Method for extracting and sequentially tracking and analyzing martian atmospheric vortices based on wind field data

Abstract

The application provides a Mars atmospheric vortex extraction, sequence tracking and analysis method based on wind field data, and comprises the following steps: using long-time sequence wind field data in a Mars atmospheric model product, identifying a Mars vortex center based on grid-based local normalized angular momentum, extracting a Mars vortex boundary according to an out-of-center streamline velocity of the vortex center, tracking a Mars vortex sequence through a loss function minimum value, and finally analyzing shape and property parameter changes in the vortex sequence in combination with a vortex sequence life cycle. The application is suitable for Mars atmospheric wind field data analysis and processing to extract a Mars vortex center, a boundary and track a vortex sequence.

Description

Technical Field

[0001] This invention belongs to the fields of planetary remote sensing and planetary geology, specifically a method for extracting, tracking and analyzing Martian atmospheric vortices based on wind field data. It is applicable to the analysis and processing of Martian atmospheric wind field data to extract Martian vortex centers, boundaries and track vortex sequences. Background Technology

[0002] Mars' atmospheric pressure is less than 1% of Earth's, but its atmosphere exhibits complex and active activity. This includes vortex motions similar to those found in Earth's atmosphere and oceans. Atmospheric vortices are a crucial component of Martian global circulation, significantly influencing the structure and dynamics of the Martian atmosphere. Transient vortices, seasonally recurring cyclones, and annular weather systems have been detected in images taken by Mariner 9, Viking 1, Mars Global Surveyor, and the Mars Reconnaissance Orbiter. These vortices form extensively in the Martian hemisphere and throughout different seasons when the meridional temperature gradient increases and baroclinic conditions develop. In many cases, mesoscale circulations (vortices) on Mars are more pronounced and prevalent than on Earth.

[0003] Previous studies on the identification and extraction of Martian atmospheric vortices have primarily relied on two methods: visual identification based on remote sensing imagery and inference based on atmospheric pressure changes observed by weather stations on in-situ Mars rovers. These two methods have the following limitations in identifying and analyzing Martian vortex sequences:

[0004] (1) It is difficult to obtain continuous, long-term sequence data of vortex objects. The existence and movement of Martian vortices can be identified through remote sensing images of Martian atmospheric cloud maps, and a sudden drop in pressure data obtained from the meteorological station on the Mars rover can also be considered as a characteristic of vortices. However, the cameras on the satellite and the meteorological instruments on the Mars rover cannot obtain full-time sequence data of Martian atmospheric activity covering the entire Mars, making it difficult to continuously and completely identify and spatiotemporally analyze Martian atmospheric vortices. Furthermore, due to the low quality of Martian images, vortex objects are affected by dust storms and clouds, and the identification results often contain errors and omissions.

[0005] (2) It is difficult to track Martian atmospheric vortex sequences and study the morphological and property changes during the vortex's life cycle. After Martian atmospheric vortices are generated, they can move along a certain route and interact with the Martian atmosphere along the way in terms of matter and energy. Such vortices that can exist stably for a certain period of time are called vortex sequences. Previous researchers have found it difficult to continuously observe and track the movement of the same vortex using remote sensing images and Mars rover weather instruments. It is also difficult to extract the morphology and movement route of Martian vortex sequences and study the changing characteristics of parameters such as velocity, radius, and water ice content during the life cycle of vortex sequences. Summary of the Invention

[0006] To overcome the shortcomings of the existing technologies, this invention proposes a method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data. It uses long-term wind field data from Martian atmospheric model products, identifies Martian vortex centers based on locally normalized angular momentum within a grid, extracts Martian vortex boundaries based on the streamline velocity surrounding the vortex centers, and tracks Martian vortex sequences by minimizing a loss function (integrating factors such as distance, time, and morphology). Finally, it studies the changes in morphological and property parameters within the vortex sequence by considering its lifecycle.

[0007] This method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data includes the following five steps:

[0008] Step 1: Data processing of Mars atmospheric model products.

[0009] Numerical simulations based on theoretical models are crucial tools for exploring the variability and physical laws of the Martian atmosphere and space environment. Over the past two decades, several leading international Martian atmospheric models have been continuously improved, and they now possess the preliminary ability to fully simulate the multi-sphere coupling processes of the Martian land surface, atmosphere, and space environment. Martian atmospheric model data spatially covers the entire Martian surface, has high temporal resolution (several hours), and can simulate Martian atmospheric characteristics from the surface to high altitudes (from hours to hundreds of kilometers). Martian atmospheric model products typically include atmospheric wind field data and atmospheric physical parameters (such as temperature, pressure, and water ice content).

[0010] Let the Martian atmospheric model product dataset used in this invention be D={Type| grid, time, height}, with a spatial resolution of r_s (in degrees), that is, dividing the Martian surface into a square grid set grid with a side length of r; a temporal resolution of r_t (in Martian hours), where time is a value of one time profile; and a height resolution of r_h, where height is a value of one layer of height profile. The Martian atmospheric data type stored in dataset D is Type, with values ​​including u, v, and I, representing the wind speed and water ice content in the longitude and latitude directions of the time profile (time) and height profile (height) in the Martian surface grid, respectively.

[0011] Because the Martian atmospheric model product covers the entire Martian surface, its spatial resolution is very low, requiring spatial interpolation of the raw data. However, the Martian atmospheric model product has a high temporal resolution, typically 1-2 Martian hours. Studying Martian atmospheric vortex sequences only requires Martian days. To reduce data volume and workload, temporal extraction of the raw data is necessary. Linear interpolation increases the spatial resolution of the original dataset D by a factor of N, resulting in r_s' = r_s / N. Data from each Martian day in the original dataset is extracted, retaining only one value at the same moment for each Martian day, thus changing the temporal resolution to r_t' (in Martian days). The Martian atmospheric model product dataset after spatial resolution interpolation and temporal resolution extraction is D1 = {Type| grid}. i time j height k}. i is the grid index, i=1, 2, …, n1, where n1 is the total number of grids in Mars; j is the Martian day index in the Mars atmospheric model product data, j=1, 2, …, n2, where n2 is the total number of Martian days in the Mars atmospheric model product data; k is the Martian atmospheric altitude index in the Mars atmospheric model product data, k=1, 2, …, n3, where n3 is the number of atmospheric altitude layers in the Mars atmospheric model product data.

[0012] Step 2: Extraction of candidate centers for Martian atmospheric vortices.

[0013] Based on previous theoretical research and experimental observations, the angular momentum is greatest at the center of an atmospheric vortex. Therefore, the angular momentum in the grid of the Martian atmospheric model product is calculated, and the location of the maximum angular momentum value in a local region is extracted as the center of the Martian atmospheric vortex. Let one Martian day be the time... j For example, the atmospheric height of Mars is height. k The grid points are grid i The wind field data in the corresponding Mars atmospheric model product is u1={u| grid i time j height k}, v1={v| grid i time j height k}, written in vector form as To calculate the normalized angular momentum of a grid point, a neighborhood centered on that grid point needs to be established. Let the side length L of the neighborhood be 2b times the side length of the grid, i.e., L = r_s' × 2b, where b is a positive integer greater than 0. The total number of grids in this neighborhood is num = (2b + 1) × (2b + 1), with the central grid being grid. i The formula for the locally normalized angular momentum (LNAM) is as follows:

[0014] (1),

[0015] in, This represents the relative vector between the m-th grid in the neighborhood and the center grid, gridi. Let be the wind field vector of the m-th grid in the neighborhood.

[0016] Then, a threshold of LNAM in the Martian atmospheric grid is set to T1. All grids with LNAM greater than this threshold are searched, and contour lines are generated. The regions enclosed by these contour lines constitute the candidate region set where LNAM is greater than the threshold T1. The grid with the maximum LNAM value is then searched within this candidate region set; this grid represents the center of the Martian atmospheric wind vortex. Let time (time...) j ) and height k Using this method on the cross-section, num1 vortex centers were calculated and searched, i.e., the set of vortex candidate centers P = {P1, P2, …, P}. num1}

[0017] Step 3: Generation of Martian atmospheric vortex boundaries.

[0018] First, extract the streamlines. A streamline is a line in which the tangent at any point is in the direction of the instantaneous velocity at that point. Let the area enclosed by the closed streamlines be A, and the perimeter be L. p Then the equivalent radius R and the average velocity V_SL of the closed streamline are respectively:

[0019] (2),

[0020] (3),

[0021] According to previous research, the mean velocity of a closed streamline centered on an atmospheric or oceanic vortex gradually increases with its equivalent radius R. Once it reaches its maximum value V_SL_max, the mean velocity of the closed streamline decreases with further increases in the equivalent radius. The closed streamline around the center of the vortex that reaches its maximum mean velocity is used to define the characteristic profile of the vortex, i.e., the vortex boundary.

[0022] Therefore, with the vortex center P i For example, the center P of the vortex is extracted from the inside out at certain distance intervals s. i Find the set of all closed streamlines SL and calculate the equivalent radius and mean velocity for each closed streamline. Locate the closed streamline SL' with the maximum mean velocity (V_SL_max) within the set of closed streamlines SL; this closed streamline SL' is considered the boundary B of the vortex center. iFurthermore, for a well-defined, coherent vortex, the closed streamlines around its center should remain coherent in both space and time. Therefore, when the streamlines around the vortex center P... i The vortex criterion is met when the average velocity (V_SL_End) of the last closed streamline is at least one threshold T2 lower than the maximum average velocity (V_SL_max): V_SL_End <= (1-T2) × V_SL_max. If this criterion is not met, the structure around which the closed curve encircles reaches its maximum velocity at the edge; this structure is not a vortex and P needs to be adjusted. i Exclude vortex candidates from the vortex center set P. Iterate through each candidate vortex center in the vortex center set P, and generate the vortex center set P' = {P1, P2, …, P} using the method described above. num2} and the corresponding vortex boundary set B={B1, B2,…, B num2 In the two sets, the indices of the vortex centers and vortex boundaries are in a one-to-one correspondence, and there are a total of num2 vortex centers and vortex boundaries.

[0023] Finally, the vorticity, or Rossby number (Ro), of the identified vortex is calculated. Vorticity is a measure used in meteorology to study vortices in ocean and atmospheric systems, and its calculation formula is:

[0024] Ro= V_SL_max / (f×R1) (4),

[0025] Where V_SL_max and R1 are the average velocity and equivalent radius of the vortex boundary (closed streamline). f is the magnitude of the Coriolis force at the location of the vortex center. The Coriolis force describes the offset of a particle moving in a straight line relative to the rotating system due to inertia. The final result is the vortex vorticity set Ro = {Ro1, Ro2, …, Ro}. num2}

[0026] Step 4: Tracking the Martian atmospheric vortex sequence.

[0027] Martian atmospheric vortices may exist over multiple Martian days and move along certain routes and directions. The center, shape, and intensity of the same vortex on different Martian days may change over time. Vortex sequence tracing is used to track vortex sequences that evolve over time on a Martian day basis, extracting their trajectories and evolutionary characteristics. Applying the methods in steps two and three to Martian wind field data at the same altitude over n consecutive Martian days yields a set of vortices e(t, i) identified over n Martian days, with corresponding vortex centers and boundaries P(t, i) and B(t, i), respectively, and corresponding equivalent radii R(t, i) and vorticity Ro(t, i). Here, t represents a Martian day, t = 1, 2, ..., n, and i represents the i-th vortex identified on the t-th Martian day.

[0028] The principle of the vortex sequence tracking algorithm is to perform correlation detection between the vortex e(t, i) in the current time step (current Martian day) t (t>1) and each vortex e(t-1, j) extracted in the previous time step t-1 (previous Martian day), select the vortex with the strongest correlation as the successor of vortex e(t-1, j), and establish or update the vortex sequence.

[0029] Let es(t-1, i) be the set of vortex sequences in the atmospheric wind field at time t-1, with a total of n. t-1 A sequence of vortices, i=1, 2, …, n t-1 The starting times of the i-th vortex sequence are es_start(t-1,i) and es_end(t-1,i), respectively. The middle time of the starting time is a continuous Martian day. The i-th vortex sequence consists of a total of es_end(t-1,i) - es_start(t-1,i) + 1 vortices.

[0030] Taking the vortex e(t, i) of Martian day t as an example. According to previous research, the farthest distance a Martian atmospheric vortex can travel in one Martian day is D' kilometers. Therefore, for vortex e(t, i), the search area is a circle with a radius of 2D' centered at its center P(t, i). The center P(t-1, j) of vortex e(t-1, j) in the previous step (Martian day) is only checked for correlation if it falls within this search area. To explain the physical similarity between vortex sequences, in addition to the distance between vortex centers in adjacent Martian days, the size and vorticity of the vortices should also be compared. Therefore, this invention uses the cost function (C) of adjacent Martian day vortices e(t, i) and e(t-1, j). ij To evaluate the correlation between vortices:

[0031] (5),

[0032] Among them, X d X R and X Ro It is a dimensionless value that quantifies the relative differences in the distance, radius, and vortex intensity between adjacent Martian solar vortex centers.

[0033] (6),

[0034] (7),

[0035] (8),

[0036] Where, d ij ΔR is the spatial distance between the centers P(t, i) and P(t-1, j) corresponding to vortices e(t, i) and e(t-1, j); ΔR is the difference between the equivalent radii R(t, i) and R(t-1, j) corresponding to vortices e(t, i) and e(t-1, j); ΔRo is the difference between the vorticities Ro(t, i) and Ro(t-1, j) corresponding to vortices e(t, i) and e(t-1, j). It is the average of the average equivalent radii of the same vortex at times t-1, t-2, and t-3. This is the average vorticity of the same vortex at times t-1, t-2, and t-3. If the same vortex is less than two times before time t-1, then the average value of all times is used. If multiple vortices are found in the search area at time t, the vortex with the lowest cost function calculated from the vortex e(t-1, j) at the previous time is most correlated, and these two can be considered as a vortex sequence.

[0037] After finding the vortex e(t, i) at time t that has the minimum cost function (correlation) with the vortex e(t-1, j) at the previous time t-1, if the vortex e(t-1, j) already exists in the vortex sequence es(t-1, k) at time t-1, then increment the number of vortices in that sequence by 1, making the last vortex e(t, i), and set the end time es_start(t-1, k) of the vortex sequence es(t-1, k) to t. If the vortex e(t-1, j) does not exist in the vortex sequence es(t-1, k) at time t-1, then generate a new vortex sequence es(t-1, n) within the vortex sequence es(t-1, k). t-1 +1), which contains only two vortices e(t-1, j) and e(t, i), and the starting times of this vortex sequence are es_start(t-1, n) and e(t, i), respectively. t-1 +1)=t-1 and es_ end(t-1, n t-1+1)=t.

[0038] Then, iterate through all vortices at time t, repeating the above steps. Use the cost function to find the vortices associated with time t-1 and update the vortex sequence at time t-1. Merge the newly generated, updated vortex sequence at time t-1 with the vortex sequences that were not updated at time t-1 to form the vortex sequence set at time t. Starting from the vortex set of the second Martian day, sequentially traverse all vortex sets for n Martian days and generate the vortex sequence for each Martian day, finally obtaining the vortex sequence set es(n, n) for the nth Martian day. n ), which contains a total of n n A vortex sequence.

[0039] Step 5: Analysis of vortex sequence life cycle parameters.

[0040] After obtaining the set of vortex sequences over n Martian days, we conduct an analysis of the changes in shape parameters (equivalent radius, average velocity) and physical parameters (water ice content, etc.) during the lifecycle of the vortex sequences. From the final vortex sequence set es(n, n n Choose a vortex sequence es(n, Index) from the given list (e.g., the longest duration, or located in the Northern Hemisphere summer, Southern Hemisphere winter, etc.). Its duration in Martian days or its lifespan is LC = es_end(n, Index) - es_start(n, Index) + 1. Let the equivalent radii during the lifespan of this vortex sequence be R1, R2, ..., R. LC Average speeds V1, V2, ..., V LC The water ice content is I1, I2,…, I LC Then, using the Martian day sequence sol1, sol2, ..., sol in the vortex sequence... LC Plot the equivalent radius, average velocity, and water ice content on the x-axis, and draw a two-dimensional polygonal plot of the equivalent radius, average velocity, and water ice content within the life cycle of the vortex sequence. Analyze the variation patterns of the relevant parameters within the vortex life cycle. Attached Figure Description

[0041] Figure 1 This is a vector arrow diagram of the near-surface wind field on the 296th Martian day of MY28, used as an example. The base image is a mosaic of full-Mars color images obtained by the US Viking spacecraft. The polygons in the diagram represent the vortex boundaries identified during this Martian day.

[0042] Figure 2 This is a schematic diagram of the vortex center for a Martian year (MY28-29) as shown in the example. The black and white dots represent the extracted vortex centers, the polygon pairs surrounding the dots represent the corresponding vortex boundaries, and the color of the polygons represents the average velocity magnitude of the vortex boundaries.

[0043] Figure 3 This is a map showing the movement of the center of a Martian vortex sequence during a Martian year (MY28-29). The colors represent the distance the Martian vortex sequence has moved. AD marks the positions and paths of the four vortex sequences.

[0044] Figure 4 The diagram shows the change in the equivalent radius of the vortex sequence AD ​​in the example.

[0045] Figure 5 The figure shows the average velocity variation of the vortex sequence AD ​​as an example.

[0046] Figure 6 The figure shown is a diagram of the water ice content variation in the vortex sequence AD ​​as an example, where figure b is a magnified diagram of the water vapor content in the vortex sequence D. Detailed Implementation

[0047] A method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data, including the following processes:

[0048] (1) Acquiring and processing Martian atmospheric wind field data. The Open Access Mars Assimilation Remote Sensing Dataset (OpenMARS) is a reanalysis product, a global surface / atmosphere reference database covering nearly nine Martian years, used to acquire atmospheric conditions (such as wind speed, temperature, dust, and carbon dioxide) and study the physical, dynamic, and chemical behavior of the Martian atmosphere. The OpenMARS product has a horizontal grid resolution of 5° in both longitude and latitude, with 35 vertical layers from the Martian surface to an altitude of 105 km, and a temporal resolution of 2 Martian hours. The OpenMARS MY24-34 product can be downloaded from the website https: / / ordo.open.ac.uk / articles / dataset / OpenMARS_MY28-32_standard_database / 7352579.

[0049] Near-surface atmospheric wind field data for 668 Martian days (MY28 Ls=109° to MY29 Ls=120°) were extracted from the OpenMARS dataset. Figure 1 As shown. Then, linear interpolation was used to interpolate the original 5° resolution wind field data by a factor of 15 to ~(1 / 3)°, and atmospheric wind field data at dawn of each Martian day were extracted to obtain a Martian atmospheric wind field dataset with a Martian year, near the Martian surface, a temporal resolution of one Martian day, and a spatial resolution of (1 / 3)°. The near-surface atmospheric wind field map of the 296th Martian day in MY28 was obtained using a vector arrow visualization method.

[0050] (2) Based on steps two and three mentioned above, vortex centers and boundaries were extracted from the atmospheric wind field data of 668 Martian days in one Martian year after processing. The results are as follows: Figure 2 As shown. A total of 4089 Martian atmospheric vortices were extracted, with the most (12) found in the 296th Martian day of MY28. The minimum and maximum areas of the extracted Martian vortices were 2.67 × 10⁻⁶. 4 km 2 and 2.23×10 6 km 2 The average area is 3.24 × 10. 5 km 2 The average equivalent radius R is 298.46 km. The northern hemisphere of Mars has significantly more vortices than the southern hemisphere, with 2408 and 1681 respectively. Atmospheric vortices are more frequent in spring and summer than in autumn and winter, mainly concentrated near Martian topography (such as basins, mountains, and plateaus).

[0051] (3) Based on step four mentioned above, correlation detection was performed on the generated vortex objects for adjacent Martian days to generate a Martian atmospheric vortex sequence for a Martian year (MY28-29). The results are as follows: Figure 3 As shown, a total of 83 Martian atmospheric vortex sequences were tracked and extracted. The longest-lasting vortex sequence appeared in Isidis Planitia, lasting approximately 124 Martian days. 90% of the vortex sequences had short lifespans, less than 40 or less. 32 and 28 vortex sequences appeared in the Martian northern hemisphere during spring and summer, respectively, while 13 and 10 sequences appeared in autumn and winter, respectively. Therefore, the vortex sequence MY28-29 mainly occurred in the northern hemisphere during spring and summer.

[0052] (4) Four vortex sequences were selected from the set of vortex sequences in the Martian year MY28-29 and named AD. Vortex sequence A is located in Isidis Planitia, with a regional average elevation of -3161.7 m and an average wind speed of 3.06 m / s; sequence B is located in Solis Planum, with a regional average elevation of 5113.0 m and an average wind speed of 10.96 m / s. The lifespans of A and B are 75 and 48 Martian days, respectively. Vortex sequence C is located near the Arctic ice cap and appears in summer (MY28 Ls=127.36-135.68°); while vortex sequence D is located in the mid-to-high latitudes of the Northern Hemisphere and appears in winter (MY28 Ls=351.6-335.4°). Sequences C and D last for 18 and 8 Martian days, respectively. Sequence D moves from south to north, while sequence C, after being generated near the Arctic ice cap, gradually moves south. For specific locations and movement paths, see [link to relevant documentation]. Figure 3 The AD serial number is marked in the middle.

[0053] Plot the changes in equivalent radius, average velocity, and water vapor content during the AD lifecycle of the vortex sequence, as follows: Figures 4-6 As shown, the average wind speeds of vortex sequences AD are 4.23, 9.95, 5.78, and 8.8 m / s, respectively, with sequences B and D having higher speeds than A and C. On terrestrial planets, wind speed increases with regional elevation. The average equivalent radii of vortex sequences AD are 335.7, 434.3, 808.5, and 705.1 km, respectively. The average radii of vortex sequences A and B are small and stable, indicating their formation is limited by Martian topography. The average radii of vortex sequences C and D are unstable, generally increasing first and then decreasing. The water vapor columns of vortex sequences A and B show little variation, with an average value of 8.63 × 10⁻⁶. -3 and 5.37×10 -3 kg / m 2 Low-altitude vortex sequence A (Isidis Planitia) has relatively high water vapor abundance, while high-altitude vortex sequence B (Solis Planum) has relatively low water vapor content. Vortex sequence C occurs near the Arctic ice cap in summer and has the highest water vapor content (average ~4.6 × 10⁻⁶). -2 kg / m 2 During spring and summer, large amounts of water from the Arctic are injected into the atmosphere from the polar caps, carried by winds across the equator, and then incorporated into the Hadley Circulation. Over 18 Martian days, the water vapor column of Vortex Sequence D decreased from ~5.33 to 3.9 × 10⁻⁶. -2 kg / m 2 It carries a total of ~6.64×10 9 kg of water vapor enters the Martian atmospheric water cycle, with a daily loss of ~3.69 × 10⁻⁶ kg. 8 kg / sol.

Claims

1. A method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data, characterized in that, Includes the following steps: Using long-term series wind field data from Mars atmospheric model products, Martian vortex centers are identified based on locally normalized angular momentum within a grid. The Martian vortex boundaries are extracted based on the outer streamline velocity of the vortex centers, and the Martian vortex sequence is tracked by minimizing a loss function. Finally, the changes in morphological and property parameters within the vortex sequence are analyzed in conjunction with the vortex sequence lifetime. Specifically: Step 1: Data processing of Mars atmospheric model products; The Mars atmospheric model product dataset is D={Type| grid, time, height}, with a spatial resolution of r_s (in degrees), which means dividing the Martian surface into a square grid set grid with a side length of r; its temporal resolution is r_t (in Martian hours), where time is a value of one time profile; its height resolution is r_h, where height is a value of one layer of height profile; the data type of the Martian atmosphere stored in dataset D is Type, with values ​​including u, v, and I, which represent the wind speed and water ice content in the longitude and latitude directions on the time profile time and the height profile height in the grid division of the Martian surface, respectively; Temporal extraction of the raw data is performed; the spatial resolution of the original dataset D is increased by a factor of N through linear interpolation, becoming r_s' = r_s / N; data from each Martian day in the original dataset are extracted, retaining only one value at the same moment for each Martian day, and the temporal resolution becomes r_t', in Martian days; the Martian atmospheric model product dataset after spatial resolution interpolation and temporal resolution extraction is D1 = {Type| grid} i time j height k }; i is the grid index, i=1, 2, …, n1, n1 is the total number of grids in Mars; j is the Martian day index in the Mars atmospheric model product data, j=1, 2, …, n2, n2 is the total number of Martian days in the Mars atmospheric model product data; k is the Martian atmospheric altitude index in the Mars atmospheric model product data, k=1, 2, …, n3, n3 is the number of atmospheric altitude layers in the Mars atmospheric model product data; Step 2: Extraction of candidate centers for Martian atmospheric vortices; Calculate the angular momentum in the Martian atmospheric model product mesh and extract the location of the maximum angular momentum value in a local region as the center of the Martian atmospheric vortex; one Martian day is time. j The atmospheric altitude of Mars is height. k The grid points are grid i The wind field data in the corresponding Mars atmospheric model product is u1={u| grid i time j height k }, v1={v| grid i ,time j height k }, written in vector form as Establish a neighborhood centered on a grid point; let the side length L of the neighborhood be 2b times the side length of the grid, i.e., L = r_s' × 2b, where b is a positive integer greater than 0; the total number of grids in this neighborhood is num = (2b + 1) × (2b + 1), and the central grid is gridd. i The formula for the locally normalized angular momentum LNAM is as follows: (1)， in, This represents the relative vector between the m-th grid in the neighborhood and the center grid, gridi. Let be the wind field vector of the m-th grid in the neighborhood; Then, a threshold for LNAM in the Martian atmospheric grid is set to T1. All grids with LNAM greater than this threshold are searched, and contour lines are generated. The regions enclosed by these contour lines constitute the candidate region set where LNAM is greater than the threshold T1. Within this candidate region set, the grid with the maximum LNAM value is identified; this grid represents the center of the Martian atmospheric wind vortex. At time... j and height k Using this method on the cross-section, num1 vortex centers were calculated and searched, i.e., the set of candidate vortex centers P = {P1, P2, …, P}. num1 }; Step 3: Generation of Martian atmospheric vortex boundaries; First, extract the streamlines. A streamline is a line in which the tangent at any point is in the direction of the instantaneous velocity at that point. The area enclosed by a closed streamline is A, and its perimeter is L. p The equivalent radius R and the average velocity V_SL of the closed streamline are respectively: (2)， (3)， The mean velocity of a closed streamline centered on an atmospheric or oceanic vortex gradually increases with the increase of its equivalent radius R. Once it reaches its maximum value V_SL_max, the mean velocity of the closed streamline will decrease with the increase of the equivalent radius. The closed streamline around the center of the vortex that reaches the maximum mean velocity is used to define the characteristic profile of the vortex, i.e., the vortex boundary. The vortex center is P i Extract the center P of the vortex from the inside out at certain intervals s. i Find the set of all closed streamlines SL, and calculate the equivalent radius and mean velocity of each closed streamline; find the closed streamline SL' in the set of closed streamlines SL with the maximum mean velocity V_SL_max, and this closed streamline SL' is considered to be the boundary B of the vortex center. i Furthermore, for a well-defined, coherent vortex, the closed streamlines around its center should remain coherent in both space and time; therefore, when the streamlines around the vortex center P... i The vortex criterion is met when the average velocity V_SL_End of the last closed streamline is at least one threshold T2 lower than the maximum average velocity V_SL_max: V_SL_End <= (1-T2) × V_SL_max. If this criterion is not met, the structure around which the closed curve encircles reaches the maximum velocity at the edge; this structure is not a vortex and P needs to be adjusted. i Exclude vortex candidates from the vortex center set P; iterate through each vortex candidate center in the vortex center set P, and generate the vortex center set P'={P1, P2, …, P} using the method described above. num2 } and the corresponding vortex boundary set B={B1, B2, …, B num2 The indices of the vortex centers and vortex boundaries in the two sets are in a one-to-one correspondence, and there are a total of num2 vortex centers and vortex boundaries. Finally, the vorticity or Rossby number Ro of the identified vortex is calculated; vorticity is a measure used in meteorology to study vortices in ocean and atmospheric systems, and its calculation formula is: Ro= V_SL_max / (f×R1) (4), Where V_SL_max and R1 are the average velocity and equivalent radius of the vortex boundary, i.e., the closed streamline; f is the magnitude of the Coriolis force at the vortex center; the Coriolis force describes the offset of a particle moving in a straight line relative to the rotating system due to inertia; finally, the vortex vorticity set Ro={Ro1, Ro2, …, Ro} is generated. num2 } 2. The method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data according to claim 1, characterized in that, Also includes: Step 4: Tracking the Martian atmospheric vortex sequence; Vortex sequence tracing is used to track vortex sequences that evolve over time on a Martian day basis, extracting their trajectories and evolutionary features. The methods in steps two and three are applied to Martian wind field data at the same altitude over n consecutive Martian days, resulting in a set of vortices e(t, i) identified over n Martian days, with corresponding vortex centers and boundaries P(t, i) and B(t, i) respectively, and corresponding equivalent radii R(t, i) and vorticity Ro(t, i); t is a Martian day, t=1, 2,…, n, and i is the i-th vortex identified in the t-th Martian day. Step 5: Analysis of vortex sequence lifecycle parameters; After obtaining the set of vortex sequences over n Martian days, we conducted an analysis of the changes in shape and physical parameters during the lifecycle of the vortex sequences.

3. The method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data according to claim 2, characterized in that, In step four, the specific method of the vortex sequence tracing algorithm is as follows: Within a given search area, the vortex e(t, i) in the current time step (i.e., the current Martian day t), t>1, is correlated with each vortex e(t-1, j) extracted in the previous time step t-1 (i.e., the previous Martian day). The vortex with the strongest correlation is selected as the successor of vortex e(t-1, j), and the vortex sequence is established or updated. Let es(t-1, i) be the set of vortex sequences in the atmospheric wind field at time t-1, with a total of n. t-1 A sequence of vortices, i=1, 2, …, n t-1 The starting times of the i-th vortex sequence are es_start(t-1, i) and es_end(t-1, i), respectively. The middle time of the starting time is a continuous Martian day. The i-th vortex sequence consists of a total of es_end(t-1, i) - es_start(t-1, i) + 1 vortices. Let the vortex of Martian day t be e(t, i); let the farthest distance a Martian atmospheric vortex can travel in one Martian day be D' kilometers. Then, for vortex e(t, i), the search area is a circle with a radius of 2D' centered at its center P(t, i). Correlation detection is only performed if the center P(t-1, j) of the previous Martian day vortex e(t-1, j) falls into this search area. The size and vorticity of the vortices are also compared. The cost function (C) of adjacent Martian day vortices e(t, i) and e(t-1, j) is used. ij To evaluate the correlation between vortices: (5)， Among them, X d X R and X Ro It is a dimensionless value that quantifies the relative differences in the distance, radius, and vorticity between the centers of adjacent Martian solar vortices; （6）， （7）， （8）， Where, d ij ΔR is the spatial distance between the centers P(t, i) and P(t-1, j) corresponding to vortices e(t, i) and e(t-1, j); ΔR is the difference between the equivalent radii R(t, i) and R(t-1, j) corresponding to vortices e(t, i) and e(t-1, j); ΔRo is the difference between the vorticities Ro(t, i) and Ro(t-1, j) corresponding to vortices e(t, i) and e(t-1, j). It is the average of the average equivalent radii of the same vortex at times t-1, t-2, and t-3. It is the average value of the vorticity of the same vortex at times t-1, t-2 and t-3. If the same vortex is less than two time points before time t-1, then the average value of all time points is used. If multiple vortices are found in the search area at time t, the vortex with the minimum value of the cost function of the vortex e(t-1, j) at the previous time point is most correlated, and the two are taken as the vortex sequence. After finding the vortex e(t, i) at time t that has the minimum cost function with the vortex e(t-1, j) at the previous time t-1, if the vortex e(t-1, j) already exists in the vortex sequence es(t-1, k) at time t-1, then increment the count of the vortices in that sequence by 1, making the last vortex e(t, i), and set the end time es_start(t-1, k) of the vortex sequence es(t-1, k) to t; if the vortex e(t-1, j) does not exist in the vortex sequence es(t-1, k) at time t-1, then generate a new vortex sequence es(t-1, n) in the vortex sequence es(t-1, k). t-1 +1), which contains only two vortices e(t-1, j) and e(t, i), and the starting times of this vortex sequence are es_start(t-1, n) and e(t, i), respectively. t-1 +1)=t-1 and es_ end(t-1, n t-1 +1)=t; Then, iterate through all vortices at time t, repeating the above steps. Use the cost function to find the vortices associated with time t-1 and update the vortex sequence at time t-1. Merge the newly generated and updated vortex sequence at time t-1 with the vortex sequences that were not updated at time t-1 to form the vortex sequence set at time t. Starting from the vortex set of the second Martian day, sequentially traverse all vortex sets for n Martian days and generate the vortex sequence for each Martian day, finally obtaining the vortex sequence set es(n, n) for the nth Martian day. n ), which contains a total of n n A vortex sequence.

4. The method for extracting, tracking, and analyzing Martian atmospheric vortices based on wind field data according to claim 3, characterized in that, Step five specifically involves: extracting the final vortex sequence set es(n, n) n Choose a vortex sequence es(n, Index) from the given list. Its duration in Martian days or its lifespan is LC = es_end(n, Index) - es_start(n, Index) + 1. The equivalent radius during the lifespan of this vortex sequence is R1, R2, ..., R. LC Average speeds V1, V2, ..., V LC The water ice content is I1, I2,…, I LC Then, using the Martian day sequence sol1, sol2, ..., sol in the vortex sequence... LC Plot the equivalent radius, average velocity, and water ice content on the x-axis, and draw a two-dimensional polygonal plot of the equivalent radius, average velocity, and water ice content within the life cycle of the vortex sequence. Analyze the variation patterns of the relevant parameters within the vortex life cycle.

Images