Analysis of land cover change in Sagara Anakan Cilacap, Central Java using principal component analysis (PCA)

Land cover in area Mangrove, an ecosystem with high productivity, is an ecological link that is very important for life in the waters. This study aims to identify land cover vegetation, water bodies, and soil using Principal Component Analysis (PCA) method with imagery using Landsat imagery for 1997 and 2021 carried out by processing Landsat 7 and Landsat 8 OLI images consisting of pre-processing processes and image classification. The results showed that in the mangrove area in Sagara Anakan, there was a significant change in land cover; for the lagoon area, the land cover was overgrown with mangrove forests, but areas far from the lagoon had changed to settlements or sparse vegetation.


Introduction
The ability to recognize various objects on the Earth surface using remote sensing technology is expanding.Based on the wavelengths from plant components reflection, remote sensing can monitor terrestrial and coastal vegetation [1].[2]The ecological connection between mangroves and other highproductivity ecosystems is crucial for aquatic life.As a transition zone between land and sea due to sea tides, resulting in highly significant environmental impacts, particularly for temperature and salinity [3].Consequently, it is crucial to safeguard the mangrove environment [4].Mangrove roots grow in moist and muddy areas to prevent erosion and abrasion of the coast and retained sediments and soil will cause the development of the coastline from time to time [5].Changes in mangrove land cover from year to 1315 (2024) 012046 IOP Publishing doi:10.1088/1755-1315/1315/1/012046 2 year have decreased a large area, based on the function of the mangrove ecosystem has an impact on coastal abrasion, decreasing the quality of coastal water and the destruction of several living species in this environment, as a coastal fortress that functions to resist erosion caused by waves or tidal currents.Therefore, the transformation of ecosystem functions can cause problems in the mangrove ecosystem For this reason, the intercorrelation can be decreased by Principal Component Analysis (PCA).To maximize the original band information in the main components, PCA seeks to minimize the dimensionality of the data, in this instance, the original band.Uncorrelated variables (principal components), which contain authentic physical information and can be investigated, are created from a group of correlated variables [6].
The Landsat satellites have supplied the longest temporal record of space-based observations of the ground surface for more than 40 years, and the successful launch of Landsat-8 in 2013 upholds this tradition.The Landsat sensor suite's data records should ideally be consistent.Compared to the previous Landsat-7 Enhanced Thematic Mapper (ETM+), the Landsat-8 Operational Land Imager (OLI) offers better calibration, signal-to-noise characteristics, a greater 12-bit radiometric resolution, and a spectrally smaller band.The variance in reflective wavelengths between the two Landsat sensors is further influenced by air conditions and surface reflectance, which are challenging to fully model.Edge swath pathways that are sensed only one day apart are provided by the orbits and sensing geometries of the Landsat-8 OLI and Landsat-7 ETM+.With this set up, a total of 59 million 30 [7].Eigenvalues and eigen factors can be found in PCA.The eigenvalues add up to 100%.There is no need to add the next PCA if the sum of the PCA eigenvalue proportions is >= 85% because it adequately reflects > 85% of the data that needs investigation.Data from time series can also be analyzed using PCA.
Segara Anakan is the mouth of several rivers, such as Citanduy, Cibereum, Cikonde and Cimeneng.The flow of these rivers flows a high sedimentation and nutrient input into Segara Anakan, there by affecting the area of mangroves in the region [8].In Segara Anakan, there are 12 mangrove species of 4 major, one minor, and two associated mangrove families [9].Segara Anakan has considerable potential for mangrove conservation, but in 2013 there was a decrease in the distribution and area of mangroves compared to previous years, with an area of 6,716 Ha [10].Changes in the cover area of Segara Anakan are due to various pressures such as sedimentation, deforestation and land conservation.Sedimentation influences the area of Segara Anakan.The Mangrove area has changed from 6,450 ha in 1903 to 1,800 ha in 1992.This situation continues to change every year.[11].The shrinkage of the Sagara Anakan Lagoon is due to the high sediment entering the lagoon [12].This study aims to identify changes of land cover vegetation, water bodies and soil using PCA imagery using Landsat imagery acquired in 1997 and 2021.

Methodology
The method used in this study is quantitative descriptive with a spatial approach.The tools used in this study were personal computers equipped with the 2015 version of ERDAS Imagine software.The materials used were Landsat 8 OLI images for 2021 and Landsat 7 OLI for 1997 in Sagara Anakan Cilacap.Data analysis, according to Jaya (2005), evaluates areas that change (change) using the concept: • The stable brightness (SB) component is defined when the eigenvector (weight) value of each channel (band) is almost the same as the positive algebraic sign.This index is generally found in the principal component one.• The stable greenness (SG) component is defined if the red band has the same algebraic sign but is opposite to the algebraic sign of the infrared band from both times.For example, the algebraic signs of the two red channels are positive in two different years, while the algebraic signs of the two infrared channels are negative, or vice versa.• The delta brightness component (DB) is characterized by the similarity of the algebraic signs of the red and infrared channels simultaneously but contradictory algebraic signs of the red and infrared channels at different times.For example, the algebraic sign in the previous year for the red and infrared channels is positive, while for the red and infrared channels in the following year, it is negative, or vice versa.` • The Greenness Delta component (DG) is the opposite of SB.For example, the red channel's algebraic sign is positive, and the infrared is negative for the previous year, and the algebraic sign for the year after.In this study, we explore a comprehensive method to monitor changes in vegetation over time.Firstly, we create a stack by aligning and combining two images captured at different times, providing a basis for further analysis.Utilizing the combined multi-temporal bands, we construct a Principal Component (PC) to condense the data and enhance the signal-to-noise ratio.Next, we calculate the eigenvectors from the PC to identify key patterns and variations in the vegetation data.These eigenvectors help us establish an eigenvalue matrix that quantifies the importance of each component in the dataset.With this foundation, we develop an equation capable of estimating changes in Greenness and Brightness, crucial indicators of vegetation health and growth.Finally, we execute the simulation, incorporating the equation to derive valuable insights into the fluctuations in Greenness and Brightness over time.The simulation results are visually presented, enabling a clear and informative display of the dynamic changes occurring within the vegetation landscape.Through this process, we establish a robust approach to monitor and analyze the evolution of vegetation, contributing to our understanding of ecological changes and land development patterns.

Results And Discussion
In tidal situations where this plant community can tolerate salt, such as in protected regions, lagoons, and river estuaries that are submerged at low tide, mangrove forests are a type of forest that can grow.Cilacap Regency has various mangrove forests.There is a fascinating backstory to Segara Anakan Lagoon.With a total size of 21,500 ha, the Segara Anakan Forest is Java's largest mangrove forest.At Create a stack between 2 image at differents times this time, the extent is difficult to predict due to high sedimentation, which has resulted in new plains being invaded by mangroves, as well as many changes to the designation of old mangrove vegetation areas, which have become the area with the widest mangrove cover on the island of Java.The tourism object in the Segara Anakan mangrove ecosystem is a habitat for diverse flora and fauna with distinctive characteristics, as well as the geological conditions of the site and its surroundings as visual objects without leaving the cultural customs of the local community.
Using the PCA method, the results of land cover image management for 1997 and 2021.If we look at Google Earth at 10-year intervals, it will be clear that changes have occurred in the Segara Anakan area.This change was very drastic.In 1995, there were still no mangrove forests which was so significant that it was only in 2016 that the increase in the area of mangrove forests was visible.Up Bred > BNIR on t1 and on t2 become Bred < BNIR (exhibit inconsistent behavior) a Based on the criteria above, a script can be created in MS.Excel as presented in From the simulation results, then classify the eigenvectors using band 3 and band 4.  In figure 5, number 1 is the 2021 Landsat image, number 2 is the 1997 Landsat image, and number 3 is the result of analysis using the PCA method.
The change in mangrove land cover between 1997 and 2021 is the expansion f mangroves in the lagoon.Changes in land cover in the lagoon in Sagara Anakan show that mangroves have been planted around the lagoon, so the mangrove rehabilitation process in 2021 is still ongoing.In 2021 it can also be seen that around this area, there has been a large addition of built-up land around the coastal area because, in 1997, settlements around the coastal area were still rather sparse.The increase in population also affected the addition of built-up land into permanent settlements.In 2021 it can also be seen that there has been an addition of land to become sparse vegetation around the area.It has not been identified whether it is rice fields, mixed gardens or empty land because in 1997, it was identified as dense vegetation or still filled with plants.

Conclusion
PCA is very good at explaining land cover, which is useful in land cover management because it can easily show the changing points of greenness and brightness.Those far away from the lagoon have changed to settlements or sparse vegetation.The dynamics of nature and human life include changes in land use and cover.This study will give an outline of the scientific scope (science domain) of the topic of land cover and use change because detailed understanding regarding the causation, magnitude, and spatial distribution of the processes of change in land cover and use is still far from adequate.It will outline the still-plentiful research potential for sensory and GIS applications for investigations of changes in land cover and use.

Figure 1 .
Figure 1.PC-Based Approach for Estimating Land Cover Change.
Create a PC from the combined multi-temporal band earlier Calculate the eigenvectors Create an eigenvalue matrix Create an equation for estimating changes in Greenness and Brightness Run the simulation results and display input data

Figure 4 .
Figure 4. Changes in Land Cover as a Result of PCA Analysis.

Figure 5 .
Figure 5. Changes in land cover as a result of PCA.

Figure 6 .
Figure 6.Changes in Land Cover Around The Lagoon (number 1 is data landsat 2021 and number 2 is data landsat 1997).

Figure 7 .
Figure 7. Change of Land Cover to Settlement (number 1 is data landsat 2021 and number 2 is data landsat 1997)

Figure 8 .
Figure 8. Change of Land Cover to Sparse Vegetation (number 1 is data landsat 2021 and number 2 is data landsat 1997)

Table 1 .
Land Cover Estimation Simulation Results.Based on the table above, a comparison relationship is made to see changes.There are several decision-making criteria, as presented in the following table.

Table 2 .
Decision-Making Criteria for Land Cover Change.

Table 3
Bands 3 and 4 Analysis Results.