Urbanization, Populations Growth, and Built-up Area Expansion of Java from 1970 to 2020
The development of the built-up area from 1970 to 2020 was analyzed using satellite imagery. In the 1970s and 1980s, the built-up area was not densely packed and the dominance was only in Jakarta. In 1990 and 2000 the built-up area began to spread in several big cities besides Jakarta, such as Bandung, Semarang, Solo, Yogyakarta and Surabaya. In addition, the development of built-up areas began to spread in the north of Java. Built-up areas in Central Java and East Java showed a large increase compared to West Java in 2010 and 2020, with development spreading across the North and South. However, in general, the Jakarta metropolitan area is the Area with the most dominating development from 1970 to 2020. This is in line with research conducted by Buchori, et al. (2017) regarding the theory of spatial dynamics of the metropolitan area with case studies on the islands of Java and Madura.
The total built-up Area on Java Island in 1970 was 1,873.53 km2, when compared to 2020, the increase in built-up Area on Java Island was 34,511.04 km2. East Java Province is the province that dominates built-up Area on Java Island (42.95%) then Central Java Province (30.6%) (Table 4). Changes in the built-up Area that continued to increase from 1970 to 2020 were in line with the increase in population (Table 5).
The increase in population is always accompanied by the development of an area, economic improvement and land cover conversion (Kumar and Sangwan 2013). Land use will increase to meet the population's needs for housing, infrastructure, and other built-up Area. The total population recorded in 2020 was 139,686,700 or almost double that of 1970. This condition also shows that more than 50% of Indonesia's population is concentrated on Java Island.
The population growth from 1970-1980 was quite significant (Figure 3). Even in Jakarta, population growth reached 42.02%. In 1980-1990, population growth decreased compared to the previous decade. It is due to the Government of Indonesia issued a Family Planning program in the 1980s. This program was successfully implemented as the population growth in 1990-2000 decreased significantly. However, in 2000-2010, the population of Java increased again. The population growth of West Java Province in 2000 was 0.98%, while in 2010 it was 20.50%. In general, population growth in Java has increased in the last five decades.
Table 4 Development of Built-up Area in Java by Province 1970-2020
Province
|
Built-up Area (km2)
|
1970
|
1980
|
1990
|
2000
|
2010
|
2020
|
Banten
|
-
|
-
|
-
|
404.19
|
1,286.19
|
2,746.60
|
Jakarta
|
42.18
|
114.54
|
144.76
|
242.88
|
355.65
|
548.21
|
Yogyakarta
|
|
|
|
|
|
|
West Java
|
53.95
|
65.07
|
307.39
|
416.41
|
420.86
|
1,690.00
|
Central Java
|
515.394
|
1,108.8672
|
2,254.171
|
3,753.19
|
4,758.73
|
7,366.76
|
East Java
|
562.25
|
1,348.82
|
1,389.39
|
2,413.26
|
4,553.60
|
11,149.64
|
Banten
|
699.76
|
1,021.84
|
2,763.01
|
3,207.22
|
5,236.27
|
15,629.96
|
Total
|
1,873.53
|
3,659.13
|
6,858.71
|
10,032.96
|
15,325.11
|
36,384.57
|
Table 5. Total Population of Java Island in 1970-2020
Year
|
1970
|
1980
|
1990
|
2000
|
2010
|
2020
|
Province
|
Banten
|
-
|
-
|
-
|
8,098,780
|
10,632,166
|
11,904,562
|
Jakarta
|
2,489,360
|
2,750,813
|
2,913,054
|
3,122,268
|
3,457,491
|
3,668,719
|
Yogyakarta
|
4,579,303
|
6,503,449
|
8,259,266
|
8,389,443
|
9,607,787
|
10,562,088
|
West Java
|
21,623,529
|
27,453,525
|
35,384,352
|
35,729,537
|
43,053,732
|
48,274,162
|
Central Java
|
21,877,136
|
25,372,889
|
28,520,643
|
31,228,940
|
32,382,657
|
36,516,035
|
East Java
|
25,516,999
|
29,188,852
|
32,503,991
|
34,783,640
|
37,476,757
|
40,665,696
|
Total
|
76,086,327
|
91,269,528
|
107,581,306
|
113,253,828
|
125,978,424
|
139,686,700
|
The slope map shows that the central part of Java Island has a slope between 25%-40% while the northern part of Java Island is almost entirely flat (0%-8%) (Figure 4a). It shows that the central and southern parts of Java Island are mountains or hills and the northern part of Java Island is lowlands. A distinctive characteristic of water is that it flows from a high place to a lower place. Therefore, rivers in Java tend to flow from the central part to the northern part of Java Island. The volume of water in the northern part will increase if the volume of surface runoff in the central part of Java Island increases. Rainwater runoff increases when there is a change in land cover from non-built-up to built-up Area. This happens because rainwater cannot be absorbed into the ground.
The map of the built-up area development in Figure 2 shows that there was a change from non-built-up to built-up Area in the central part of Java Island from 1970 to 2020. This condition increases the volume of rainwater runoff towards the northern part of Java Island. It is evidenced by the increase in the red grid (higher-high) on the flood distribution map (Moran Scatterplot) in the North of Java Island in 2010 and 2020 (Figure 6).
Flood distribution map using Moran Scatterplot can show the average distribution pattern of flood events in an area. The pattern can be dispersed, random, or cluster. The grid size used is 500 x 500 meters. When a grid in which there are > 4 flood points and adjacent to a similar grid it will form a red grid (higher-high). Meanwhile, the yellow grid (higher-low) depicts a grid with low flood points (1-4 points) adjacent to the grid with high flood points (> 4 points). Flood events on the north coast of Java Island form a cluster pattern, where the higher-high grids are close together to form one group. The red grid increased from 2010 to 2020, it means that in 2010 the grid with low flood points changed to the high category within a decade.
In addition, slopes and river patterns can also show the relief of the earth's surface. The location of the centrifugal radial pattern is usually on the mountain because the pattern spreads from a peak. The pattern can be seen in the southern part of Jakarta, specifically in Bogor (Figure 4b). The area is the upstream of Jakarta. If the land cover/ land use in upstream conditions change, the downstream areas will also be affected. The hydrological conditions in the upstream Ciliwung watershed have changed due to climate change and land use. Any changes in this Area can increase the frequency of flood which has high consequences in the downstream watershed where Jakarta is located (Emam et al., 2016). In the map of the development of built-up land, it is evident that there has been an expansion of built-up land towards the hinterland of Jakarta, including Bogor, from 1970-2020. Therefore, the number of flood events in 2020 in the Jakarta Metropolitan Area is quite large up to 82 flood events.
There are many river branches in Subang Regency and Indramayu Regency which are part of the Cipunegara watershed. This river flows from North Bandung to the Java Sea. Based on the map of the built-up area development, it can be seen that land cover change in Greater Bandung is quite fast. The two regencies often experience floods. In 2020, there are 102 flood points in Subang Regency. This makes Subang Regency the location with the most flood points in West Java.
Jakarta and West Java and Central Java Province also experiences frequent flooding, especially in Pekalongan, Demak, Semarang, and Kendal. In addition to the upstream part of the river, land subsidence is also a driving factor for floods. According to Azeriansyah & Harintaka (2019), overall, the Semarang City area has experienced land subsidence from 0 to 6.753 cm/year.
There are also many flood events in several regencies/cities in the Surabaya Metropolitan Area, such as in Surabaya City, Gresik Regency, Sidoarjo Regency, and Lamongan Regency. The Bengawan Solo watershed plays a vital role in flood events in Lamongan and Gresik regencies. The map of the built-up area development shows the built-up Area in Solo is increasing and continues to grow towards Sukoharjo, which is currently known as Solo Baru (the new solo). It can be a factor in the occurrence of floods in the Surabaya Metropolitan Area.
Spatial Autocorrelation-Moran Index
The value of the Moran index can indicate the presence or absence of spatial autocorrelation in a study area. This value can also be positive or negative. The Moran index value is in the range of -1 to +1. The closer the value to -1 or +1, the higher the autocorrelation. If the Moran index value is at 0 <I ≤1, the spatial autocorrelation is positive. On the other hand, if the Moran index value is -1≤ I<0, the spatial autocorrelation is negative. If the Moran index value is 0, there is no spatial autocorrelation.
In addition to the Moran index value, the value of Z count (Z(I)) can also indicate the presence or absence of spatial autocorrelation in the observed Area. The value of used is 5%, so (Z(α)) is 1.96. If Z(I) > Z(α), there is spatial autocorrelation in the observed area. On the other hand, the value of Z(I) < Z(α) means that there is no spatial autocorrelation.
Using the Moran Scatterplot, the flood distribution map can show the average distribution pattern of flood events in an area. The pattern can be dispersed, random, or cluster. The grid size used is 500 x 500 meters. Based on the value of the Moran index, the average distribution pattern of flood events on the north coast of Java Island is illustrated in Figure 6.
The value of the Moran index on the distribution of floods in Java in 1970 is -0.000225 and the value of Z(I) is 142.83 > Z(α), it means that there is a negative spatial autocorrelation. This shows that adjacent locations have different values so that the spatial pattern tends to spread. The 1970 Moran Scatterplot depicts a blue grid (lower-low) only in greater Jakarta.
The value of the Moran index in 1980 was -0.000372 and the value of Z(I) was 71.38 > Z(α). Same as in 1970, the Moran index value means that there is a negative spatial autocorrelation. This is shown by the Moran Scatterplot, where the spatial pattern is spread out. There is a blue grid (lower-low) in Jakarta and a red grid (higher-high) in East Java.
The Moran Scatterplot in 1990 has begun to show a grouping of the average flood events on the north coast of Java Island. It can be seen in Figure 8 that the yellow (lower-high) grid began to appear and form clusters in West Java, Central Java, and East Java. This is reinforced by the Moran index of 0.042 and the value of Z(I) of 10.37 > Z(α). This means that there is a positive spatial autocorrelation with spatial patterns that tend to be clusters.
In 2000, red (higher-high) and yellow (lower-high) grids increased. The grid forms a cluster pattern. The red grid forms groups in Jakarta and Central Java. Meanwhile, the yellow grid groups are located in Banten, Jakarta, Central Java, and East Java. This also shows that the incidence of flooding on the northern coast of Java Island is increasing compared to the previous few decades. The value of the Moran index in 2000 was 0.1462 and Z(I) was 18.43 > Z(α), meaning that according to the Moran Scatterplot, there is a positive spatial autocorrelation with a cluster pattern.
The 2010 Moran Scatterplot also forms a cluster pattern. It can be seen that the red (higher-high) and yellow (lower-high) grids form clusters in several cities/districts on the north coast of Java Island. Moran index value also shows the same thing. The value of the Moran index in 2010 was 0.3478 and the value of Z(I) was 42.09 > Z(α) so that there was a positive spatial autocorrelation that formed a cluster pattern. Figure 8 shows more red grids in 2020 than in the previous decade. In addition, the grid also forms a large group on the north coast of Java Island. This shows that more and more flood points are located in adjacent locations. The value of the Moran index in 2020 is positive and close to +1, which is 0.5248. The value of Z(I) is also 63.69 > Z(α). This means that in 2020 there will be a high positive spatial autocorrelation with a clustered pattern.
In general, the Moran Scatterplot shows that the incidence of flooding on the north coast of Java Island is increasing from 1970 to 2020. Spatial patterns that initially spread, starting in 1990, these patterns formed clusters. This is also in accordance with population growth and the development of built-up land on the Island of Java. The more people and the wider LULCC become built-up land, the greater the number of flood events on the Island of Java.
Based on Pearson's Bivariate Correlation Analysis, the relationship between the number of flood points with the number of residents and the Area of built-up land can be known. The flood point will be the dependent variable while the population and land area will be the independent variables. The relationship between variables can be seen from the value of Sig. (2-tailed) and Pearson Correlation scores.
Table 6 Moran I-Analysis
Years
|
I (Moran's Index)
|
z-score
|
p-value
|
Result
|
2020
|
0.5348
|
63.5316
|
0
|
Clustered
|
2010
|
0.3478
|
41.7881
|
0
|
Clustered
|
2000
|
0.1462
|
17.3193
|
0
|
Clustered
|
1990
|
0.042
|
5.1041
|
0
|
Clustered
|
1980
|
-0.000372
|
-0.03599
|
0.9712
|
Random
|
1970
|
-0.000225
|
-0.014991
|
0.988
|
Random
|
Correlation Analysis
Based on the Significance Value of Sig. (2-tailed): Given the value of Sig. (2-tailed) between Population (X1) and Flood Point (Y) is 0.029 < 0.05, meaning that there is a significant correlation between the population and the flood point. Given the value of Sig. (2-tailed) between Built-Up Area (X2) and Flood Point (Y) is 0.037 < 0.05, meaning that there is a significant correlation between the Area of the built-up area and the flood point.
Based on the calculated r-value (Pearson Correlations): It is known that the estimated r value for the relationship between Population (X1) and Flood Point (Y) is 0.857 > 0.811 (r table because N = 6). It means that there is a relationship or correlation between the population and the flood point. Because r arithmetic or Person Correlations is positive, the relationship between the two variables is positive. As the population increases, the number of flood points also increases. The calculated r-value between the Built-Up Area (X2) and the Flood Point (Y) is 0.840 > 0.811 (r table N = 6), meaning that there is a relationship or correlation between the Area of the built-up area and the flood point. Because r arithmetic or Person Correlations is positive, the relationship between the two variables is positive. As the Area of built-up increases, the number of flood events also increases.
Table 7 Pearson Bivariate Correlation Analysis
Correlations
|
|
Population
|
Built_Up_Area
|
Flood_Point
|
Population
|
Pearson Correlation
|
1
|
.879*
|
.857*
|
Sig. (2-tailed)
|
|
.021
|
.029
|
N
|
6
|
6
|
6
|
Built_Up_Area
|
Pearson Correlation
|
.879*
|
1
|
.840*
|
Sig. (2-tailed)
|
.021
|
|
.037
|
N
|
6
|
6
|
6
|
Flood_Point
|
Pearson Correlation
|
.857*
|
.840*
|
1
|
Sig. (2-tailed)
|
.029
|
.037
|
|
N
|
6
|
6
|
6
|
*. Correlation is significant at the 0.05 level (2-tailed).
|