S

O

pen Access

J Geol Geosci Volume 3(2): 20191

ReseaRch aRticle

Journal of Geology and Geoscience

Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare CratersWU Yuan ling1,2*, CHEN Jian-ping1,2

1School of Earth Sciences and Resources, China University of Geosciences (Beijing), Beijing 100083, China2Key Laboratory of Land and Resources Information Research & Development in Beijing, Beijing 100083, China

AbstractThe Mare is an important topographical unit while the impact crater is an important annular structure on the surface of the moon. The purpose of this paper is to study the temporal and spatial distribution characteristics of impact craters in the mare area and lay a good foundation for further determination of impact events in the lunar mare area. The LU106016 database is selected to classify the mare chronologically, which is divided into five periods, namely Aitkenian, Nectarian, Late Imbrian, Early Eratothenian and Late Eratothenian. The relationship between the periods of the impact craters and the number of geological units, the number of impact craters and the impact areas are analyzed statistically. According to the results of statistical analysis, Late Imbrian is an era in which impact events occured frequently in the lunar mare area, and the number of geological units, the number of impact craters and the impact areas are much larger than those in the other four periods, from Aitkenian to Late Eratothenian, the impact strength shows a trend of first increasing and then decreasing. At last, combining the Kernel Density Estimation and latitude-longitude distribution, the spatial distribution characteristics of impact craters in each period are analyzed in details. The Aitkenian impact craters are mainly distributed in the Mare Orientale, Mare Humorum, Mare Serenitatis and Mare Marginis areas, with a relatively balanced distribution in latitude and longitude. The Nectarian craters are mainly distributed in the Mare Spumans and Mare Australe area, and is extremely uneven in latitude and longitude distribution, the number of craters in the western hemisphere is significantly larger than that in the eastern hemisphere, and in latitude, the number of craters in the southern hemisphere is significantly larger than that in the northern hemisphere. The number and concentration of late Imbrian craters in the northern hemisphere are higher than those in the southern hemisphere, while the number and concentration in the eastern hemisphere are higher than those in the western hemisphere. The impact craters of the early Eratothenian are evenly distributed in the Mare Imbrium area, with relatively scattered distribution and uneven latitude and longitude distribution, mainly in the western and northern hemispheres. The number of impact craters in Late Eratothenian is very small, but the distribution is very regular, they are mainly distributed in the Mare Imbrium and Mare Anguis area, with 20o longitude and 35o latitude as axisymmetric distribution respectively.

Key Words: Mare, Craters, Chronological Division, Space Distribution

IntroductionThe successful launch of Chang ‘e - 4 set up another milestone in China’s lunar exploration project, and the mysterious veil of the moon will be gradually revealed. The internal energy of the moon basically consumed 3 billion years ago, and the surface of the moon has not been affected by life activities, water bodies, plate movement, and atmosphere and so on. Therefore, the moon completely retains its history of being impacted by small celestial bodies, providing an example for us to explore the evolution of the moon, other planets and even solar system [1]. Based on the different driving forces, the evolution of the moon can be divided into three periods: the first period is dominated by internal dynamic geological processes, beginning with the formation of the moon and ending with the evolution of magma ocean [2- 4]; The second period is dominated by both internal and external dynamic geological processes, when two major events took place, namely the impact event and the flood of mare basalt [5- 8]; The third period is dominated by external dynamic geological processes. During this period,

Correspondence to: WU Yuan ling, School of Earth Sciences and Resources, China University of Geosciences (Beijing), Beijing 100083, China, Email: wylcugb[AT]163[DOT]com

Received: Jan 23, 2019; Accepted: Jan 25, 2019; Published: Jan 28, 2019

internal dynamic geological processes basically stopped, there were mainly small-scale impact events, and the moon entered a relatively calm period [9-10]. Impact is considered to be the most important surface effect on the moon. The study on the classification and tectonic evolution of the lunar impact crater has become the key to understanding the lunar impact events.

The mare is an important topographic unit on the surface of the moon. There are 22 mares on the moon, of which 19 are located on lunar nearside and 3 are located on the farside [11]. The mare area has not only experienced large-scale basalt eruptions, but also contains many traces of impact events. The study of mare impact events will be helpful for further study of full-month impact events.

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20192

Method and DataMethod

Crater Size-Frequency Distribution

The method to determine the absolute age used in this paper is the Crater Size-Frequency Distribution (CSFD) method, which is the most commonly, used method to obtain the unit age of the surface of planets hit by impact. The CSFD method can establish a functional relationship between the number of impact craters and the age by combining SFD data with isotope dating results. The method is based on two basic assumptions: impact craters are randomly distributed; and the rate of disappearance of impact craters is much slower than the rate of generation of impact craters [12- 13].

1. Production Function (PF). PF refers to the functional relationship between the diameter of the impact craters and the number of impact craters in per unit area (i.e. the frequency of impact craters). Hartmann initially established the impact crater production function [14], which was later applied and improved by Hartmann and other scholars.

Hartmann Production Function (HPF)

Hartmann (2005) used a three-part piecewise function to represent the impact crater size-frequency distribution [15] according to the different diameter ranges of the impact craters.

log 2.616 3.82logH LN D= − − 0.3Km<DL<1.41Km

log 2.920 1.80logH LN D= − − 1.41Km<DL<64Km

log 2.198 2.20logH LN D= − − DL>64Km

In the formula, DL and DR respectively represent the left and right boundaries of the unit (the standard unit width DR/ DL=

2 ). HPF can only be used in areas where the lunar surface has not yet reached a saturated equilibrium state [16].

Neukum Production Function (NPF)

Neukum et al. (1994) believed that the shape of the crater production function curves of different geological units on the moon in different lunar historical periods is approximately the same in all the diameter ranges available for measuring the crater size-frequency distribution, which means the crater production function is independent in time. Neukum then used a curve to fit the lunar crater size-frequency distribution statistics [17] obtained from the lunar mare area with an age of 1Ga.

11

01

log (log )kcum k

kN a a D

=

= +∑

2. Age Function. After obtaining the relationship between the diameter and frequency of the impact craters in the geological units, it is also necessary to establish a functional relationship between the obtained SFD data and the isotope dating results of Apollo samples. According to the difference in the diameter of the impact craters, there are usually three methods.

Melosh and Vickery method (4Km <D<100Km)

Melosh and Vickery established a functional relationship between the cumulative density of impact craters and the

isotopic ages of Apollo samples in 1989.5

7

(4) 2.68 10[ 4.57 10 exp(4.53 ) 1]NT T

−

−

= ×

+ × −

This method is suitable for impact craters with diameters between 4Km and 100Km, where N(4) represents the cumulative density of impact craters with diameters greater than 4Km, and T represents the geological age [18].

Neukum method (D>1Km)

At present, the most commonly used method is the algorithm proposed by Nuekum in 1983, and then Neukum improved the production function in 2001 [19] and calculated the age of each unit [20] with the improved algorithm. The improved expression is as follows:

14

4

(1) 5.44 10[exp(6.93 ) 1] 8.38 10n

T T

−

−

= ×

− + ×

This method is suitable for impact craters with a diameter greater than 1Km, where N (1) represents the cumulative density of impact craters with a diameter greater than 1Km, and T represents the geological age.

Li Kun et al method (D<1Km)

Kun Li et al. (2012) obtained the relationship between the cumulative density of the impact craters and the isotopic age of Apollo samples when the diameter of the impact craters is less than 1Km based on the production function proposed by Neukum.

12

1

[ ( ) ( )][ ( ) ( )](1) (1)

[ ( ) ( )]

ni ii

nii

PF D PF D N D N DN PF

PF D PF D=

=

− −= •

−∑

∑

Among them,11

01

lg ( ) (lg )nn

nPF D a a D

=

= + ×∑ , ( )N D and ( )PF D

Represents the average of the cumulative density of impact craters and the sampling values of production functions, respectively [21]. The diameter of the impact craters selected in this study is more than 500m, and the production function and age function of the impact craters are proposed and modified by Neukum in 1994 and 2001 respectively.

Kernel Density Estimation

Kernel Density Estimation (KDE) is a nonparametric estimation method used to estimate the density function of an unknown distribution. It can effectively obtain the spatial distribution of a data set without training samples [22]. It is based on the idea that the occurrence of events in space has some randomness, but this randomness will be affected by certain spatial processes. Therefore, the probability of occurrence at different locations is different [23-26]. It considers that any location in the region has a measurable event density (also called intensity), and the event density at that location can be estimated by the number of event points per unit area around it [27- 28].

Let X1,…XN be an independent identically distributed sample extracted from a population with a distribution density function of f, and estimate the value f(x) of f at a certain point x, usually there will be Rosenblatt-Parzen [29-31] kernel estimate:

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20193

1

1( ) ( )n

in

i

x xf x knh h=

−= ∑ , where k (·) is called the kernel function.

2 2/22

3( )2

d ij hk d ehπ

−= , h> 0 is band width; dij is the distance

from point i to point j; (x -Xi) indicates the distance from the valuation point to event Xi. The method centers on the position of a specific feature point, distributes the attributes of the point within a specified threshold range (circle with radius h), density at the center is highest, and decays with distance until 0 [32-33] at the limit distance.

Data

In this study, Clementine multispectral images of 415nm, 750nm and 950nm were used for ratio calculation, in which 415 / 750 jointly reflects the changes of lunar surface maturity and Ti content, while 750 / 950 ratio reflects Fe content, and then false color images were synthesized according to the above ratios [34-35]. Different colors in the false color image reflect different compositions and albedo, on the premise that the same geological body has the same material composition [16], different geological units can be relatively accurately divided according to the color change of the image, and then the crater size - frequency distribution can be fixed by combining Chang ‘e- 1 DEM data (with resolution of 500m) [36] and LROC data

(with resolution of 100m) [37].

This study selected LU106016 crater database, which has a crater diameter greater than 500m [38] and it, contains the largest number, most complete type and most systematic shape index [39] so far.

Temporal Distribution of Mare Impact Craters

There are 5175 craters that we can use in the mare area from LU106016 crater database (figure 2.1). Based on Clementine UV/VIS image data, the mare area is divided into 78 geological units (figure 2.2), and each geological unit is dated by combining Chang ‘e 1 DEM and LROC data. The area of each geological unit, the number of craters and the dating results are shown in (table 2.1).

Referring to the geochronology scheme proposed by Guo Dijun (2014) and Ouyang Ziyuan (2014) [40-41], the 78 geological units are chronologically classified into five categories (Table 2.2) which are Aitkenian, Nectarian, Late Imbrian, Early Eratothenian and Late Eratothenian, and the Period of impact craters and number of geological units, number of impact craters and impact area are statistically analyzed respectively (Figure 2.3). According to the results of statistical analysis, the Late Imbrian is a time full of frequent impact events, and

Figure 2.1: Data distribution of lunar mare impact craters in database LU106016 (base image is LROC data)

Figure 2.2: Division of mare units (base image is LROC data)

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20194

Mare Name Geological Unit Name Age (Ga) Area (103km2) Number of craters

Anguis

A1 2 1.89 3A2 3.8 1.79 9A3 3.9 0.193 6A4 3.9 0.185 3

Australe

A1 3.7 115 753A2 3.9 0.547 2A3 3.9 6.71 42A4 3.8 1.98 18A6 3.9 1.11 4

CognitumCo1 3.7 15.8 37Co2 3.8 1.7 8Co3 3.9 1.16 7

Crisium

Cr1 3.5 32 131Cr2 3.7 7.52 48Cr3 3.8 2.35 10Cr4 3.7 2.16 14Cr5 3.8 6.28 39

FecunditatisFe1 3.7 39.8 100Fe2 3.8 28.9 27Fe3 3.8 9.64 33

Frigoris Fr1 3.3 155 185Fr2 3.5 61.3 177

Humboldtianum Hu1 3.5 1.71 6Hu2 3.2 27.8 183

Humorum

Hu1 3.8 8.77 90Hu2 3.8 4.71 11Hu3 3.7 14.7 71Hu4 3.9 1.66 9Hu5 3.8 0.955 12Hu6 4.1 2.22 24

Imbrium

I1 3.5 35.4 130I2 2.8 1.92 5I3 3.4 79.9 267I4 3.8 1.46 4I6 3.7 9.32 23I7 3.6 20.7 69I8 3.6 36.4 81I9 3 73.2 64

I10 3.6 48.7 141I11 3.3 197 232I12 3.4 38.8 20I13 3.3 86.3 197I14 2.9 217 441I15 3.2 2.08 4I16 3.7 0.807 4I17 3.6 3.48 16I18 3.2 8 8I19 3.6 3.35 2I20 3.8 1.82 3I21 3.7 0.896 2I24 3.7 0.530 2I25 3.9 0.999 1

Ingenii In1 3.7 11.8 43

Marginis M1 3.8 10.5 12M2 4 6.57 28

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20195

Moscoviense Mo1 3.7 3.32 23Mo2 3.8 4.1 52

NectarisNe1 3.9 6.71 13Ne2 3.9 4.98 5Ne3 3.8 6.23 22

Nubium Nu1 3.8 49.9 97

Orientale

O1 3.8 6.99 10O2 3.8 2.4 9O3 3.9 2.51 2O4 4 0.426 2

Orientale- south Os1 3.5 1.38 3

SerenitatisS1 4 1.21 3S2 3.6 40.9 202S3 3.5 56.6 343

Smythii S 3.7 10.4 45Spumans Sp 3.9 4.08 62Temporis Te 3.8 0.0888 2

TranquillitatisTr1 3.8 11.6 58Tr2 3.8 16.4 38Tr3 3.7 74.1 220

Undarum Un 3.8 6.07 73

Vaporum Va1 3.6 10.5 10Va2 3.8 6.68 20Table 2.1: Dating results of lunar mare geological unit

Period Geological unit name Age(Ga) Area(103km2) Number of craters Amount of craters

Aitkenian

Hu6 4.1 2.22 24M2 4 6.57 28 57O4 4 0.426 2S1 4 1.21 3

Nectarian

A3 3.9 0.193 6A4 3.9 0.185 3A2 3.9 0.547 2A3 3.9 6.71 42A6 3.9 1.11 4

Co3 3.9 1.16 7 156Hu4 3.9 1.66 9I25 3.9 0.0999 1Ne1 3.9 6.71 13Ne2 3.9 4.98 5O3 3.9 2.51 2Sp 3.9 4.08 62

Late Imbrian

A2 3.8 1.79 9A1 3.7 115 753A4 3.8 1.98 18

Co1 3.7 15.8 37Co2 3.8 1.7 8Cr1 3.5 32 131Cr2 3.7 7.52 48Cr3 3.8 2.35 10Cr4 3.7 2.16 14Cr5 3.8 6.28 39Fe1 3.7 39.8 100Fe2 3.8 28.9 27

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20196

Fe3 3.8 9.64 33Fr1 3.3 155 185Fr2 3.5 61.3 177Hu1 3.5 1.71 6Hu2 3.2 27.8 183Hu1 3.8 8.77 90Hu2 3.8 4.71 11Hu3 3.7 14.7 71Hu5 3.8 0.955 12I1 3.5 35.4 130I3 3.4 79.9 267I4 3.8 1.46 4I6 3.7 9.32 23I7 3.6 20.7 69I8 3.6 36.4 81

I10 3.6 48.7 141I11 3.3 197 232I12 3.4 38.8 20I13 3.3 86.3 197 4449I15 3.2 2.08 4I16 3.7 0.807 4I17 3.6 3.48 16I18 3.2 8 8I19 3.6 3.35 2I20 3.8 1.82 3I21 3.7 0.896 2I24 3.7 0.530 2In1 3.7 11.8 43M1 3.8 10.5 12

Mo1 3.7 3.32 23Mo2 3.8 4.1 52Ne3 3.8 6.23 22Nu1 3.8 49.9 97O1 3.8 6.99 10O2 3.8 2.4 9Os1 3.5 1.38 3S2 3.6 40.9 202S3 3.5 56.6 343S 3.7 10.4 45Te 3.8 0.0888 2Tr1 3.8 11.6 58Tr2 3.8 16.4 38Tr3 3.7 74.1 220Un 3.8 6.07 73Va1 3.6 10.5 10Va2 3.8 6.68 20

Early EratothenianI9 3 73.2 64 505

I14 2.9 217 441

Late EratothenianA1 2 1.89 3 8I2 2.8 1.92 5

Table 2.3: Time division results of geological units in the lunar mare

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20197

its number of geological units, number of craters and impact area are much larger than those in the other four periods. From Aitkenian to Late Eratothenian, the strength of impact shows a trend of increasing first and then decreasing.

Spatial Distribution of Mare Impact Craters

Kernel Density Estimation

In the space-time domain of geography, an impact crater exists only at one spatial point at one time point, and the formation of the impact craters on the lunar surface satisfies the basic assumption of isotropy, but the distribution characteristics of the impact craters are different in different regions and different spatial scales. Based on the above principles, the space analysis of the point pattern is carried out by regarding the lunar impact craters as the points [8].The method of kernel density estimation is used to analyze the spatial distribution characteristics of lunar mare impact craters in different periods. In the kernel density estimation, the bandwidths of 100Km, 200Km, 300Km, 400Km and 500Km were selected for comparison. The smaller the bandwidth is, the larger the density value within the bandwidth and the more abrupt the density curve is. The larger the bandwidth selection is, the smaller the density value within the bandwidth and the smoother the density value curve are [42]. After comparison and analysis, the bandwidth of 300Km was finally selected, which can clearly show the density center of the impact craters and density difference?

The total number of Aitkenian impact craters is 60, and the total number used to determine the age is 57. The analysis of the nuclear density map shows that the Aitkenian impact craters are concentrated in four regions, located in Mare Orientale, Mare Humorum, Mare Serenitatis and Mare Marginis areas respectively, with a total of two nuclear sites (Table 3.1.1) located in Mare Humorum and Mare Marginis respectively (the red part in the figure), and the nuclear area of the Mare Marginis is larger than that of the Mare Humorum, indicating that the Aitkenian impact craters are clustered on a large scale

in the Mare Marginis, with the same high concentration in the Mare Humorum, but relatively small scale.

The total number of impact craters in Nectarian is 185, and the number used to date is 156. On the whole, the Nectarian impact craters are mainly located in the southern hemisphere and form two cores in the Mare Spumans and Mare Australe area (table 3.1.2). According to (figure 3.1.2), Nectarian impact craters are clustered on a large scale in the Mare Spumans area and on a small scale Mare Australe area. In Mare Nectaris and Mare Nubium regions, some impact craters are clustered, but the degree of aggregation is relatively scattered with respect to the nuclear locations. There are a few scattered Nectarian impact craters in the Mare Orientale, Mare Cognitum and other regions.

The total number of impact craters in the Late Imbrian is 4994, and the number of impact craters used to date is 4449. On the whole, they are distributed in nearly the whole mare regions, mainly with four cores (Table 3.1.3), of which the largest one is

Figure 2.3: Relationship between the age of the impact craters and the number of geological units, the number of impact craters and the impact areas, respectively

Mare Location Area (Km2) Mare Humorum 46.35 W 25.18 S 65022.03Mare Marginis 83.85 E 12.87 N 112938

Table 3.1.1: Cores of KDE of Aitkenian craters

Mare Location Area (Km2) Mare Spumans 65.28 E 1.55 N 114333.6Mare Australe 79.72 E 40.65 S 76427.02

Table 3.1.2: Cores of KDE of Nectarian craters

Mare Location Area (Km2) Mare Imbrium 17.3 W 35.9 N 144840.9

Mare Serenitatis 22.2 E 27.47 N 267026.8Mare Crisium 62.2 E 16.07 N 19116.33Mare Australe 88.43 E 45.82 S 104671.7

Table 3.1.3: Cores of KDE of Late Imbrian craters

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20198

Figure 3.1.1: The map of KDE of Aitkenian craters (base image is Chang’ e - 1 DEM data, the bandwidth is 300 km)

Figure 3.1.2: The map of KDE of Nectarian craters (base image is Chang’ e - 1 DEM data, the bandwidth is 300 km)

Figure 3.1.3: The map of KDE of Late Imbrian craters (base image is Chang’ e - 1 DEM data, the bandwidth is 300 km)

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 20199

Figure 3.1.4: The map of KDE of Early Eratothenian craters (base image is Chang’ e - 1 DEM data, the bandwidth is 300 km)

Figure 3.1.5: The map of KDE of Late Eratothenian craters (base image is Chang’ e - 1 DEM data, the bandwidth is 300 km)

Figure 3.2.1: Statistical Chart of Latitude and Latitude Distribution of Aitkenian craters

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 201910

Figure 3.2.2: Statistical Chart of Latitude and Latitude Distribution of Nectarian craters

Figure 3.2.3: Statistical Chart of Latitude and Latitude Distribution of Late Imbrian craters

Figure 3.2.4: Statistical Chart of Latitude and Latitude Distribution of Early Eratothenian craters

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 201911

located in Mare Serenitatis area and the other three are located in the Mare Imbrium, Mare Crisium and the Mare Australe. Among the four nuclear sites, the one in Mare Crisium is the smallest. The total area of the three small cores is not as large as that in Mare Serenitatis. It can be seen from this that a large number of impact craters in the Late Imbrian are clustered in Mare Serenitatis area, and there are also impact craters in the areas of Mare Imbrium, Mare Crisium and the Mare Australe, but the cluster area is much smaller than Mare Serenitatis area.

Early Eratothenian contains 552 impact craters, of which 505 are used for dating. According to the kernel density distribution map, there is only one core (Table 3.1.4), which is located in Mare Imbrium area, and all of the 552 craters are distributed in Mare Imbrium area with relatively scattered distribution and low kernel density.

The total number of impact craters included in Late Eratothenian is 12, and the total number of impact craters used to determine the age is 8. The number is small, but the distribution is regular. From the kernel density map, we can see that there are two cores (Table 3.1.5) located in Mare Imbrium and Mare Anguis regions respectively. The two cores are close in area and symmetrical in distribution.

Distribution in Latitude and Longitude

Using the latitude and longitude coordinate information of impact craters contained in LU106016 database, the impact craters of different periods in mare area were statistically classified, and the distribution characteristics of impact craters in latitude and longitude in each period were obtained.

Aitkenian impact craters are distributed evenly in longitude, mainly concentrated in four sections, corresponding to the four distribution areas in the kernel density map, with the largest number of impact craters in the range of - 60o to - 30o and 60o to 105o, especially in the range of 75o to 90o, accounting for about one-third of the total number of Aitkenian impact craters. In latitude, the northern hemisphere contains slightly more craters than the southern hemisphere, and the distribution is relatively balanced, mainly concentrated in the range of three segments, with the largest number of impact craters distributed in the range of - 30o to - 20o, accounting for about one third of the total number of Aitkenian impact craters.

The Nectarian impact craters are distributed in the four regions in longitude, corresponding to the four regions distributed in the kernel density map, in both the eastern and western hemispheres, but the number of impact craters in the western hemisphere is significantly larger than that in the eastern hemisphere, especially in the range of 60o to 80o, and the number of impact craters is about half of the total number of Nectarian impact craters. In latitude, the craters are continuously and unevenly distributed in the range of -60o

to 10o, and are distributed in both the northern and southern hemispheres, but unevenly distributed, and the number of impact craters in the southern hemisphere is significantly larger than that in the northern hemisphere.

The Late Imbrian impact craters are distributed continuously in longitude, mainly between - 40o and 100o. On the whole, the impact craters are mostly distributed in the eastern hemisphere, and the number of impact craters increases first and then decreases gradually from the western hemisphere to the eastern hemisphere, with the highest concentration at the boundary between the eastern and western hemispheres, especially between -20o to 0o. In latitude, the craters are distributed between -70o to 70o, but the number in the northern hemisphere is significantly larger than that in the southern hemisphere. With the equator as the center, the number of craters first increases and then decreases from the southern hemisphere to the northern hemisphere, then increases again

Figure 3.2.4: Statistical Chart of Latitude and Latitude Distribution of Late Eratothenian craters

Mare Location Area (Km2) Mare Imbrium 22.57 W 34.63 N 21801.45

Table 3.1.4: Cores of KDE of Early Eratothenian craters

Mare Location Area (Km2) Mare Imbrium 21.05 W 42.98 N 19072.54Mare Anguis 65.52 E 26.32 N 22046.01

Table 3.1.5: Cores of KDE of Late Eratothenian craters

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 201912

at the equator, then decreases gradually further north, and become smaller when it is closer to the poles.

The early Eratothenian impact craters are distributed continuously from -40o to 10o in longitude, mostly in the western hemisphere, especially in the range of -30o to -15o, there are more than two-thirds of the total number of early Eratothenian impact craters. In latitude, the impact craters are all distributed in the northern hemisphere, continuously and intensively distributed between 18o to 48o, especially in the range of 30o to 36o, the number of impact craters distributed in this range is more than half of the total number of early Eratothenian impact craters.

The late Eratothenian impact craters are concentrated in two areas in latitude and longitude, showing a symmetrical distribution trend. Longitudinally, the craters are symmetrically distributed around the 20o meridian, respectively in the range of -40o to -20o and 60o to 80o, with the total number of impact craters in both ranges being 6. In terms of latitude, the impact craters are distributed in the northern hemisphere, mainly symmetrically distributed in the latitude coordinates of 35o latitude, in the range of 20o to 25o and 45o to 50o respectively, and the total number of impact craters in both ranges is 6.

ConclusionBased on the LU106016 impact crater database, this paper analyzes the spatial distribution characteristics of impact craters in different periods, and draws the following conclusions:

1. The geological units of lunar mare area can be divided into five periods, namely Aitkenian, Nectarian, Late Imbrian, Early Eratothenian and Late Eratothenian, in which Late Imbrian contains the largest number of impact craters, the largest impact area and the widest distribution.

2. Aiken impact craters are mainly distributed in the Mare Orientale, Mare Humorum, Mare Serenitatis and Mare Marginis, among which there are a large number of concentrated distributions in the Mare Humorum and Mare Marginis areas, and the latitude and longitude distribution is relatively balanced.

3. The impact craters in Nectarian Period are clustered at the height of the Mare Spumans and Mare Australe, while some impact craters are clustered in the region of Mare Nectaris and Mare Nubium, but the degree of aggregation is more dispersed than that in the Mare Spumans and Mare Australe area. In terms of latitude and longitude distribution, the distribution of impact craters is extremely uneven, the number of impact craters in the western hemisphere is significantly larger than that in the eastern hemisphere, and in latitude, the number of impact craters in the southern hemisphere is significantly larger than that in the northern hemisphere.

4. The Late Imbrium impact crates are distributed in the whole lunar mare area, especially in Mare Serenitatis, Mare Imbrium, Mare Crisium and the Mare Australe.

Overall, the spatial distribution is uneven. The number and concentration of Late Imbrium impact crates in the northern hemisphere is higher than that in the southern hemisphere, and the number and concentration in the eastern hemisphere is higher than that in the western hemisphere.

5. The impact craters of the early Eratothenian are evenly distributed in Mare Imbrium area, and they are relatively scattered mainly in the western and northern hemispheres.

6. The number of impact craters in Eratothenian is very small, but the distribution isvery regular, they are mainly distributed in the Mare Imbrium and Mare Anguis area, with 20o longitude and 35o latitude as axisymmetric distribution respectively.

References1. Heiken G, Vaniman D, French B M (1991) Book-Review - Lunar

Sourcebook - a User’s Guide to the Moon. Sky & Telescope, 254. [View Article]

2. OUYANG Zi-yuan (2005) Introduction to Lunar Science [M] Beijing: China Aerospace Publishing House. [View Article]

3. Jolliff BL, Gillis JJ, Haskin L A, et al. (2000) Major lunar crustal terranes: Surface expressions and crust-mantle origins. Journal of Geophysical Research 105: 4197-4216. [View Article]

4. Wood JA, Dickey JS, Marvin UB, et al. (1970) Lunar anorthosites and a geophysical model of the moon. Geochimica Et Cosmochimica Acta Supplement 1: 965. [View Article]

5. Cadogan PH (1974) Oldest and largest lunar basin? Nature 250: 315-316. [View Article]

6. Miljkovićć K, Wieczorek MA, Collins GS, et al. (2013) Asymmetric distribution of lunar impact basins caused by variations in target properties. Science 342: 724-726. [View Article]

7. Yao Mei-juan, CHEN Jian-ping, WANG Xiang, et al. (2016) The Grading and Evolution Analysis of Lunar Crater Based on Optimum Partition and Grading Method. Acta Petrologica Sinca 32: 119-126. [View Article]

8. WANG Jiao, ZHOU Cheng-hu, CHENG Wei-ming (2017) The Spatial Pattern of Lunar Craters on a Global Scale. Geomatics and Information Science of Wuhan Uniersity 42: 512-519. [View Article]

9. YAO Mei-juan, CHEN Jian-ping, XU Bin, et al. (2017) Difference Comparison and Evolution Analysis of Lunar Nearside and Farside. Journal of Earth Sciences and Environment 39: 428-438. [View Article]

10. Head JW (1976) Lunar volcanism in space and time. Reviews of Geophysics 14: 265-300. [View Article]

11. Heiken GH, Vaniman DT, French BM (1991) Lunar sourcebook: A User’s Guide to the Moon [M]. Cambridge: Cambridge University Press. [View Article]

12. Kreiter TJ (1960) DATING LUNAR SURFACE FEATURES BY USING CRATER FREQUENCIES. Publications of the Astronomical Society of the Pacific 72: 393. [View Article]

13. Bell SW, Thomson BJ, Dyar MD, et al. (2012) Dating small fresh lunar craters with Mini-RF radar observations of ejecta blankets. Journal of Geophysical Research Atmospheres 117: E00H30. [View Article]

Yuan ling WU (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters

J Geol Geosci Volume 3(2): 201913

14. Hartmann WK, Davis DR (1975) Satellite-sized planet esimates and lunar origin. Icarus 24: 504-515. [View Article]

15. Hartmann WK, Clifford SM (2005) Martian cratering 8; isochron refinement and the chronology of Mars. Icarus 174: 294-320. [View Article]

16. ZHAO Jian-nan, HUANG Jun, XIAO Long, et al. (2013) Crater Size-Frequency Distribute Measurements and Age Determination of Sinus Iridum. Earth Science- Journal of China University of Geosciences 38: 351-361. [View Article]

17. Neukum G, Ivanov BA (1994) Crater Size Distributions and Impact Probabilities on Earth from Lunar, Terrestrial-planet, and Asteroid Cratering Data. pp. 359-416. [View Article]

18. Melosh HJ, Vickery AM (1989) Impact erosion of the primordial atmosphere of Mars. Nature 338: 487-489. [View Article]

19. Neukum G (2001) Cratering Record of the Inner Solar System in Relation to the Lunar Reference System. Springer Netherlands. [View Article]

20. LI Yan-qiu, CHEN Sheng-bo, GUO Peng-ju, et al. (2014) Crater Size-Frequency Distribution Measurements and the Secondary Craters Features of Lunar High-Lands. Earth Science Frontiers 21: 38-44. [View Article]

21. LI Kun, LIU Jian-jun, MOU Ling-li, et al. (2012) Crater Retention Age Calculation of Lunar Geological Unit Using Small- Scale Craters. Journal of Jilin University (s2): 452-459. [View Article]

22. XIONG Kai-ling, PENG Jun-jie, YANG Xiao-fei, et al. (2017) K- means Clustering Optimization Based on Kernel Density Estimation. Computer Technology and Development 27:1-5. [View Article]

23. Anderson TK (2009) Kernel density estimation and K-means clustering to profile road accident hotspots. Accident Analysis & Prevention 41: 359-364. [View Article]

24. YAN Qing-wu, BIAN Zheng-fu, ZHAO Hua (2005) A Method of Spatialization of Population Density. Geography and Geo- Information Science 21: 45-48. [View Article]

25. ZHANG Gui-ming, ZHU A-xing, Yang Sheng-tian, et al. (2013) Mapping Wildlife Habitat Suitability Using Kernel Density Estimation. Acta Ecologica Sinica 33: 7590-7600. [View Article]

26. LU Min, YANG Liu, WANG Jin-yin, et al. (2017) Applications of Pointgroup Density Cartography Based on Kernel Density Estimation. Engineering of Survering and Mapping 26: 70-74. [View Article]

27. Xie Z, Yan J (2008) Kernel Density Estimation of traffic accidents in a network space. Computers Environment & Urban Systems 32: 396-406. [View Article]

28. Xie Z, Yan J (2013) Detecting traffic accident clusters with network kernel density estimation and local spatial statistics: an

integrated approach. Journal of Transport Geography 31: 64-71. [View Article]

29. Parzen E (1962) On estimation of a probability density function and mode. Ann Math Statis 33:1065-1076. [View Article]

30. Rosenblatt M (1956) Remarks on Some Nonparametric Estimates of a Density Function. Annals of Mathematical Statistics 27: 832-837. [View Article]

31. Haben S, Giasemidis G, Hyndman RJ (2016) A hybrid model of kernel density estimation and quantile regression for GEFCom2014 probabilistic load forecasting. International Journal of Forecasting 32:1017-1022. [View Article]

32. Silverman BW (1986) Density estimation for statistics and data analysis. Chapman and Hall [View Article]

33. O’Brien, Travis A, Kashinath, et al. (2016) A fast and objective multidimensional kernel density estimation method: fastKDE. Computational Statistics & Data Analysis 101:148-160. [View Article]

34. Pieters CM, Staid MI, Fischer EM, et al. (1994) A Sharper View of Impact Craters from Clementine Data. Science 266:1844-1848. [View Article]

35. Fischer EM, Pieters CM (1994) Remote Determination of Exposure Degree and Iron Concentration of Lunar Soils Using VIS-NIR Spectroscopic Methods. Icarus 111: 475-488. [View Article]

36. LUO Lei, WANG Xin-yuan, GUO Hua-dong, et al. Automatic Detection of Lunar Craters Based on CE-1 DEM Data. Journal of Remote Sensing [View Article]

37. Robinson MS, Bowmancisneros E , Brylow SM, et al. (2009) LROC - Lunar Reconnaissance Orbiter Camera[C]// Agu Fall Meeting. AGU Fall Meeting Abstracts [View Article]

38. Wang J, Cheng W, Zhou C (2015) A Chang\”E-1 global catalog of lunar impact craters. Planetary and Space Science 112: 42-45. [View Article]

39. WANG Jiao, CHENG Wei-ming, ZHOU Cheng-hu, et al. (2014) Identification and Morphologic Expression of Lunar Impact Craters. Geographical Research 33:1251-1263. [View Article]

40. GUO Di-jun, LIU Jian-zhong, ZHANG Li, et al. (2014) The Methods of Lunar Geochronology Study and the Subdivisions of Lunar Geologic History. Earth Science Frontiers 21: 45-61. [View Article]

41. OUYANG Zi-yuan, LIU Jian-zhong (2014) The Origin and Evolution of the Moon and its Geological Mapping. Earth Science Frontiers 21: 1- 6. [View Article]

42. Rui L, Wei-Ping HU, Hong-Liang W, et al. (2011) The Road Network Evolution of Guangzhou-Foshan Metropolitan Area Based on Kernel Density Estimation. Scientia Geographica Sinica 30: 81-86. [View Article]

Citation: Yuan ling WU, Jian-ping CHEN (2019) Analysis of Temporal and Spatial Distribution Characteristics of Lunar Mare Craters. J Geol Geosci 3: 001-013.

Copyright: © 2019 Yuan ling WU, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.