieve the above purpose, the present invention provides a method for segmenting, encoding and accessing solar wind data in large-scale solar-terrestrial space. The method includes:

Step 101) receiving the original solar wind data output by the three-dimensional solar wind model;

Step 102) subdivide the ecliptic plane or the meridian plane with the PDQG-R grid, and set the LOD level to be subdivided according to the requirements, thereby automatically constructing the PDQG-R grid of the corresponding resolution level, wherein the PDQG-R grid Grid: when the radial subdivision level is greater than the meridional subdivision level, continue to subdivide independently in the radial direction on the basis of the PDQG grid;

Step 103) Coding the PDQG-R grid, that is, giving each divided sub-grid a unique identifier, so that the coordinates of the grid and the coding form a one-to-one correspondence;

Step 104) calculate the coordinate position corresponding to the center point of each PDQG-R grid, and the coordinate position is expressed as: (latitude, longitude, radius);

Step 105) According to the coordinates of the center point of the PDQG-R grid calculated in step 104), the original solar wind data is put into the PDQG-R grid, and when the original data is matched to a PDQG-R grid, the The attribute value of the original solar wind data is assigned to the center point of the grid;

step 106)

Based on the grid code in step 103), the original solar wind data is retrieved, that is, the grid position corresponding to the code is searched, and the content placed in the grid is extracted; it is used for the visual display of multi-resolution data, and different resolutions are given. Visual display of rate-level data.

Optionally, the above step 102) further includes:

Step 102-1) Subdividing the ecliptic plane into a PDQG grid, the specific steps are as follows:

Step 102-1-1) Dividing the ecliptic plane into four parts, each part is a quarter circle with a longitude range of 0° to 90° and a radius of about 1AU (average distance between the sun and the earth);

Step 102-1-2) Use the PDQG grid to recursively subdivide each quadrant face, and for the first subdivision, take the midpoints of the three sides on the quadrant face to obtain 3 new points; Two new points on the two waists of the quadrant are connected to form a parallel, and then the midpoint of the parallel and another new point are connected to form a radial line to form a new sub-quadrant and two sub-quadrilaterals;

Step 102-1-3) The second subdivision is the recursive subdivision of the two types of sub-grids generated for the first time; the sub-triangles can be subdivided according to the first subdivision method, and For the sub-quadrangle subdivision, the quadtree method can be used to get the midpoints of the four sides to obtain 4 new points, connect the midpoints on the two waists with the center of the circle with concentric arcs, and directly connect the midpoints of the two latitude lines to a straight line, Then get 4 new sub-quadrilaterals, and then generate 1 new sub-triangle and 10 quadrilaterals to achieve higher resolution subdivision of the quadrangular surface;

Step 102-1-4) Repeat step 102-1-3) until the resolution meets the set requirements;

Step 102-1-5) When the radial resolution is greater than the meridional resolution, that is, when the radial subdivision level is greater than the meridional subdivision level, continue to subdivide independently in the radial direction on the basis of the PDQG grid to obtain PDQG -R grid.

Further optionally, the specific coding of the above step 103) PDQG-R grid is:

Step 103-1) First, the ecliptic plane divided into four is numbered 0, 1, 2, and 3 counterclockwise, and the quadrants in which each quadrant is located are sequentially determined;

Step 103-2) After the first subdivision, for each quadrant, the outer two quadrilaterals are coded as 2 and 3 in counterclockwise order, and the central triangle is coded as 0;

Step 103-3) After the second subdivision, the coding method of the central triangle is the same as that of the first subdivision unit. The lower left and upper left subquadrons are coded as 0 and 1 respectively, and the lower right and upper right are respectively 2 and 3. By analogy, and every time the subdivision level increases by 1 layer, the code will increase by one bit;

Step 103-4) In the radial direction, each PDQG grid has to be subdivided three times in the radial direction, and 1 is divided into 8, then any grid is coded with an additional code after its own code, and then it is identified that the grid has been subdivided in the radial direction. The specific position of the grid after division;

Among them, the numbers from the center of the circle to the outside are represented by 0 and 1 in sequence, the first subdivision is marked by 0 and 1, the second subdivision is represented by 00, 01, 10, 11 in sequence, and the third subdivision is sequentially represented by 000 , 001, 010, 011, 100, 101, 110, 111 means;

Step 103-5) the number of digits behind the dividing line identifies the subdivision of the radius, and every digit of the remaining Morton codes is a quaternary number not greater than 3, and the corresponding Morton code digits are longer and can be expressed as:

Morton＝q ₁ q ₂ q ₃ ...q _n q _n+1 —p

Among them, q ₁ is the tetrad quadrant identification bit, "—" is the separator, and p is the radial distinguishing identification bit after three radial subdivisions based on the PDQG grid.

Optionally, the above step 105) further includes:

Step 105-1) Starting from the original data, first judge the radius, and find the radius layer corresponding to the grid; in the grid-based radius layer, match the longitude range; specifically:

a) The difference between the radius of the grid center point and the radius of the original sampling point is less than half of the grid radius span;

b) The difference between the longitude of the grid center point and the longitude of the original sampling point is less than half of the longitude span of the grid

If the above two conditions are met at the same time, find the grid corresponding to the original data;

Step 105-2) After matching to the grid, assign the attribute value of the original data to the center point of the grid, cycle all the original data, and map all the original data to the PDQG-R grid one by one.

Optionally, the above step 106) further includes:

Step 106-1) Decode the received grid code, and calculate the grid coordinate position corresponding to the code;

Step 106-2) According to the code, extract the corresponding data from the divided grid.

Due to the grid code, each grid has a unique identifier, and each grid has a unique coordinate, so the grid coordinates can be directly calculated from the grid code, and the grid can also be retrieved Which attribute data is stored in .

The above-mentioned data used in the present invention is the data output by a three-dimensional solar wind model (the solar interplanetary conservation element solution element (SIP-CESE) three-dimensional solar wind model developed by my country's State Key Laboratory of Space Weather Science SIGMA research group). Under coordinates, the three coordinate axes are latitude, longitude, and radius. The solar wind data has the characteristics of irregular sampling. In the sampling space of a sphere centered on the sun, the latitude range is -90° to 90°, and the sampling time is 55 times. The longitude range is 0 to 360°, and the sampling time is 80 times, radial sampling The range is about 1AU (astronomical unit, the mean distance between the sun and the earth, 1AU=149597870.691km), and 154 samples are taken.

Compared with prior art, the technical advantage of the present invention is:

Due to the dense latitude and longitude lines at the two poles of the ordinary latitude and longitude grid, the ratio of the grid divided by the two poles and the right ascension does not converge, resulting in serious deformation of the grid drawing effect. In order to solve this problem, the present invention introduces the idea of degenerate quadtree subdivision, and proposes a new PDQG-R grid model for the two typical planes of the ecliptic plane and the meridian plane, according to the characteristics of solar wind data, and for the grid The model proposes a corresponding encoding scheme.

This subdivision model not only solves the problem of too dense grid near the sun's barycenter, but also meets the requirement that the radial resolution is not synchronized with the longitude (latitude) resolution. Moreover, it can provide data of multiple resolution levels, which significantly improves the efficiency of data retrieval, and is suitable for the subdivision of irregularly sampled data of the solar wind.

Description of drawings

Fig. 1-a is the flow chart of splitting solar wind data provided by the present invention;

Fig. 1-b is the flow chart of solar wind data visualization provided by the present invention;

Figure 2 is a schematic diagram of the ecliptic plane;

Figure 3 is a schematic diagram of the meridian plane;

Figure 4 is a schematic diagram of the three subdivisions of the PDQG grid;

Fig. 5 is a schematic diagram of the three subdivisions of the PDQG-R grid;

Fig. 6 is a schematic diagram of quadrant coding;

Fig. 7 is a schematic diagram of PDQG grid coding;

Fig. 8 is a schematic diagram of three divisions in the radial direction.

detailed description

The method of the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

In the solar-terrestrial space with the sun as the center, in order to study the changes of the magnetic field, temperature and density of ejected particles with the distance from the heliocenter, as well as the change of data on this plane, it is necessary to compare two typical The tangent plane (ie ecliptic plane, meridian plane) constructs a plane grid subdivision model, divides the data into different resolution levels, and encodes the grid, so as to realize the data organization of a huge circular plane, thereby improving the quality of massive data. access efficiency.

The invention supports the rapid organization of massive and large-scale solar wind data in the solar-terrestrial space. On the basis of the subdivision model of the two-dimensional space of the solar wind, a pyramid model is established, which can provide data of different resolution levels. Users can use it according to their own It is necessary to select the data with appropriate resolution.

The main functions of this patent include:

1. Two-dimensional space subdivision model based on solar wind

In order to study the distribution characteristics of solar wind data on the plane that has passed the center of the sun, a new PDQG-R model is proposed for the spatial division of the ecliptic plane and the meridian plane, and for the huge circular plane, that is, the radial subdivision The planar degenerate quadtree meshing model, which not only solves the problem of too dense mesh at the center of the sphere, but also meets the requirement that the radial resolution is greater than the longitude (latitudinal) resolution.

2. The coding scheme corresponding to the subdivision model

For the PDQG-R grid, the improved degenerate Z curve is used for encoding, and the corresponding encoding scheme is proposed.

Morton＝q ₁ q ₂ q ₃ ...q _n q _n+1 —p

Among them, q ₁ is the tetrad quadrant identification bit, "—" is the separator, and p is the radial distinguishing identification bit after three radial subdivisions based on the PDQG grid.

By counting the number of grids after 5 subdivisions, the comparison is as follows:

It can be summarized that when the subdivision level is n, the number of grids generated by quarter-circle subdivision is num, and the following formula is satisfied:

That is, num=1+2 ¹ +...+2 ^2n-1 n≥1

which is

While the number of grids in the common latitude and longitude model is num=4 ⁿ (n>0), it can be calculated that the PDQG grid number compression limit is

When the traditional latitude and longitude grid is used for subdivision, the number of grids increases at a rate of 2 ² . After the degeneration, the number at the center of the ball is significantly reduced, and the ratio of the grid size between the center and the periphery is reduced. It effectively improves the shortcoming of too dense grid at the center of the sun, making the grid size tend to be uniform.

On the basis of the PDQG grid, in order to meet the requirement that the radial resolution is larger than that of the longitude (latitude) direction, it is further subdivided in the radial direction to form a PDQG-R grid, which can not only retain the advantages of the PDQG grid, The grid at the center is not too dense, and the radial data characteristics can be preserved to the greatest extent.

Example

Taking the ecliptic plane as an example and the particle density data output by the solar wind model at a certain moment as an example, the steps of subdivision, encoding, and retrieval are described in detail.

Step 101) receiving the output data of the three-dimensional solar wind model, and extracting the data of the ecliptic plane;

Step 102) Divide the ecliptic plane with PDQG-R grid, set the LOD level to be divided according to requirements, thereby automatically constructing the PDQG-R grid of corresponding resolution level, the PDQG-R grid is:

PDQG grid refers to the plane degenerated quadtree grid model (Plane Degenerated Quadtree Grid), which introduces the idea of degenerated quadtree and combines with ordinary latitude and longitude grids to form a higher quality grid model. In order to solve the problem of inconsistency in two-dimensional resolution, when the radial subdivision level is greater than the meridional subdivision level, continue to subdivide independently in the radial direction on the basis of the PDQG grid to solve the problem of radial resolution and meridional resolution. The problem of rate desynchronization, that is, the PDQG-R grid.

Let’s take the solar ecliptic plane as an example, first introduce the subdivision method of the PDQG grid, and the steps of the subdivision method to be used are as follows:

1) First, the ecliptic plane is divided into four parts, each part is a quarter circle with a longitude range of 0°-90° and a radius range of about 1AU, as shown in Figure 2.

2) Use the PDQG grid to recursively subdivide each quadrant face. For the first subdivision, take the midpoints of the three sides on the quadrant face to obtain 3 new points, and divide the quarter circle two waists 2 new points on the above are connected into a latitude line, and then the midpoint of the latitude line is connected with another new point to form a radial line, forming a new sub-quadrant circle (which can be approximately considered as a triangle) and 2 sub-quadrilaterals, See Figure 4 for details, where the solid line represents the result of the first division.

3) The second subdivision is the recursive subdivision of the two types of sub-grids generated in the first time. Sub-triangles can be subdivided according to the first subdivision method, and for subquadrangle subdivision, the ordinary quadtree method can be used to obtain 4 new points from the midpoints of the four sides, and the two waists The midpoint and the center of the circle are connected by concentric circular arcs, and the midpoints of the two latitudes are directly connected with a straight line, and then 4 new sub-quadrilaterals are obtained, thus producing 1 new sub-triangle and 10 (that is, 2+4+4 ) quadrilaterals to achieve a higher resolution subdivision of the quadrant. The dotted line in Figure 4 represents the second subdivision.

4) Repeat step 3 until the resolution meets the requirements. Among them, the dotted line in Figure 4 represents the result of the third division.

In order to solve the problem of inconsistency in two-dimensional resolution, mainly when the radial resolution is greater than the meridional resolution, that is, when the radial subdivision level is greater than the meridional subdivision level, it can be considered to continue on the basis of the PDQG grid. Independent subdivision to meet the needs of radial resolution and meridional resolution not synchronized. Therefore, the PDQG-R grid is proposed here.

Since the data dimension of the quadrant is 2 ⁵ *2 ⁸ , it can be seen that the sampling density in the radial direction is 2 ³ times that in the meridional direction. Therefore, on the basis of PDQG, continue to subdivide three times in the radial direction to meet the resolution requirement, as shown in Figure 5.

Figure 5 is a schematic diagram of a certain grid subdivided three times in the radial direction based on the three subdivisions of the PDQG grid. The solid line is the first subdivision in the radial direction, and the dashed line is the radial The second subdivision, the dotted line is the third radial subdivision. Thus, a PDQG grid becomes 8 sub-grids. Thus, the PDQG-R grid was generated.

The PDQG-R grid at each level is generated on the basis of the PDQG grid at the corresponding level, and the number of radial subdivisions can also be changed as needed. When the number of radial subdivisions is 0, it is PDQG grid.

This step is to divide the huge circular plane of the ecliptic plane and the meridian plane into small grids of different resolution levels, so as to facilitate the organization and management of the solar wind data of this plane. It can be used to process the actual detection data, or the solar wind data output by the model.

Step 103) encode the PDQG-R grid, the encoding method is:

The code is to give each divided sub-grid a unique identifier, so that the coordinates, attribute data and codes in the grid form a one-to-one correspondence relationship. In order to perform efficient data indexing, the PDQG-R grid The specific encoding rules are as follows:

1) First, the ecliptic plane divided into four is numbered 0, 1, 2, and 3 counterclockwise, and the quadrants of each quadrant are sequentially determined.

2) After the first subdivision, for each quadrangular surface, the outer two quadrilaterals are coded as 2 and 3 in counterclockwise order, and the central triangle is coded as 0, which can be considered as 0 and 1 combined to form 0, as shown in Figure 7a .

3) After the second subdivision, the coding method of the central triangle is the same as that of the first subdivision unit. The lower left and upper left subquadrons are coded as 0 and 1, respectively, and the lower right and upper right are 2 and 3, respectively, and so on. And, every time the subdivision layer increases by one layer, the code increases by one bit accordingly.

4) In the radial direction, each PDQG grid has to be subdivided three times in the radial direction, and 1 is divided into 8, then any grid can add an additional code after its own code to identify the radial subdivision The specific location of the grid, the grid numbered 010 in Figure 8 is one of the eight sub-grids generated after PDQG-R subdivision. It can be separated by a dividing line after encoding itself, and the following digits identify the subdivision of the radius, expressed in binary code, 1 digit represents 1 subdivision, 2 digits represent 2 subdivisions, and each time subdivision, the code increases One bit, the more times of subdivision, the smaller the obtained sub-region.

The numbers from the center of the circle to the outside are represented by 0 and 1 in sequence, the first subdivision is marked by 0 and 1, the second subdivision is represented by 00, 01, 10, 11 in sequence, and the third subdivision is represented by 000 and 001 in sequence , 010, 011, 100, 101, 110, 111 said. The number of radial subdivisions can be determined according to the number of coding bits, and the coding is dynamic, not only hereditary, but also well combined with multi-resolution levels.

5) The number of digits behind the dividing line identifies the subdivision of the radius, which is a binary code, and each digit of the remaining Morton codes is a quaternary number not greater than 3. The longer the corresponding Morton code digits, it can be expressed as:

Morton＝q ₁ q ₂ q ₃ ...q _n q _n+1 —p

Among them, q ₁ is the tetrad quadrant identification bit, "—" is the separator, and p is the radial distinguishing identification bit after three radial subdivisions based on the PDQG grid.

This step is to code each grid, so that the data in the grid corresponds to the code, so as to support data retrieval, support data organization, and provide support for subsequent data applications and even data analysis.

Step 104) Calculate the coordinate position (radius, latitude, longitude) (radius, latitude, longitude) corresponding to the center point of each PDQG-R grid, the following is the calculation method of coordinates in the range of [0, pi/2], And so on for the other quadrants.

Radius＝maxradius*2^(-lod)*(i-1)+maxradius*2^(-lod-1)

lontitude=0.25*pi/lon_num(i)+(0.5*pi/lon_num(i))*(j-1)

Latitude=0

Among them, maxradius is the maximum radius of the ecliptic plane, lod is the lod value of the subdivision level set, lon_num(i) is the number of grids divided by the i-th layer in the radial direction, and j refers to the j-th grid of the i-th layer .

Step 105) Put the original data into the PDQG-R grid, that is, calculate the attribute value of the grid:

Comparing the coordinates of each data in the original data with the central coordinates of the PDQG-R grid, put the original data into the nearest grid, that is, each data is only placed in the nearest grid. The possible results of this are: some grids have only one data; some grids may have multiple data, and the attribute value of the grid can be finally calculated according to a certain suitable algorithm; Just leave it empty to ensure data integrity and consistency.

1) Starting from the original data, first judge the radius, find the grid radius layer where it is located, and based on this layer, to match the longitude range, the conditions to be satisfied are:

a) The difference between the radius of the grid center point and the radius of the original sampling point is less than half of the grid radius span;

b) The difference between the longitude of the grid center point and the longitude of the original sampling point is less than half of the longitude span of the grid.

If both are satisfied at the same time, find the grid corresponding to the original data.

2) After matching to the grid, the attribute value of the original data is assigned to the grid. Loop through all raw data and map all raw data into a PDQG-R grid.

Take the data as an example (L, B, R, den) (3.673205e-068.246700e-014.732900e+07

3.585800e-10), the units are (radian, radian, kilometer), the radius of the farthest sampling data is 0.15391E+09km as the basis for spatial division, and the recursive division is performed based on this range. (A space distance between the sun and the earth is 1AU=1.469E+8km).

After the above steps, the PDQG-R grid corresponding to the data can be found as:

002113-1103.673205e-068.246700e-014.732900e+073.585800e-10

From left to right are: code, latitude, longitude, radius, density attribute value

Similarly, the grid code corresponding to 3.673205e-062.081300e+001.080500e+081.026500e-10 is:

121221-011

Step 106) Retrieve the data of the two-dimensional solar wind model based on the code.

This step is actually a decoding process. At the same time, find out the grid position corresponding to the code, and extract the content placed on the grid.

Take the code 033123-100 as an example to find the data corresponding to the code.

1) Decoding, analyzing the code, and estimating the grid position corresponding to the code. The first digit 0 indicates the first quadrant, within the range of [0,pi/2], the five digits of 33123 indicate that the subdivision level is 5, and the radial subdivision code 100 has 3 digits, indicating that the radial subdivision has been subdivided 3 times, and is Layer 5 of radial subdivision.

The coordinates (latitude, longitude, radius) corresponding to 33123 are calculated below.

The grid center point corresponding to the first digit 3:

Radius(1)=maxradius*2^(-1-1)

Lontitude(1)＝0.25*pi/2+(0.5*pi/2)*(2-1)

The grid center point corresponding to the second digit 3:

Radius(2)=Radius(1)+maxradius*2^(-2-1)

Lontitude(2)＝Lontitude(1)+0.25*pi*2^(-2-1)

The grid center point corresponding to the third digit 1:

Radius(3)＝Radius(2)-maxradius*2^(-3-1)

Lontitude(3)＝Lontitude(2)+0.25*pi*2^(-3-1)

The grid center point corresponding to the 4th bit 2:

Radius(4)=Radius(3)+maxradius*2^(-4-1)

Lontitude(4)＝Lontitude(3)-0.25*pi*2^(-4-1)

The grid center point corresponding to the 5th bit 3:

Radius(5)=Radius(4)+maxradius*2^(-5-1)

Lontitude(5)＝Lontitude(4)+0.25*pi*2^(-5-1)

Adding radial subdivision, the final coordinate value is:

Radius＝Radius(5)-maxradius*2^(-5-1)+maxradius*2^(-5-3-1)*5

Lontitude=Lontitude(5)

The coordinate value is: 3.673205e-061.055379e+001.325670e+08

Due to the grid coding, each grid has a unique identifier, and each grid has unique coordinates, so the grid coordinates calculated in step 104) can be directly retrieved from the grid coding, and can also be Retrieve which attribute data is in the grid by encoding this index. In short, because of the previous steps, this step is the simplest. In fact, the data retrieval can be completed without calculating the coordinates, because the code is unique, and the coordinates and attribute data are associated through the code. But extrapolated coordinates can be used as a function of the present invention.

2) According to the code, the corresponding data can be extracted from the grid.

The original data corresponding to code 003031-001 is, there are 2:

3.673205e-061.129010e+004.899869e+073.491200e-101.381500e+05

3.673205e-061.129010e+004.899869e+073.649800e-101.391600e+05

The original data corresponding to the code 231121@010 is, there is 1:

3.673205E-064.589671E+008.807740E+073.783500E-103.104800E+05

In addition to being used for retrieval, the above-mentioned models and codes can also be used for visual display of multi-resolution data, giving intuitive display of data of different resolution levels, which is convenient for users to intuitively understand the characteristics of data, so as to quickly locate the area of interest. data.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than limit them. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent replacements to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all of them should be included in the scope of the present invention. within the scope of the claims.

Claims (4)

1. a kind of solar wind data partition, coding and the access method of large scale solar-terrestrial physics, methods described are included：

Step 101) receive the original sun wind data that three-dimensional solar wind model is exported；

Step 102) ecliptic plane or meridian plane with PDQG-R grid are subjected to subdivision, the LOD grades for subdivision are set according to demand, So as to build the PDQG-R grid of corresponding resolution grade automatically, wherein the PDQG-R grid is：When radial direction subdivision level is big When warp-wise subdivision level, continue to carry out independent subdivision in radial direction on the basis of PDQG grid；

Step 103) PDQG-R grid is encoded, sub-grid that will be each after subdivision gives unique mark so that net The coordinate and coding of lattice form one-to-one incidence relation；

Step 104) the corresponding coordinate positions of central point of each PDQG-R grids is calculated, the coordinate position is expressed as：(latitude Degree, longitude, radius)；

Step 105) according to step 104) calculate obtained PDQG-R grids central point position coordinates, by original solar wind Data are put into PDQG-R grid, after initial data matches a PDQG-R grid, then by original sun wind data Property value be assigned to the grid element center point；

Step 106) be based on step 103) grid coding retrieve original sun wind data, that is, search the corresponding grid of the coding Position, and the contents extraction placed in the grid is come out；Visualization for multiresolution data is shown, provides different resolutions The display directly perceived of the data of rate grade；

The step 102) further include：

Step 102-1) by the subdivision of ecliptic plane progress PDQG grid, comprise the following steps that：

Step 102-1-1) by one point of ecliptic plane it is four, per being partly that longitude range is 90 ° and radius is ground average departure on the 1st From a quarter disc；

Step 102-1-2) recursion is carried out to each quadrant face with PDQG grid, the 1st subdivision is taken on quadrant face Three sides midpoint, obtain 3 new points；The new point of 2 on the waist of quadrant two is linked to be a parallel, then by the parallel Point is linked to be a radial line with another new point, forms 1 new sub- quadrant and 2 sub- quadrangles；

Step 102-1-3) the 2nd subdivision is then the recursion of the two types sub-grid to the 1st generation；To sub- triangle Shape carries out subdivision by the method for first time subdivision, and for sub- quadrangle grids, then can use quadtree approach to take four edges Midpoint obtains 4 new points, and the midpoint and the center of circle on two waists are connected with concentric arc line, the midpoint of two parallels are connected, then again 4 new sub- quadrangles are obtained, and then just produce 1 new sub- triangle and 10 quadrangles, are realized higher to quadrant face The subdivision of resolution ratio；

Step 102-1-4) repeat step 102-1-3), until resolution ratio meets the requirement of setting；

Step 102-1-5) when radial resolving power is more than warp-wise resolution ratio, i.e., radial direction subdivision level is more than warp-wise subdivision level When, continue, in radially independent subdivision, to obtain PDQG-R grid on the basis of PDQG grid.

2. solar wind data partition, coding and the access method of large scale solar-terrestrial physics according to claim 1, its feature It is, the step 103) specific coding of PDQG-R grid is：

Step 103-1) first by one point for four ecliptic plane according to numbered counter-clockwise 0,1,2,3, each four points are determined successively Quadrant where disc；

Step 103-2) after the 1st subdivision, for each quadrant face, two, outside quadrangle according to encoding successively counterclockwise For 2,3, Center-triangle is encoded to 0；

Step 103-3) after the 2nd subdivision, Center-triangle encodes, sub four side identical with the coding method of the 1st subdivision unit Shape encoded left, upper left are respectively 0,1, and bottom right, upper right are respectively 2,3, and the rest may be inferred, and subdivision level often increases by 1 layer, volume Code-phase should increase by one；

Step 103-4) diametrically, each PDQG grid will be by radially 3 subdivisions, and 1 point is 8, then any one grid Extra-code is added after itself coding, and then identifies the grid particular location after radial direction is segmented；

Wherein, the numbering from the center of circle outward is represented with 0,1 successively, and the 1st subdivision is segmented for the 2nd time with 0,1 mark, successively with 00, 01st, 10,11 represent, the 3rd subdivision is represented with 000,001,010,011,100,101,110,111 successively；

Step 103-5) digit mark radius subdivision position behind cut-off rule, the bits per inch word of remaining Morton code is all no more than 3 quaternary number, corresponding Morton code bit numbers are represented by：

Morton=q₁q₂q₃…q_nq_n+1—p

Wherein, q₁Tetrad quadrant flag, "-" is separator, p be on the basis of PDQG grid after radially 3 subdivisions Radial direction distinguishing identifier position.

3. solar wind data partition, coding and the access method of large scale solar-terrestrial physics according to claim 1, its feature It is, the step 105) further include：

Step 105-1) from initial data, first judge radius, the corresponding radius layer of the grid where finding；Grid is based on again Radius layer, matches longitude range；Specially：

A) difference of the radius of grid element center point and the radius of original sample point is less than the half of the mesh radius span；

B) difference of the longitude of grid element center point and the longitude of original sample point is less than the half of the grid longitude span；

If meeting two above condition simultaneously, the corresponding grid of initial data is found；

Step 105-2) match grid after, then the property value of initial data is assigned to the grid element center point, circulated all Initial data, all initial data are mapped in PDQG-R grid one by one.

4. solar wind data partition, coding and the access method of large scale solar-terrestrial physics according to claim 1, its feature It is, the step 106) further include：

Step 106-1) grid coding of reception is decoded, calculate the coordinate position of the corresponding grid of coding；

Step 106-2) according to coding, corresponding data are extracted from the grid that subdivision is obtained.